TURKEY
ECONOMIC
MONITOR
APRIL 2021
NAVIGATING
THE WAVES
TURKEY ECONOMIC MONITOR,
APRIL 2021:
NAVIGATING THE WAVES
© 2021 International Bank for Reconstruction and Development / The World Bank
1818 H Street NW
Washington DC 20433
Telephone: 202-473-1000
Internet: www.worldbank.org
Standard Disclaimer:
This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank.
The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the
Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the
accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any
map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory
or the endorsement or acceptance of such boundaries.
Copyright Statement:
The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without
permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The
World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work
promptly.
For permission to photocopy or reprint any part of this work, please send a request with complete information to the
Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-
750- 4470, http://www.copyright.com/.
All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher,
The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail
pubrights@worldbank.org.
Contents
EXECUTIVE SUMMARY ............................................................................................................................................... i
I. TAKING STOCK .................................................................................................................................................... 1
A. COVID-19 determined the course of the year worldwide and in Turkey over 2020 ................................ 1
Countries worldwide struggled to control COVID-19, while supporting their economies .......................... 1
Resurgent COVID-19 had called for re-imposition of control measures by late 2020.................................. 4
B. A sharp rebound in late 2020, accompanied by rising pressures ................................................................... 6
The economy rebounded strongly in the second half… ..................................................................................... 6
…as COVID-19 restrictions were relaxed .......................................................................................................... 10
..and the government unveiled a large fiscal package and supported a wave of credit ................................ 12
Turkey faced severe external pressure, and the Lira depreciated sharply ....................................................... 13
Inflation accelerated as the effects of monetary and credit stimulus played out ........................................... 16
Corporate and bank balance sheets faced pressure, obscured by forbearance measures ............................ 21
C. Policy frameworks were adjusted to restore macroeconomic stability ....................................................... 26
The fiscal policy framework remained strong, despite COVID-19 pressures ............................................... 26
Exceptional regulation aimed at supporting domestic lending was withdrawn............................................. 30
Supporting monetary policy tightening as vulnerabilities grew ........................................................................ 31
D. Millions of people left behind after a year of turmoil ................................................................................... 34
Firms have been severely affected ........................................................................................................................ 34
Job vulnerabilities have widened labor market disparities ................................................................................ 36
Elevated inflation hit lower income households ................................................................................................ 40
And millions more are now in poverty ................................................................................................................ 42
II. LOOKING AHEAD ............................................................................................................................................. 46
A. The outlook for growth is tentative in a risky environment ........................................................................ 46
B. There are major uncertainties which could materially affect the path of growth this year ..................... 50
COVID-19 is a clear and present danger until the vaccination program nears completion ....................... 50
Turkey faces pronounced risk of disruptive external adjustment .................................................................... 52
C. Policy should be geared towards lowering risks in the near-term ............................................................... 54
Tighter monetary policy should be maintained until inflation is under control ............................................ 56
Fiscal policy plays an important role in securing the recovery, but increases in debt should be limited... 56
NPL resolution and corporate insolvency frameworks will be crucial this year ........................................... 60
D. As a foundation for higher medium-term inclusive growth......................................................................... 67
The COVID stimulus offers an opportunity for building a more resilient, sustainable economy ............. 67
Turkish firms can have doors opened worldwide with a new drive to reduce trade barriers ..................... 70
Revitalizing the EU-Turkey Customs Union can help realize greater gains from trade and GVCs .......... 70
A more diversified and greener financial system would support a resilient, sustained recovery ................ 72
Addressing the youth, especially female youth, employment challenge critical for recovery ...................... 75
Annexes .............................................................................................................................................................................. 79
List of figures
Figure 1: COVID cases peaked late in 2020.................................................................................................................... i
Figure 2: Very high credit growth in 2020 ....................................................................................................................... i
Figure 3: Recovery only partly due to reopening ........................................................................................................... ii
Figure 4: Consumption and investment drove growth ................................................................................................. ii
Figure 5: Inflation accelerated ........................................................................................................................................... ii
Figure 6: Exchange rate deprecated sharply .................................................................................................................. iii
Figure 7: Interest rates tightened late in 2020 ............................................................................................................... iv
Figure 8: Workers’ hours cut and others put on leave ................................................................................................. iv
Figure 9: Labor market recovery unbalanced ................................................................................................................ iv
Figure 10: Poverty estimated to have risen in 2020...................................................................................................... iv
Figure 11: Growth in 2021 driven by exports ................................................................................................................ v
Figure 12: Vaccination proceeding well .......................................................................................................................... v
Figure 13: Substantial external financing needs ............................................................................................................ vi
Figure 14: NPLs could be significantly higher than reported under forbearance ................................................... vi
Figure 15: Simulations show government budget may be hard hit by external financing shocks ....................... vii
Figure 16: COVID-19 restrictions were being re-imposed by the end of 2021 ....................................................... 1
Figure 17: Globally new COVID cases ........................................................................................................................... 2
Figure 18: and deaths continued rising into 2021 .......................................................................................................... 2
Figure 19: and fiscal balances saw a massive deterioration .......................................................................................... 2
Figure 20: Government debt rising rapidly..................................................................................................................... 2
Figure 21: Interest rates falling in EMs over 2020 ........................................................................................................ 3
Figure 22: EM portfolio flows have been robust in late 2020 and into 2021 ........................................................... 3
Figure 23: But trade volumes have rebounded .............................................................................................................. 3
Figure 24: Global PMI shows weak growth, especially for services ........................................................................... 3
Figure 25: Tourism remains in the doldrums ................................................................................................................. 4
Figure 26: Investment remains low across advanced economies and EMDEs ........................................................ 4
Figure 27: Methodological changes in reporting............................................................................................................ 5
Figure 28: Resurgence of the pandemic in Autumn...................................................................................................... 5
Figure 29: The second wave of the pandemic in autumn outpaced the first wave across countries .................... 5
Figure 30: Relatively low Covid-19 cases….................................................................................................................... 6
Figure 31:… and also low losses per million population.............................................................................................. 6
Figure 32: Sharp rebound in GDP in 2020 H2 .............................................................................................................. 7
Figure 33: Solid yoy growth in three quarters more than made up for losses in Q2 ............................................... 7
Figure 34: …driven by both industry and services ........................................................................................................ 7
Figure 35: Not only consumption, but also a large accumulation of finished goods drove growth ..................... 7
Figure 36: Turkey’s rebound was more notable compared to other countries. ........................................................ 8
Figure 37: Capacity utilization reaching almost pre-pandemic levels ......................................................................... 8
Figure 38: Large exporting sectors recovered fast ......................................................................................................... 8
Figure 39: Labor productivity per person recovered .................................................................................................... 8
Figure 40: A sharp recovery in all services sectors ........................................................................................................ 9
Figure 41: Gradual improvement in the tourism sector ............................................................................................... 9
Figure 42: Confidence picked up with eased restrictions ........................................................................................... 10
Figure 43: The pandemic and insufficient demand are top factors limiting services sector activity ................... 10
Figure 44: Credit markets have been flat so far this year… ....................................................................................... 10
Figure 45: …and domestic activity is declining............................................................................................................ 10
Figure 46: COVID restrictions were relaxed in two phases, in June and July ........................................................ 11
Figure 47: Electricity use rebounded strongly in June…............................................................................................ 11
Figure 48: …but only partially explained by reopening .............................................................................................. 11
Figure 49: Turkey’s fiscal package relies mostly on guarantees rather than direct expenditures ......................... 12
Figure 50: Huge credit stimulus, led by state banks .................................................................................................... 13
Figure 51: Turkey’s credit stimulus was among the largest amongst EMs and Advanced Economies .............. 13
Figure 52: Turkey experienced a sudden stop in 2020................................................................................................ 14
Figure 53: Turkey’s central bank released a massive amount of international reserves ........................................ 14
Figure 54: But the lira nevertheless depreciated throughout the year ...................................................................... 14
Figure 55: Portfolio flows recovered in line with other EMs .................................................................................... 14
Figure 56: Lira volatility has been lower........................................................................................................................ 14
Figure 57: Exports recovered relatively well… ............................................................................................................ 15
Figure 58: …and the trade balance is narrowing ......................................................................................................... 15
Figure 59: Export growth to non-EU markets is stronger......................................................................................... 16
Figure 60: Net services receipts also strengthening… ................................................................................................ 16
Figure 61: …but limited by weak demand in destination markets ........................................................................... 16
Figure 62: …and new demand is weakening ................................................................................................................ 16
Figure 63: Deviation from the target and inflation expectations move in parallel ................................................. 17
Figure 64: Turkey shows the largest deviation among peer countries ..................................................................... 17
Figure 65: Inflation resurged in the second half of 2020… ....................................................................................... 17
Figure 66: …with strong broad-based price pressures ............................................................................................... 17
Figure 67: Driven mostly by core goods and food prices .......................................................................................... 18
Figure 68: Rising international food prices add pressure on domestic prices ......................................................... 18
Figure 69: Inflation of some durable goods jumped backed by massive credit expansion................................... 18
Figure 70: Consumer consumption patterns have changed with the COVID-19 pandemic ............................... 18
Figure 71: Time Varying Parameters of the Model 1 .................................................................................................. 20
Figure 72: Contributions to Core Inflation (annualized, percentage points) .......................................................... 21
Figure 73: Corporate debt burden remains relatively high ......................................................................................... 22
Figure 74: Net FX position of corporates has declined ............................................................................................. 22
Figure 75: Turkish traded companies increased their leverage during the pandemic ............................................ 23
Figure 76: Contributing to solvency pressures ............................................................................................................. 23
Figure 77: Corporate vulnerability elevated .................................................................................................................. 23
Figure 78: Reflected to a rising number of illiquid firms............................................................................................ 23
Figure 79: Forbearance masks true NPLs… ................................................................................................................ 24
Figure 80: …and Stage 2 loans… ................................................................................................................................... 24
Figure 81: …but rebalancing continues ........................................................................................................................ 25
Figure 82: Banks remain prudent with provisions ....................................................................................................... 25
Figure 83: Capital Adequacy may deteriorate due to the rise in NPLs and stage 2 loans ..................................... 25
Figure 84: Improvement in risk premia encouraged banks to tap external finance ............................................... 25
Figure 85: State banks’ FX net open position recovers .............................................................................................. 26
Figure 86: Deposit dollarization increased over 2020 ................................................................................................. 26
Figure 87: Total expenditure was in line with NEP projection ................................................................................. 27
Figure 88: Revenue outturn was higher amid a surge in indirect taxes .................................................................... 27
Figure 89: Rising fiscal imbalances… ............................................................................................................................ 29
Figure 90: …driven largely by rising current transfers… ........................................................................................... 29
Figure 91: …and interest expenditures.......................................................................................................................... 29
Figure 92: But an uptick in tax on G&S contained the deterioration ...................................................................... 29
Figure 93: Deficit increasingly financed through external debt................................................................................. 29
Figure 94: Contributing an increase in central government debt .............................................................................. 29
Figure 95: Offshore swap restrictions were relaxed .................................................................................................... 30
Figure 96: Required Reserves were increased............................................................................................................... 30
Figure 97: Asset Ratio repeal eased pressure on banks to lend ................................................................................. 30
Figure 98: Real interest rates negative and lowest amongst EMs ............................................................................. 30
Figure 99: Both FX and TL deposits grew strongly .................................................................................................... 31
Figure 100: Money supply among the fastest amongst EMs ..................................................................................... 31
Figure 101: The Central Bank’s effective interest rate more than doubled between Sept-Dec 2020.................. 32
Figure 102: Monetary and financial policy reversals began in August and gained pace towards end-2020 ....... 32
Figure 103: Policy rates moved out of negative real territory .................................................................................... 33
Figure 104: As did Treasury bond yields ....................................................................................................................... 33
Figure 105: TL commercial lending rates have increased........................................................................................... 33
Figure 106: Gross international reserves were built up .............................................................................................. 33
Figure 107: Logistics hardest hit in Turkey................................................................................................................... 34
Figure 108: Turkish firms cut wages and provided leave with pay instead of firing employees .......................... 34
Figure 109: More than a third of firms received public support ............................................................................... 35
Figure 110: Informal workers… ..................................................................................................................................... 36
Figure 111: lower skilled….............................................................................................................................................. 36
Figure 112: female… ........................................................................................................................................................ 37
Figure 113 and young workers saw greatest job losses ............................................................................................... 37
Figure 114: Youth NEET rates very high for women ................................................................................................ 37
Figure 115: across all levels of education ...................................................................................................................... 37
Figure 116: Youth LFPR has fallen................................................................................................................................ 37
Figure 117: and is far below comparators for women ................................................................................................ 37
Figure 118: More youth dropping out of labor force.................................................................................................. 38
Figure 119: Women report more home duties due to COVID-19 ........................................................................... 39
Figure 120: More youth are NEET................................................................................................................................ 40
Figure 121: And far fewer women in the labor force in Turkey ............................................................................... 40
Figure 122: Consumption patterns vary greatly across income groups ................................................................... 41
Figure 123: Lower income households experienced higher inflation ....................................................................... 41
Figure 124: High inflation depressed incomes of the poor........................................................................................ 41
Figure 125: The poverty rate is estimated to reach its highest level since 2012 ..................................................... 42
Figure 126: Poverty can increase by up to 1.6 pp in 2020, corresponding to 1.6 million new poor ................... 43
Figure 127: Job losses at the peak of the COVID-19 crisis affected chiefly the bottom deciles ......................... 45
Figure 128: More than 3 million people dropped out of the middle-class since the 2018 currency shock........ 45
Figure 129: Covid-19 may increase poverty globally by up 124 million in 2020 .................................................... 44
Figure 130: Exports will be the main driver of growth in 2021 ................................................................................ 47
Figure 131: Debt overhang tends to be a drag on investment. ................................................................................. 47
Figure 132: Rebound in EMDE investment insufficient to reverse losses during 2020. ...................................... 47
Figure 133: Weak investment can cause long-lasting damage to economic potential. .......................................... 47
Figure 134: Global growth to rebound in 2021 ........................................................................................................... 48
Figure 135: But still be much lower than the no-COVID baseline .......................................................................... 48
Figure 136: COVID-19 estimated to have damaged potential growth .................................................................... 49
Figure 137: Government debt is on the rise ................................................................................................................. 49
Figure 138: Including in EMDEs ................................................................................................................................... 49
Figure 139: There remains a long way to reaching near-full vaccination rates ....................................................... 50
Figure 140: Vaccinations have proceeded quickly ....................................................................................................... 51
Figure 141: In line with some of the fastest worldwide.............................................................................................. 51
Figure 142: Market-based inflation expectations increased sharply… ..................................................................... 52
Figure 143: …and indicators of risk spiked .................................................................................................................. 52
Figure 144: External financing need remains substantial ........................................................................................... 53
Figure 145: And reserve adequacy metrics suggest a substantial need in the event of a shock ........................... 53
Figure 146: International reserves have fallen well below the level of short-term possible drains ..................... 54
Figure 147: Banks maintain liquidity to meet ST FX liabilities ................................................................................. 54
Figure 148: Support to health, households, and businesses was provided .............................................................. 57
Figure 149: Primary expenditure rise to be driven mainly by current transfers...................................................... 57
Figure 150: Turkey has relatively low public debt…................................................................................................... 58
Figure 151: …with a relatively higher interest burden ................................................................................................ 58
Figure 152: Debt breakdown differs among EMs ....................................................................................................... 59
Figure 153: Conditions for external funding diverge .................................................................................................. 59
Figure 154: Volatile trend in bond yields ...................................................................................................................... 59
Figure 155: High correlation with CPI expectations ................................................................................................... 59
Figure 156: Increased debt burden with a real interest rate shock............................................................................ 60
Figure 157: With rising interest burden ......................................................................................................................... 60
Figure 158: Impulse-Responses to negative demand and currency volatility shocks ............................................ 65
Figure 159: Counterfactual NPL ratio: Private banks ................................................................................................. 66
Figure 160: Counterfactual CAR: Private Banks ......................................................................................................... 66
Figure 161: Counterfactual NPL ratio: State banks..................................................................................................... 66
Figure 162: Counterfactual CAR: State Banks ............................................................................................................. 66
Figure 163: Counterfactual NPL ratio: Foreign banks ............................................................................................... 66
Figure 164: Counterfactual CAR: Foreign Banks ........................................................................................................ 66
Figure 165: Range of environmental policy instruments ............................................................................................ 68
Figure 166: Turkey and the EU’s preferential trade agreements in force ................................................................ 71
Figure 167: Coverage rate of ALMPs has not been growing… ................................................................................ 77
Figure 168: ...although unemployment benefits have been scaling up ..................................................................... 77
Figure 169: On-the-job training more used by men, but more women access vocational training..................... 78
Figure 170: And unemployment benefits are primarily claimed by men ................................................................. 78
List of tables
Table 1: Phillips Curve Estimation Results................................................................................................................... 19
Table 2: COVID-19 measures introduced by the Ministry of Family, Labor and Social Services ...................... 27
Table 3: Fiscal operations (CG, percent of GDP)....................................................................................................... 28
Table 4: Population targeting for Turkey’s vaccination program.............................................................................. 51
Table 5: Options to respond to the COVID-19 challenges impacting solvency of firms and individuals? ....... 62
Table 6: Sign restrictions enabling the identification of demand and currency shocks on NPLs and CARs .... 64
List of Boxes
Box 1: Drivers of Core inflation-Phillips Curve Estimation...................................................................................... 19
Box 2: Constraints to Women’s Economic Empowerment and Covid-19 ............................................................. 39
Box 3: The Effects of Covid-19 on Global Poverty ................................................................................................... 44
Box 4: Global Outlook .................................................................................................................................................... 48
Box 5: Turkey’s vaccination program to date............................................................................................................... 50
Box 6: Turkey’s new Economic Reform Package ....................................................................................................... 54
Box 7: Preventing an expected surge in NPLs from becoming a drag on economic growth .............................. 61
Box 8: Conditional Counterfactual Analysis of NPLs and CARs in the absence of government actions ......... 63
Box 9: Green COVID-19 recovery measures for various sectors: Country examples .......................................... 69
Box 10: Greening the Financial Sector in Turkey........................................................................................................ 73
List of Abbreviations
ADV Advanced Economies
AFR Africa
ALMP Active labor market programs
ALB Albania
AR Asset Ratio
BAT Banks Association of Turkey
BGR Bulgaria
BPS Business Pulse Survey
BRA Brazil
BVAR Bayesian Vector Auto Regression
BRSA Banking Regulation and Supervision Agency
CAB Current Account Balance
CAR Capital Adequacy Ratio
CB Central Bank
CBAM Carbon Border Adjustment Mechanism
CBRT The Central Bank of the Republic of Turkey
CG Central Government
CGF Credit Guarantee Fund
CDS Credit default swap
CHN China
CMB Capital Markets Board
COVID-19 2019 Novel Coronavirus Disease
CPI Consumer Price Index
CVI Corporate Vulnerability Index
ECA Europe and Central Asia
EFI Equitable Growth, Finance and Institutions
EMDE Emerging Markets and Developing Economies
EMBI Emerging Markets Bond Index
EM Emerging Markets
ENB Environment, Natural Resources and Blue Economy
ERP Economic Reform Package
ESG Environment, Social and Governance
EU European Union
FCI Finance, Competitiveness and Innovation
FinSAC Financial Sector Advisory Center
FX Foreign exchange
G-20 Group of 20
GDP Gross Domestic Product
GEFR Gross External Financing Requirement
GEO Georgia
GIR Gross International Reserves
GP Global Practice
GW Gigawatt
HCI Human Capital Index
HD Human Development
HICES Household Income Consumption Expenditure Survey
HRV Croatia
HUN Hungary
ICT Information and Communication Technologies
ICR Interest Coverage Ratio
ICLS International Conference of Labour Statisticians
IDN Indonesia
IFC International Finance Corporation
IFRS-9 International Financial Reporting Standard 9
ILO International Labor Organization
IMF International Monetary Fund
IND India
ISKUR Turkish Employment Agency
ITA Italy
KGZ Kirgizstan
LAC Latin America and Caribbean
LFPR Labor Force Participation Rate
LHS Left Hand Side
MAR Morocco
MAX Maximum
MENA Middle-East and North Africa
MFN Most favored nation
MIN Minimum
MOTF Ministry of Treasury and Finance
MOM Month-on-month
MSCI Morgan Stanley Capital International
MSME Micro, Small and Medium Enterprises
MTI Macroeconomics, Trade and Investment
NEET Not in Education, Employment, or Training
NEP New Economic Program
NGA Nigeria
NSA Not seasonally adjusted
NPL Non-performing Loan
OECD Organization for Economic Co-operation and Development
OJT On-the-Job Training Program
OxCGRT Oxford COVID-19 Government Response Tracker
PAK Pakistan
PISA Program for International Student Assessment
PMI Purchasing Managers’ Index
POL Poland
POV Poverty
PPP Purchasing power parity
PPI Producer Price Index
PTA Preferential Trade Agreement
QOQ Quarter-on-quarter
R&D Research and development
RHS Right hand side
ROM Romania
ROU Romania
RUS Russia
RWA Risk-weighted assets
SA Seasonally adjusted
SD Sustainable Development
SDIF Savings-Deposit Insurance Fund
SGK Social Security Institution
SME Small and medium size entrepreneurs
SPJ Social Protection and Jobs
SSI Social Sustainability and Inclusion
TEM Turkey Economic Monitor
TFP Total factor productivity
TL Turkish Lira
TJK Tajikistan
TRY Turkish Lira
TUR Turkey
TURKSTAT Turkish Statistical Institute
TWF Turkey Wealth Fund
UK United Kingdom
UKR Ukraine
US United States of America
USA United States of America
US$ United States of America Dollar
UZB Uzbekistan
VT Vocational and Technical Courses Program
WB World Bank
WBG World Bank Group
YOY Year-on-year
ZAR South Africa
The Turkey Economic Monitor (TEM) periodically analyzes economic developments, policies and
prospects in Turkey. The TEM was prepared under the guidance of Auguste Tano Kouame (WB Country
Director, Turkey), Lalita Moorty (Regional Director for Equitable Growth, Finance and Institutions,
ECA) and Sandeep Mahajan (Practice Manager, Macroeconomics, Trade and Investment GP) by a core
team including David Knight (Acting Program Leader, EFI Turkey), Pinar Yasar (Senior Country
Economist, MTI GP), Erdem Atas (Research Analyst, MTI GP), Javier Baez (Senior Economist, POV
GP), Metin Nebiler (Economist, POV GP), Etkin Ozen (Senior Financial Sector Specialist, FCI GP), and
Gunhild Berg (Senior Financial Sector Specialist, FCI GP).
The team is very grateful to the following colleagues for advice and technical inputs: Asli Demirguc-Kunt
(Chief Economist, ECA region), Michael Lokshin (Lead Economist, ECA Chief Economist’s Office),
Heba Elgazzar (Program Leader, HD Turkey), Laurent Debroux (Sector Leader, SD Turkey), Daniel
Besley (Senior Climate Change Specialist, Climate Change Advisory and Operations), Craig Meisner
(Senior Environmental Economist, ENB GP), Sasa Eichberger (Senior Environmental Specialist, ENB
GP), Thanh Phuong Ha (Environmental Analyst, ENB GP), Umut Kilinc (Economist, FCI GP), Michael
Ferrantino (Lead Economist, MTI GP), Mattia Makovec (Senior Economist, SPJ GP), Laurent Bossavie
(Economist, SPJ GP), Sirma Seker (Senior Economist, SPJ GP, Joel Reyes (Senior Education Specialist,
Education GP), Safir Sumer (World Bank Consultant), Nadwa Rafeh (Senior Health Specialist, Health
GP), Efsan Nas Ozen (World Bank Consultant), Arzu Uraz (Social Development Specialist, SSI GP),
Andrea Woodhouse (Senior Social Development Specialist, SSI GP), John Grinyer (World Bank
Consultant), Ibrahim Unalmis (World Bank Consultant), and Kaan Inan (World Bank Consultant).
The team is very grateful to Tom Farole (Lead Economist, SD PG), Kim Edwards (Senior Economist,
MTI GP) and Fernando Blanco (Lead Economist, IFC) for peer review comments and advice.
The team is very grateful to colleagues from the Central Bank of the Republic of Turkey, the Ministry of
Treasury and Finance, the Presidency of Strategy and Budget, and the Ministry of Trade for very helpful
discussions on economic developments and policy priorities. The team greatly appreciates insights
provided by business associations and the private sector during the preparation of the TEM.
The TEM is a product of the staff of the World Bank Group. The findings, interpretations, and
conclusions expressed in this report do not necessarily reflect the views of the Executive Directors of the
World Bank (or the governments they represent), or the Government of the Republic of Turkey.
EXECUTIVE SUMMARY
Turkey responded to COVID-19 with a large
I. TAKING STOCK economic stimulus program, focused on credit
COVID-19 has taken a heavy toll on Turkey, as channels. In fiscal terms, Turkey’s COVID-19
it has across much of the world. New estimates stimulus package amounted to nearly 12 percent of
of total COVID-19 cases indicate that the epidemic GDP when including tax deferrals and contingent
grew rapidly over November and into December liabilities. This is larger than the average for
(Figure 1), before a new round of restrictions were emerging market, G20 countries, and is similar in
put in place to control the spread of the virus in late size to the stimulus packages of the United States,
December. The expansion of COVID-19 in late Australia and Canada. While the program utilized a
2020 followed a similar path to many European broad range of fiscal tools, uniquely amongst G20
countries. countries, Turkey’s support was overwhelmingly
provided through the banking sector, and was not
Figure 1: COVID cases peaked late in 2020 realized as direct fiscal costs on the budget, but as
70,000
contingent liabilities to the government in future.
Covid-19 infections in Turkey (daily)
60,000 Daily new cases, old Credit stimulus, loose monetary policy, and
50,000
method other regulatory measures to promote credit
Daily change, expansion drove a sharp increase in economic
synthetic
40,000
Daily change, new activity in late 2020. In addition to government
30,000 method (actual data) measures to recapitalize a partial credit guarantee
fund and state banks, policy interest rates were
20,000 maintained below inflation and a series of financial
10,000 regulatory measures introduced that incentivized
banks to increase lending activities. The combined
0
effect of these policies was to generate one of the
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Jul-20
Oct-20
Feb-21
Jan-21
largest credit expansions that the world saw in 2020
(Figure 2).
Turkey’s economic growth performance in Figure 2: Very high credit growth in 2020
2020 was strong compared to other countries,
Real private credit growth in 2020
but poverty spread, and unemployment (cumulative % change)
became more prevalent. Turkey’s GDP grew by 30%
1.8 percent in 2020, which was the fastest amongst 25%
G20 countries in 2020 aside from China. This was 20%
on the back of one of the strongest economic 15%
rebounds in the second half of 2020, largely 10%
achieved in the third quarter, but with positive 5%
growth continuing into the fourth quarter (Figure 0%
4). Despite measures to support jobs and -5%
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Dec-19
Mar-20
Jun-20
Dec-20
Feb-20
Jul-20
Jan-20
Oct-20
households, labor market and poverty outcomes
nevertheless deteriorated. Indicators of domestic
demand showed some signs of economic growth Turkey Max EMs Max Adv
cooling early in 2021, although supply side
Note: Credit growth is not adjusted for FX valuation effects for
measures such as industrial output remained Turkey or comparators.
robust.
i
The direct fiscal response to the COVID-19 Figure 4: Consumption and investment drove
crisis was smaller in size and targeted, but growth
provided important support to firms and GDP (growth contributions, yoy)
10%
households. On-budget fiscal expenditure was
much smaller than the run-up of contingent 5%
liabilities, but supported important, targeted fiscal
objectives. The major items supported with fiscal 0%
expenditure were payments to furloughed workers
and those put on unpaid leave, additional lump- -5%
sum social support to households, and increased -10%
unemployment benefits. On the revenue side,
2019
2020
several tax and government fee payments were
deferred to reduce the cost burden on firms while
Priv. Consumption Pub. Consumption Investment
economic activity was reduced.
Exports Imports Stocks
Figure 3: Recovery only partly due to reopening
Growth in electricity consumption relative Economic risks emerged alongside recovery.
to counterfactual baseline, 2020 Policy frameworks geared towards ensuring a
20
strong recovery in 2020 also contributed to
growing imbalances and risks, especially domestic
10 inflation, currency depreciation and loss of external
reserves. Inflation gradually accelerated, to reach an
0 18-month high of 16.2 percent by March 2021
(Figure 5). Facing capital outflows, falling exports,
and rising imports the Central Bank financed most
-10
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec of Turkey’s rising current account deficit using
Reopening impact
international reserves in 2020Q2. As reserves, net
Residual of short-term liabilities, reached unprecedented
Seasonally adjusted change in electricity lows, the Central Bank limited further losses of
reserves and the Lira began depreciating, ending
the year almost 30 percent lower, further adding to
Reopening after the first wave also played a inflationary pressures (Figure 6).
role in economic recovery. Turkey had swiftly
introduced COVID-19 related restrictions that, Figure 5: Inflation accelerated
although effective in limiting COVID transmission, Inflation and Exchange Rate (yoy, %
were somewhat less disruptive than a full-scale 25 change) 60
lockdown. Following the first wave, mobility and
20
business operation restrictions were relaxed in June 40
and July. Based on the experiences of reopening 15
around the world and using high-frequency 20
10
electricity consumption data as a proxy for
0
economic activity, reopening is estimated to have 5
accounted for between one third and one half of
0 -20
the rebound in economic activity, indicating that
other factors – namely stimulus – played a pivotal
role (Figure 3). CPI Inflation Core Inflation
Exc. Rate (TL/$), RHS
ii
Monetary and credit policies were significantly signaled a softening monetary policy stance toward
strengthened toward the end of 2020 to address price stability. The Turkish Lira experienced a 9.3
the growing vulnerabilities. As vulnerabilities percent depreciation in one day, bond yields
grew, the authorities implemented a raft of increased by close to 500 basis points in just two
significant policy changes, geared towards days, and the 5-year CDS spread experienced its
stabilizing the economy and reducing the sharpest one-day rise on record. Since then,
imbalances that had built up. The Central Bank statements committing to tight monetary policy
raised its policy interest rate in a series of steps meant volatility has eased for the time being.
between September 2020 and March 2021, bringing
the real interest rate back into positive territory The credit boom, along with a sharp fall in
(Figure 7). The Central Bank also took steps to tourism receipts, drove the current account
improve transparency; reverting to a single policy deeply into deficit in 2020. The economic
rate, providing more forward guidance in Monetary rebound in the third quarter was driven by
Policy Committee statements, and publishing more domestic consumption and investment, much
high-frequency data of market interest. Both the spent on imported goods. A fast-rising import bill
Central Bank and the Banking Regulation and was driven by this demand as well as rising demand
Supervision Agency (BRSA) withdrew regulation for gold imports as a safe haven investment. This
that had been introduced to support credit growth combined with sharply lower exports of goods and
such as the “Asset Ratio” that had mandated services – especially tourism – to drive the current
commercial banks to actively use their assets, and account from a surplus in 2019 to a deficit of 5.1
reserve requirements linked to real credit growth. percent of GDP in 2020. After major disruption in
the first half, exports began recovering later in the
Figure 6: Exchange rate deprecated sharply year with the resumption of supply chains.
Cumulative change of NEER in 2020 Corporates and banks felt balance sheet
10% pressure as corporate debt increased. The
corporate deleveraging process, ongoing since
0% 2018, was interrupted by the COVID-19 pandemic.
-10%
Debt levels in the non-financial corporate sector
increased by 9 percent of GDP in the first six
-20% months of 2020, the sharpest rise amongst
emerging markets. The number of illiquid
-30%
Q4-19 Q1-20 Q2-20 Q3-20 Q4-20
companies spiked in the second quarter of 2020
and measures of corporate vulnerability remained
TUR CHN RUS heightened throughout the year. Corporate sector
BRA ZAR vulnerability continues to be the main risk to banks
susceptible to currency and asset quality shocks.
Entering 2021, volatility was reduced, and risks
were beginning to be addressed. Off the back of
these revised policy frameworks, portfolio flows
began returning to Turkey, the Lira stabilized and
appreciated markedly compared to its low point in
October 2020 and country risk premia, as measured
by credit default swap (CDS) spreads, began falling.
However, in March 2021, Turkey saw market
turbulence and price volatility, reversing earlier
gains. The exchange rate and Lira asset markets
were buffeted by intense volatility following the
replacement of the Central Bank Governor in
March—the second change at the helm of the
Central Bank in four months—amid fears that it
iii
Figure 7: Interest rates tightened late in 2020 informal workers lagged behind (Figure 9). This
asymmetric recovery exacerbated existing
13 Real policy interest rate (percentage)
disparities as youth already had much higher
11
9 unemployment rates and female labor force
7 participation is far lower than expected given
5 Turkey’s levels of income and human capital.
3
1 Figure 9: Labor market recovery unbalanced
-1
-3 Percentage change in employment (year on
-5 year to November 2020)
Jan-19
Mar-19
May-19
Jul-19
Sep-19
Nov-19
Jan-20
Mar-20
May-20
Jul-20
Sep-20
Nov-20
Jan-21
Mar-21
5% Formal Skilled
Max Min Turkey 0%
-5% Male Adult
Note: Turkey compared to range of 12 other EMs; real ‘ex
post’ interest rate. Female
-10% Youth
Unskilled
Firms in Turkey were badly affected by -15%
COVID-19 in 2020. Firms in Turkey experienced Informal
a sharper drop in sales than comparators. -20%
Gender Age Formality Skill level
Constrained by a government-imposed redundancy
moratorium since the onset of COVID-19, Turkish
firms responded by reducing working hours and Poverty rose significantly in 2020. In
putting more employees on paid leave (Figure 8). conjunction with job losses affecting the poorer
Female-headed firms were disproportionately more, the proportion of people in poverty is
affected, being concentrated in the hard-hit estimated to have risen from 10.2 percent in 2019
accommodation sector. to 12.2 percent in 2020 – following a significant
uptick in 2019 – and presents a growing challenge.
Figure 8: Workers’ hours cut and others put on (Figure 10). Were it not for the fiscal measures to
leave
support workers and households, the increase in
Change in employment conditions (percent
poverty would have been far worse.
of employees)
35 Figure 10: Poverty estimated to have risen in 2020
30
25 Poverty headcount rate (percentage of
20
people)
15.2%
15
10
13.1%
16%
12.3%
12.2%
5
14%
10.7%
10.7%
10.2%
0
10.0%
9.5%
12%
Wage cut
Hours cut
Leave w/o pay
Leave w/ pay
8.7%
8.5%
10%
8%
6%
Turkey ECA-EU ECA-nonEU Others 4%
2%
Note: Initial impact up to time of survey (June-July 2020)
0%
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Job losses compounded existing labor market
disparities. Despite the redundancy moratorium,
millions of jobs were reported lost in the second
quarter of 2020. While overall the labor market saw
a good recovery towards the end of the year, the
recovery among female, youth, lower-skilled and
iv
and the ongoing risk that uncontrolled outbreaks of
II. LOOKING AHEAD COVID-19 can continue to emerge rapidly, the
Turkey is expected to see a rebound in real possibility that more restrictive controls would
economy activity in 2021, but in an need to be reintroduced throughout the year could
environment of macroeconomic uncertainty lead to lower than forecast growth.
and elevated vulnerabilities. Turkey and much of
Figure 12: Vaccination proceeding well
the rest of the world will struggle to shake off the
COVID-19 pandemic this year. Annual growth in Total vaccinations per 100 people since first
Turkey is nevertheless expected to be a respectable 50 vaccination
5 percent since the ‘base’ for comparison in early 45
40
2020 was exceptionally low (Figure 11). However, 35
market pressures in March have heightened macro- 30
25
financial risk and dampened Turkey’s economic 20
outlook, particularly by weakening the policy grip 15
10
over inflation and exacerbating pressures over 5
corporate and bank balance sheets. 0
10
20
30
40
50
60
70
80
90
100
-
Growth in 2021 is expected to be driven by UK United States Chile
export recovery and base effects. Turkey’s Morocco Greece Poland
economy had already recovered from its COVID- Romania Portugal Turkey
induced low by the end of 2020, and this level of
activity is expected to be broadly sustained. Exports Heightened external volatility could derail
are expected to account for two-thirds of growth in growth. Rising inflation expectations in developed
2021, normalizing from the severe shock of last countries (the US in particular) and an
year. Private consumption is expected to only intensification of market speculation over an end to
contribute slightly to growth. very loose monetary policy in advanced economies
could cause destabilizing movements in global
Figure 11: Growth in 2021 driven by exports
liquidity away from emerging markets, akin to the
10% Contribution to Growth taper tantrum episode of 2013. As in 2013, this could
8% pressure central banks of emerging-market
6% economies to raise interest rates to prevent
4% significant weakening of their currencies and a
2%
0% build-up of inflationary pressures.
-2%
-4% Sustained macro-financial pressures since 2018
-6% have depleted policy buffers and left the
-8% Turkish economy vulnerable. Although external
2019
2020
2021
2022
2023
financing requirements remain sizeable (Figure 13),
Stocks Imports macroeconomic buffers such as international
Exports Investment reserves are now lower and more reliant on the
Government Spending Private Consumption
Growth financial sector remaining healthy. Gross
International Reserves (GIR) at the Central Bank,
The resurgence of COVID-19 and risks to the for example, fall short of the Central Bank FX
progress of vaccination roll-out could apply the liabilities to the commercial banking sector. Turkey
brake to growth. While Turkey’s vaccination may be less able to weather any severe bout of
program, begun in January, has made good external volatility with fewer resources available to
progress (Figure 12), risks to the security of supply smooth excess external volatility. Such events
of vaccination and capacity to scale-up at speed would prove damaging to corporates, households
mean there is a wide range of estimates as to when and the financial sector, and inevitably lower
the majority of the population could be fully growth prospects. Given these risks, the Central
vaccinated. Coupled with resurgent COVID-19 Bank would be advised to substantially raise the
v
level of GIR and to bring the level of reserves net of the banking sector’s distressed assets not
of short-term drains back into positive territory. observable at present. But simulations to estimate
non-performing loans (NPLs) and capital adequacy
Figure 13: Substantial external financing needs ratios suggest banks do have a substantial backlog
of distressed assets to deal with. While unlikely to
Estimated gross external financing req. present major systemic risks, especially given the
12,000
(US$m) banking sector’s prudent buffers, the NPL and
10,000 Stage II loan ratios are estimated to significantly
8,000 increase compared to their reported levels, were
6,000 forbearance, loan deferrals and restructurings not
4,000 in place (Figure 14). In addition to withdrawing
2,000 forbearance measures, the authorities can support
0 the orderly and efficient resolution of distressed
assets by strengthening insolvency and debt
Apr-21
May-21
Aug-21
Sep-21
Nov-21
Mar-21
Jun-21
Dec-21
Jul-21
Jan-21
Feb-21
Oct-21
resolution frameworks. Recently, the authorities
Non-interest CAB Fin. corp. made commitments to support NPL resolution, yet
Non-fin. corp. Government a holistic approach remains needed.
Public banks
Figure 14: NPLs could be significantly higher than
A firm commitment by the Central Bank to reported under forbearance
tame inflation will be necessary to restore
macroeconomic stability. While Turkey has Banking Sector NPL Forecasts
15%
committed to a robust monetary policy stance to
achieve a permanent reduction in inflation, the
10%
frequent changing of Central Bank governors in
recent years has elevated uncertainty over the
future course of policy, which has fed into 5%
deteriorating expectations of inflation and market
volatility. The Central Bank would be well served 0%
by staying unambiguously committed to inflation-
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
2020.Q4
targeting and price stability as its primary mandate.
Any premature softening of monetary policy would
Dom. Private State
be contrary to the current policy position of the Foreign Dom. Priv. Proj
Central Bank and is likely to elevate inflationary State proj. Foreign proj.
expectations and lead to external pressures and
weaker financial and corporate balance sheets. The Although fiscal space is limited, there is room
Economic Reform Package (ERP) launched in for important, targeted support to firms and
March 2021 provides a useful anchor to support households to help ensure an inclusive
economic recovery and stability, but needs a sound recovery. Turkey maintained a modest fiscal deficit
macroeconomic policy framework to achieve over a difficult year in 2020, but a significant
desired results. Steps to consolidate Central Bank amount of public support in the form of guarantees
independence as part of the authorities’ ongoing and state bank capitalization may be realized on the
reform program would also support the objective budget in future. Coupled with Turkey’s continuing
of the ERP. high exposure to external market risk, this means
that authorities should remain prudent to avoid
A focus on supporting the banking sector to large run-ups of public debt. While interest
restore its balance sheets will be key to expenses are not presently high, Turkey would be
reducing risks and paving the way for future challenged by an adverse financing cost shock
growth. Rapid credit growth over the past year has (Figure 15). Nevertheless, the costs of continuing
added to problems of distressed assets in the key, targeted post-COVID-19 on-budget fiscal
financial sector. Forbearance measures, still in place support measures will be limited, and may be
since the onset of COVID-19, make the true state
vi
important to secure a recovery that is likely to A more diversified and greener financial
remain tentative throughout this year. system would support a resilient, sustained
recovery. Developing domestic sources of long-
Figure 15: Simulations show government budget term finance and a green finance taxonomy would
may be hard hit by external financing shocks alleviate existing imbalances in the financial system
and raise economic growth. However, a volatile
Interest expenditure to GDP and a potential
10 shock (in percent)
economic environment and its symptoms have
acted as a constraint on capital market and long-
8 term finance development.
6 Finally, seismic shocks to labor markets are
likely to have long-lasting effects, and social
4
assistance, as well as active and targeted labor
2 market policies, will be increasingly important to
ensure people – especially women and the youth of
0 today – do not miss the opportunity to fulfil their
2018 2019 2020 2021 2022 2023 potential as productive members of society.
Mexico Turkey Indonesia
Romania USA Germany
Measures that increase international market
access and economic integration will support
Turkish firms to grow and become more
productive. While in most cases contingent on
bilateral or multilateral agreements, the ability for
Turkish firms to trade in goods and services and to
invest internationally is an important route to
increase demand and realize gains in productivity
through the adoption of new technology and
practices. As a Customs Union partner with the
EU, it is important that Turkey moves together
with the EU to secure common access to tariff-free
trade with third-party countries.
Measures to incentivize a green recovery – and
begin a green transformation – can keep
Turkey at a competitive advantage as global
markets decarbonize. While considerable
uncertainty persists, developments in Turkey’s key
economic partner, the European Union, indicate
that decarbonization will be proceeding with
increasing speed. Turkey can use the opportunity
of adapting existing policy frameworks, including
those related to COVID-19, to support firms in
planning for an adjustment to low-carbon markets
over time. Important measures would include both
the planned strengthening of the institutional
framework for greening the financial sector and a
national emission trading scheme, as well as tax and
spending programs that are linked to carbon
emissions.
vii
I. TAKING STOCK
Policy responses to the pandemic and its economic impact led to a sharp rebound in GDP in the
second half of 2020. Taking account of quasi-fiscal measures to support credit, Turkey’s
stimulus program was larger than the average of G20 emerging market countries. A stimulus-
induced credit boom and reopening in June and July drove a sharp rebound in economic activity,
leaving Turkey as one of the few countries with positive growth in 2020. But this growth came
with rising inflation, falling international reserves, a weaker Lira, a sharply expanded current
account deficit and increasing corporate stress. In response, interest rates were tightened, and a
series of measures aimed at promoting credit growth were removed. Entering 2021, volatility
was reduced, and risks were beginning to be addressed. The Lira was more stable, capital inflows
supported a recovery in international reserves and the banking sector was able to maintain its
FX liquidity buffers. However, following the dismissal of the Central Bank Governor, Turkey
saw pronounced market turbulence and price volatility in March, reversing earlier gains. While
overall the labor market saw a good recovery towards the end of the year, the recovery among
female, youth, lower-skilled and informal workers lagged behind. This, in conjunction with high
inflation is likely to have hurt the poor more. Poverty is estimated to have risen substantially
over 2020, and presents a growing challenge.
A. COVID-19 determined the course of the year worldwide and in Turkey over 2020
Countries worldwide struggled to control COVID-19, while supporting their economies
After a catastrophic economic impact in the first part of the year, the global economy has begun to
recover, but the ongoing effects of COVID-19 are Figure 16: COVID-19 restrictions were being re-
still very much apparent. COVID-19 in 2020 caused imposed by the end of 2021
the worst global recession since World War II and a
severe loss of life, while tipping millions more people
Stringency of pandemic control index
into extreme poverty. This episode is expected to inflict
lasting scars that push activity and income well below 90
80
their pre-pandemic trend for a prolonged period. 70
Global economic activity began growing again in the 60
50
second half of the year, but remains some way from 40
recovering to levels seen prior to the onset of the 30
COVID-19 pandemic. 20
10
0
Overall, the global economy is estimated to have
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Mar-21
Jul-20
Jan-20
Feb-20
Oct-20
Jan-21
Feb-21
contracted by 4.3 percent in 2020 1. Advanced
economies are expected to have contracted by 5.4
Emerging Markets Advanced Economies
percent. While China followed a different, and more
robust, path compared to many other countries, the Source: WB Global Economic Prospects, Jan 2021; University of
group of emerging market and development Oxford.
Notes: Figure shows the simple average stringency index for EMDEs.
economies excluding China are expected to have The shaded area indicates the regional range. The stringency index
contracted by 5.0 percent in 2020. range is between 0 and 100, with 100 being the most stringent.
1 This, and all projections presented in this section draw on the World Bank Group’s Global Economic Prospects report, published in
January 2021.
1
Since they relaxed COVID-19 restrictions in mid-2020, many countries have experienced new waves
of COVID and needed to reintroduce restrictions. An incipient recovery was initially supported by a partial
easing of stringent lockdowns. Resurgent outbreaks, however, have forced many governments to maintain or
reintroduce lockdown measures (Figure 16). Some countries have experienced a sharp resurgence of infections
and deaths, particularly amongst the high-income countries grouping (Figure 17, Figure 18). Nonetheless,
pandemic-control measures in general have become better targeted and less economically disruptive.
Figure 17: Globally new COVID cases Figure 18: and deaths continued rising into 2021
Monthly new COVID cases (index, 2020 Monthly COVID deaths (index, 2020 avg=100)
avg=100)
350 300
300 250
250
200
200
150
150
100 100
50 50
0 0
06-Jul-20
06-Oct-20
06-Jan-21
06-Feb-21
06-Apr-20
06-May-20
06-Aug-20
06-Sep-20
06-Nov-20
06-Mar-20
06-Jun-20
06-Dec-20
06-Jul-20
06-Oct-20
06-Jan-21
06-Feb-21
06-Apr-20
06-May-20
06-Aug-20
06-Sep-20
06-Nov-20
06-Mar-20
06-Jun-20
06-Dec-20
World UMICs HI World UMICs HI
Source: World Health Organization via Haver.
Strong policy responses and automatic stabilizers have led to gaping fiscal deficits in 2020. Drastic falls
in consumption and trade taxes, measures to delay or lower tax payments and the COVID-19 related health,
economic and social spending have led to some of the largest global fiscal deficits in recent history (Figure 19).
Fiscal expansion has led to sharp increases in debt levels around the world. While deficits were trimmed by Q3,
many countries’ public finances remain deeply in deficit, with public sector debt rising commensurately (Figure
26).
Figure 19: and fiscal balances saw a massive Figure 20: Government debt rising rapidly
deterioration
Quarterly fiscal balance (%), simple average, Quarterly change in government debt
seasonally-adjusted (percentage of GDP)
2 10
0
8
-2
-4 6
-6
4
-8
-10 2
-12
0
-14
-16 -2
Q1-19
Q2-19
Q3-19
Q4-19
Q1-20
Q2-20
Q3-20
Q4-20
Q1-2019
Q2-2019
Q3-2019
Q4-2019
Q1-2020
Q2-2020
Q3-2020
Q4-2020
Advanced economies Emerging economies Advanced Emerging
Source: Haver Analytics.
Notes: Based on a sample of 19 emerging and 12 advanced economies with data available for the whole period.
2
After collapsing in the aftermath of the pandemic outbreak, portfolio flows to EMDEs strengthened
later in 2020, led by inflows to China and other parts of Asia. The aggressive policy actions of central banks
prevented the global financial system from falling into crisis last year. Financial conditions are generally loose
(Figure 20), as suggested by low borrowing costs, abundant credit issuance, and a recovery in equity market
valuations. Portfolio flows to Asia have gained in strength, but have been positive in other emerging markets
as well (Figure 21).
Figure 21: Interest rates falling in EMs over 2020 Figure 22: EM portfolio flows have been robust in late
2020 and into 2021
Policy interest rates (%) Portfolio flows to emerging markets (US$bn)
8
7 60
6
5 20
4
3 -20
2
1 -60
0
-100
May-19
May-20
Sep-19
Nov-19
Sep-20
Nov-20
Mar-19
Mar-20
Jul-19
Jul-20
Jan-20
Jan-21
Jan-19
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Feb-20
Jul-20
Oct-20
Jan-21
US UK Eurozone EMs avg Europe Asia LAC AFR & MENA
Source: Haver Analytics. Source: Haver Analytics.
While globally, manufacturing and merchandize trade have recovered, services and investment remain
sharply lower. Whereas activity and trade in goods has improved (Figure 23), the services sector remains
anemic (Figure 24), with international tourism, in particular, still depressed (Figure 25). The fall in global
investment has been pronounced, particularly for EMDEs excluding China (Figure 26). Most commodity prices
rebounded from their mid-2020 lows as strict lockdowns were gradually lifted and demand firmed, especially
from China; however, the recovery in oil prices was more modest amid concerns over the pandemic’s lasting
impact on oil demand.
Figure 23: But trade volumes have rebounded Figure 24: Global PMI shows weak growth, especially
for services
World trade volumes Global Purchasing Manager's Index, output
(SA, >50 is expansion)
105
100
95
90
50 threshold
85
80
World trade Adv. Econ. Imports Services Manufacturing
Source: Haver Analytics. Source: IHS Markit via Haver Analytics.
3
Figure 25: Tourism remains in the doldrums Figure 26: Investment remains low across advanced
economies and EMDEs
Monthly tourist arrivals (millions)
Year-on-year fixed investment growth, simple
average across countries
120
15%
100 10%
5%
80
0%
60 -5%
-10%
40
-15%
20 -20%
Q1-2017
Q2-2017
Q3-2017
Q4-2017
Q1-2018
Q2-2018
Q3-2018
Q4-2018
Q1-2019
Q2-2019
Q3-2019
Q4-2019
Q1-2020
Q2-2020
Q3-2020
0
May
Sep
Mar
Aug
Oct
Jun
Jul
Dec
Jan
Feb
Apr
2019 2020 Nov World EMDEs Advanced
Source: Haver Analytics. Source: Haver Analytics.
Note: Sum over countries with data available for whole period
Resurgent COVID-19 had called for re-imposition of control measures by late 2020
Data indicate a large increase in confirmed COVID-19 cases by the end of the year. As a nationwide
testing program was rolled out, the authorities ceased reporting asymptomatic COVID-19 cases on July 28,
2020 before reverting to a case definition that includes asymptomatic cases on December 10, 2020. When the
methodology was adjusted on December 10, as well as reporting daily cases from that date according to the
newly revised definition, the authorities also reported the cumulative total number of cases detected to date
according to that definition. The structural break is clear in Figure 28, as the total number of Covid-19 cases
jumped by 1.2 million in a day. Extrapolating daily data on total confirmed cases suggests the COVID-19 spread
may have peaked 2at around 60,000 cases around mid-November, before falling again to the point where the
new data series began with around 30,000 daily cases at end-November (Figure 28).
2 By a process of fitting data based on the consistent and related ‘old method’ series and the partial data for the ‘new series’, we are
able to produce a synthetic series of what daily cases may actually have looked like over the whole period based on the new definition.
In creating a synthetic series two criteria must be satisfied and a further two criteria are desirable:
Criteria 1: The number of daily infections must be equal to the historical series up to July 29 (essential)
Criteria 2: The total number of cumulative total infections must equal 1,296,539 on November 26, when the new series starts
(essential)
Criteria 3: The ‘new’ series between July 29 and November 26 be a function of the ‘old’ series.
Criteria 4: There are no structural breaks (jumps) in the total number of cases and daily cases series.
Method
The daily percentage change in total cases is calculated for the period July 29 and November 26 using the following equation:
������������������������������������_������������������������������������������������������������������������������������
������������������������������������_������������������������������������������������������������������������������������ = ������������ � � + ������������������������������������������������������������������������������������������������
������������������������������������ _������������������������������������������������������������������������������������−1
Where g is determined so to satisfy:
������������������������������������ 26
� ������������������������������������������������������������ = 1,296,539
������������������������������������������������ 29
And the constant is set so to satisfy:
������������������������������������ _������������������������������������������������������������������������������������������������������������_30 − ������������������������������������_������������������������������������������������������������������������������������������������������������_29 = ������������������������������������������������������������(������������������������������������ _������������������������������������������������������������ )
This gives a value of g of 3.397, and a value of the constant of 2,326. The value of g ensures that the series grows sufficiently to hit
1,296,539 total cases on November 26.
4
Figure 27: Methodological changes in reporting Figure 28: Resurgence of the pandemic in Autumn
Covid-19 infections in Turkey (cumulative) 70,000
3,000,000 Covid-19 infections in Turkey (daily)
Total Number of 60,000 Daily new cases, old
2,500,000 infections (patients) Old method
method 50,000 Daily change, synthetic
Total Number of
2,000,000 infections (patients) new 40,000
method Daily change, new
1,500,000 method (actual data)
30,000
1,000,000 20,000
500,000 10,000
0
0
Source: Ministry of Health, Turkey, WB staff calculations.
Source: Ministry of Health, Turkey. Note: Daily new cases, old method corresponds to the number of patients
announced by the Ministry of Health.
The second wave of the pandemic evidentially hit Turkey hard. As well as several other countries
including Germany, Italy, Russia and United Kingdom, Turkey has experienced a significant surge in the
number of COVID-19 cases starting from October 2020 (Figure 29). The size of the second wave was well
above the first wave, as daily cases rose above 40,000 in selected countries 3.
Figure 29: The second wave of the pandemic in autumn outpaced the first wave across countries
Evolution of Covid-19 pandemic (14-day moving average of daily cases)
60,000
50,000
40,000
30,000
20,000
10,000
0
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Mar-21
Jul-20
Jul-20
Oct-20
Jan-21
Feb-21
Germany Italy Russia Turkey United Kingdom France
Source: University of Oxford, WB staff calculations.
Note: Turkey’s Covid-19 cases in the July 29-December 10 period based on a WB Staff estimation as described in this section, for compatibility
with other countries.
Like many countries that experienced a surge in the pandemic, Turkey re-imposed restrictions as
COVID-19 increased again. The reopening that had started in June 2020 was reversed through the
reintroduction of lockdown measures starting in November, including all-day weekend curfews, night curfews,
restaurant and school closures, public transportation restrictions for the elderly, and reduced office utilization
for civil servants. Moreover, flight suspensions to several countries were introduced, particularly following
reports of emergent COVID-19 variants. The introduced measures had a relatively limited impact on
manufacturing activity compared to the early phase of pandemic as restrictions were less strict and mostly
3 This could also be related to the improved testing capacity of countries relative to the first wave.
5
focused on the services sectors, rather than manufacturing. Following an easing of restrictions in the first
quarter of 2021, restrictions were again tightened again in the middle of April following a sharp increase in the
number of cases.
The authorities extended salary support and introduced new measures to support suffering
households and businesses. Salary support schemes including short-term work allowance and support for
workers on unpaid leave was extended until the end of March 2021. The tax payments of closed businesses
were deferred, and a direct income support package for cafes and restaurants was introduced. In addition,
several private and state banks postponed loan repayments and eased conditions for NPL restructuring for
MSMEs and households.
Despite a significant rise in the number of cases, Turkey experienced relatively fewer deaths. The
cumulative number of COVID-19 cases in Turkey had reached 2.8 million by the middle of March 2021, with
more than 34 million COVID-19 tests conducted over the same period. Despite these high numbers, Turkey
has relatively fewer COVID-19 cases per million people than comparator counties, and has recorded amongst
the lowest deaths per million population (Figure 30, Figure 31).
Figure 30: Relatively low Covid-19 cases… Figure 31:… and also low losses per million
Total cases per million population Total deaths per million population
100,000 2500
90,000
80,000 2000
70,000
60,000 1500
50,000
40,000 1000
30,000
20,000 500
10,000
0 0
Source: Our World in Data.
B. A sharp rebound in late 2020, accompanied by rising pressures
The economy rebounded strongly in the second half…
Policy responses to the pandemic and its economic impact led to a sharp rebound in GDP in the
second half of 2020. GDP recorded the sharpest contraction (-11 percent, qoq) of the past decade in the
second quarter of 2020. However, driven by credit expansion and reopening, there followed a rapid rebound
in GDP. GDP grew by around 16 percent over the third quarter, largely driven by private consumption (Figure
32). The contributions of exports and investment were sizable as well. Positive growth in Q4 as well, driven by
exports, meant that Turkey was among the few countries to record positive overall growth, of 1.8 percent in
2020 (Figure 33). Both industry and services contributed to this rebound in the second half (Figure 34).
6
Figure 32: Sharp rebound in GDP in 2020 H2 Figure 33: Solid yoy growth in three quarters more
than made up for losses in Q2
GDP (growth and contribution, qoq, SA)
35%
25%
15.9%
15%
-11.0% 1.7%
0.4% 2.0% 0.1%
5%
-5%
-15%
2019Q3
2019Q4
2020Q1
2020Q2
2020Q3
2020Q4
Priv. Consumption Pub. Consumption
Investment Exports
Imports Stocks
GDP
Sources: Haver Analytics, TURKSTAT, WB Staff estimates.
Figure 34: …driven by both industry and services Figure 35: Not only consumption, but also a large
accumulation of finished goods drove growth
Gross Value Added (growth and contribution, GDP (growth and contribution, yoy)
qoq, SA) 8%
6%
15%
4%
10%
2%
5% 0%
0% -2%
-5% -4%
-10% -6%
-15% -8%
2019
2020
2019Q2
2019Q3
2019Q4
2020Q1
2020Q2
2020Q3
2020Q4
Industry Construction Priv. Consumption Pub. Consumption Investment
Services Residual
Agriculture Gross Value Added Exports Imports Stocks
Source: Haver Analytics, TURKSTAT, WB Staff estimates.
Manufacturing rebounded sharply in response to these developments. The Purchasing Managers’ Index
(PMI) for manufacturing indicates the speed at which manufacturing sector activity accelerated after April 2020
(Figure 36), industrial production restarted, and capacity utilization rates returned to pre-pandemic levels
(Figure 37). Turkey’s rebound in 2020Q3 was much sharper than the average for both emerging markets and
the Euro zone. However, Turkey’s PMI fell slightly in 2021Q1 whilst PMI in Euro area and the U.S. continued
to rise.
7
Figure 36: Turkey’s rebound was more notable Figure 37: Capacity utilization reaching almost pre-
compared to other countries. pandemic levels
PMI: Manufacturing (SA, 50+=Expansion) Capacity Utilization Rate (SA) and Industrial
65 Production (SA, mom % change)
60
80 30
55 20
75 10
50
70 0
45 -10
40 65 -20
-30
35 60 -40
30
Turkey Emerging Markets Capacity Utilization, LHS
Euro Area Benchmark
U.S. Industrial production, RHS
Source: Haver Analytics, TURKSTAT, CBRT and IHS Markit Economics.
The recovery in manufacturing was broad-based across all sub-sectors. Major export sectors, basic
metals, apparel and road vehicles sectors recorded strong rebounds (more than 25 points) in their PMI output
in 2020Q3 compared to the previous quarter (Figure 38). Following the stoppage in the second quarter,
manufacturing led firms in the EU started growing again, as indicated by PMI levels of above 50 from June
onwards, thereby increasing demand for intermediate inputs from Turkey. In addition to external demand, the
vehicles sector benefitted from a surge in domestic demand backed by favorable credit conditions specific to
vehicle purchases. The food sector, which contracted modestly unlike other sectors, had the lowest rebound.
The second round of restrictions led to a decline in the PMI output of all sectors at the end of the year.
Excepting the food and chemicals sectors, all sectors’ PMI (output) fell to below 50, marking the lowest reading
since the sectors began recovery from the pandemic-related downturn in June 2020.
Losses in labor productivity of the industry sector were reversed to some extent with the recovery in
economic activity. Public measures to stem job losses coupled with the drop in production in Q2 led to a fall
in production per employee of almost 20 percent, year-on-year (Figure 39). However, industry production per
hour worked actually rose, as hours worked fell dramatically. And as production in industry picked up again,
these trends reversed.
Figure 38: Large exporting sectors recovered fast Figure 39: Labor productivity per person recovered
Sector PMI, Output (SA. qoq changes) Index of Total Industry Production per Person
Employed and per Hours Worked (2015=100)
35 145
25
15 125
5
-5 105
-15
-25 85
-35
65
Index of Production Per person Employed
2020Q2 2020Q3 2020Q4 Index of Production per hours worked
Hours worked Index
Source: IHS Markit Economics, TURKSTAT, Ministry of Industry and Technology and CBRT.
8
All service sub-sectors rebounded as restrictions were eased in the second half of 2020. With strict
restrictions in place during most of 2020Q2, all services sectors except ICT suffered dramatically. Food and
accommodation, administration and support and transport and storage sectors were especially impacted.
Reopening that began in early June coupled with government stimulus led to a quick rebound in most services
sub-sectors (Figure 40). The reopening of restaurants and cafes and the resumption of domestic flights
supported the rebound in accommodation and food sectors. Foreign visitor arrivals increased gradually as
international flights resumed and land borders reopened in early June, and hotel occupancy rates had increased
to 45 percent by September (Figure 41). For real estate, abundant credit drove a record level of house sales.
Figure 40: A sharp recovery in all services sectors Figure 41: Gradual improvement in the tourism sector
Turnover Index deflated by CPI (SWDA) Foreign Visitors and Occupancy Rate
70
7000 90%
60
6000 80%
50 70%
5000
60%
40 4000 50%
30 3000 40%
30%
20 2000
20%
1000 10%
10
0 0%
Apr-19
May-19
Apr-20
May-20
Aug-19
Sep-19
Nov-19
Aug-20
Sep-20
Nov-20
Mar-19
Jun-19
Dec-19
Mar-20
Jun-20
Dec-20
Jan-19
Feb-19
Jul-19
Oct-19
Jan-20
Feb-20
Jul-20
Oct-20
Jan-21
May-19
May-20
Sep-19
Nov-19
Sep-20
Nov-20
Mar-19
Mar-20
Jul-19
Jul-20
Jan-19
Jan-20
Jan-21
Wholesale&Retail Transp.&Storage
Accom.&Food ICT Number of Foreign Visitors (in thous.)
Real Estate Prof.,Scient.&Tech.
Hotel Occupancy Rate, RHS
Admin.&Support
Source: Haver Analytics, TURKSTAT and Hotel Association of Turkey.
Confidence in services sectors also recovered, almost reaching pre-pandemic levels. Following a plunge
in confidence 4 in April, confidence in retail trade and other services sectors picked up quickly, driven mainly
by improvements in demand and price expectations over the following three months (Figure 42). A
TURKSTAT survey reports the factors limiting services sector activity. According to the survey results, whilst
more than 60 percent of respondents were reporting no limiting factor before the pandemic, this ratio dropped
to 30 percent in 2020Q2 and then rose to 50 percent recently with eased restrictions (Figure 43). In addition to
pandemic related factors (other factors), insufficient demand seems to be one of the main factors limiting
services sector activity, as expected. There is no remarkable change in limiting factors such as shortage of labor
force and financial constraints between pre- and post-pandemic periods.
As policy frameworks tightened, credit growth cooled, and the credit-fueled growth momentum
subsided. The tightening of credit conditions was quickly felt through a fall in credit growth (Figure 44). By
mid-February, Lira-denominated credit growth was at around 1 percent quarter-on-quarter. Consumption
indicators such as vehicle sales, imports and electricity usage also fell over this period, and on a seasonally-
adjusted, quarterly basis showed no, or negative, growth in the first quarter of 2021 (Figure 45).
4 Based on the TURKSTAT confidence survey. Sectoral confidence indices can take values of between 0 and 200. An index reading of
above 100 indicates an optimistic outlook and a reading of below 100 a pessimistic one.
9
Figure 42: Confidence picked up with eased restrictions Figure 43: The pandemic and insufficient demand are
Confidence Index (SA) Main factors limiting activities in the services
70 sector (%)
130
60
120
110 50
100 40
90 30
80 20
70 10
60 0
50
40
May-19
May-20
Sep-19
Nov-19
Sep-20
Nov-20
Mar-20
Mar-21
Mar-19
Jul-19
Jul-20
Jan-20
Jan-21
Jan-19
Insufficient demand
Shortage of labour force
Shortage of space and/or equipment
Financial constraints
Services Retail Trade Construction Other factors
No Limiting Factor
Source: TURKSTAT. Note: Figures indicate the ratio of the number of enterprises affected by
factor limiting activity, and the total number of survey responses.
Figure 44: Credit markets have been flat so far this Figure 45: …and domestic activity is declining
year…
Credit Growth Trend
(13-week, annualized trend, FX adjusted) High frequency activity indicators (SA, 3 month
100% percentage change)
120%
80% 100%
80%
60% 60%
40%
40% 20%
20% 0%
-20%
0% -40%
Aug-20
Aug-20
Sep-20
Sep-20
Nov-20
Nov-20
Dec-20
Dec-20
Dec-20
Jul-20
Jul-20
Jul-20
Oct-20
Oct-20
Jan-21
Jan-21
Feb-21
-60%
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Jul-20
Jan-20
Feb-20
Oct-20
Jan-21
Feb-21
Domestic Private Banks Foreign Banks
Sector State Banks Electricity Vehicle sales Imports
Source: Haver Analytics, BRSA, CBRT, WB Staff calculations. Source: Haver Analytics.
…as COVID-19 restrictions were relaxed
After imposing a comparatively moderate lockdown in March, Turkey began to slowly reopen from
June 1. The Oxford COVID-19 Government Response Tracker (OxCGRT) has become the most widely used
record of cross-country COVID-19 restrictions, and can be used to compare restrictions across countries 5.
Figure 46 shows the evolution of this index for Turkey over 2020, together with a selection of peer countries 6
and a record of Turkey’s daily COVID-19 deaths. The period of peak restriction in Turkey lasted between
March 28 and May 31, with reopening only beginning 40 days after the peak in the daily death rate, one of the
longest “wait times” in Europe. Whilst workplaces were never fully closed, the ban on non-essential travel was
only lifted in June, and public transport fully reopened in July. Schools only reopened in late September, which
together with Latvia was the longest school closure in Europe.
5 The database tracks information on containment and closure policies, such as school closures and restrictions in movement, and
compiles a daily index ranging from 0 (no restrictions) to 100 (full lockdown).
6 Bulgaria, Spain, France, United Kingdom, Georgia, Greece, Italy, Poland and Romania.
10
Figure 46: COVID restrictions were relaxed in two phases, in June and July
Turkey, Stringency index and daily deaths, Jan-Dec 2020
100 Stringency index (LHS) 300
90 Maximum lockdown restrictions Daily deaths (RHS)
(March 28 to June 1) Peer countries 250
80
70
200
Stringency index
60
Peak deaths
50 150
(first wave)
40
100
30
20
50
10
0 0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Source: Oxford COVID-19 Government Response Tracker, World Health Organization, World Bank staff estimates.
Turkey had the strongest post-lockdown economic performance amongst peers, but this was only
partially explained by reopening. Turkey’s economic activity, proxied by electricity consumption, increased
rapidly compared to the level predicted for a typical year, and rose much faster than peer countries, several of
which never saw their electricity consumption record a level above the predicted value (Figure 47). But the size
of Turkey’s actual rebound is considerably larger than predicted by COVID-19 reopening alone. Turkey
experienced a 15 percent increase in electricity use in June and a further 10 percent in July (Figure 48). Based
on recent research 7, the predicated impact of reopening on growth only explains around one third to one half
of the increase, suggesting that other factors, namely the large credit stimulus, drove most of the observed
economic expansion in the latter part of 2020.
Figure 47: Electricity use rebounded strongly in Figure 48: …but only partially explained by reopening
June…
Electricity consumption compared to predicted, Growth in electricity consumption relative to
Turkey and comparators, 2020 counterfactual baseline, 2020
25 20
20 15
15
10
10
5 5
0 0
-5 -5
-10
-10
-15 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
-20 Reopening impact
-25 Residual
Jan Feb Mar Apr May Jun Jul Aug Sep Seasonally adjusted change in electricity
Source: Demirgüç-Kunt et al (2020), TEİAŞ World Bank staff data.
7 Demirgüç-Kunt, Asli, Michael Lokshin, and Iván Torre. "Opening-up Trajectories and Economic Recovery: Lessons after the First
Wave of the COVID-19 Pandemic." World Bank Policy Research Working Paper 9480 (2020). Based on a cross-country analysis of
electricity consumption and OxCGRT data on COVID-19 restrictions, this research estimates that the move from a full to partial
lockdown sees an average 5 percent increase in economic activity (proxied by seasonally-adjusted electricity consumption), and the
move from partial to fully open a further 6 percent. That longer wait times and slower reopening is associated with a faster rebound
suggests that public trust and confidence that the pandemic is declining is important for economic recovery.
11
..and the government unveiled a large fiscal package and supported a wave of credit
Turkey’s overall fiscal support package was one of the highest among G20 emerging market
economies, but was overwhelmingly focused on support through the credit channel. Turkey’s fiscal
support package, at 11.5 percent of GDP (10.1 percent of GDP excluding deferred revenue), was the second
largest amongst emerging markets after Brazil, and the 12th largest amongst all G-20 countries (Figure 49). But
unlike most other economies, Turkey predominantly relied on indirect fiscal measures with 75 percent of the
economic support package consisting of contingent liabilities (loan guarantees and quasi-fiscal operations).
These measures were central to facilitating a large increase in credit, primarily via the state-owned banks. Direct
fiscal measures, though smaller in magnitude, were generally well-targeted and effective. This included support
to furloughed workers and public payments to workers on unpaid leave, additional social support transfers to
households, rent and revenue support to small businesses, and tax and social security contribution deferrals for
businesses.
Figure 49: Turkey’s fiscal package relies mostly on guarantees rather than direct expenditures
60.0
Fiscal support packages across G-20 countries (in percent of GDP)
50.0
40.0
30.0
20.0
10.0
0.0
Additional spending and foregone revenues Asset purchases and equity injections Contingent Liabilities Deferred revenue
Source: IMF Fiscal Monitor Database, January 2021.
The policy response action was effective in stimulating a sharp credit expansion in 2020. Rapid loan
growth was fueled by the countercyclical lending of state banks, an expanded CGF, and policy measures such
as the BRSA’s Asset Ratio, in place until end 2020, and that obliged banks to ease their cautious stance and
provide more credit. By the end of 2020, overall domestic Lira lending from deposit-taking institutions had
expanded by 42 percent over the year, of which state banks were responsible for 25 percent, private banks 10
percent and foreign banks 7 percent (Figure 50). The majority of the new lending went to corporate and
mortgage loans, and the vast majority of credit growth was Lira denominated 8. Overall, Turkey’s real credit
growth in 2020 was one of the highest amongst advanced and emerging economies (Figure 51).
8 FX credit expanding by only a tenth (in nominal Lira terms), largely as a result of measures put in place recently to discourage
domestic FX lending. FX lending declined around $6 billion dollars throughout 2020. This decline was partly due to liability de-
dollarization by firms incentivized by interest rate differential as well as the level of nominal exchange rate until after mid-year.
12
Figure 50: Huge credit stimulus, led by state banks Figure 51: Turkey’s credit stimulus was among the
largest amongst EMs and Advanced Economies
Credit growth in 2020 (cumulative % change) Real private credit growth in 2020 (cumulative
% change)
45% 30%
40% 25%
35%
20%
30%
25% 15%
20% 10%
15%
5%
10%
5% 0%
0% -5%
10/3/20
11/3/20
12/3/20
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Dec-19
Mar-20
Jun-20
Dec-20
1/3/20
2/3/20
3/3/20
4/3/20
5/3/20
6/3/20
7/3/20
8/3/20
9/3/20
Jul-20
Jan-20
Feb-20
Oct-20
Turkey Max EMs Max Adv
Private Foreign Public
Source: BRSA; Haver; World Bank Staff calculations. Source: Haver; World Bank Staff calculations.
Note: Deposit money banks Lira loans only. Notes: Maximum amongst 17 other emerging market economies and 11
advanced economies with data available. Nominal credit discounted using
consumer price inflation. Credit growth is not adjusted for FX valuation
effects for Turkey or comparators.
Turkey faced severe external pressure, and the Lira depreciated sharply
Turkey was exceptionally hard hit by capital outflows over 2020, with the central bank running down
reserves in response. In the aftermath of the first wave, many emerging markets (EMs) saw capital outflows,
and Turkey, along with Brazil and South Africa, was particularly hard hit, with these three countries
experiencing a ‘sudden stop’ 9(Figure 52). 10 Turkey’s very loose monetary policy, coupled with high inflation is
likely to have exacerbated capital outflows in Turkey over the period. As Turkey faced the full force of capital
flight, the Central Bank began to use international reserves at an unprecedented rate. In the first three quarters
of the year, more than US$40bn was recorded as use of reserves in official balance of payments data 11 (Figure
53). The lira nevertheless depreciated in the first half of the year, by 14 percent against a trade-weighted basket
of currencies (Figure 54). However, pressure on the lira continued, and by the end of the year, despite a
continued run-down of reserves, it had depreciated by 30 percent in 2020Q4 compared to 2019Q4 to become
the worst-performing of thirteen EM currencies. 12
9 Capital inflows fell to just over 2 standard deviations below their 5-year average in 2020 Q2. Eichengreen and Gupta (April 2016)
classify an episode as a sudden stop when: (i) non-resident portfolio and other investment inflows decline below the average of the
previous 20 quarters by at least one standard deviation; (ii) when the decline lasts for more than one quarter; (iii) and when flows are
two standard deviations below their prior average of at least one quarter. Episodes end when capital flows recover to the prior mean
minus one standard deviation
10 Capital flows in Turkey are highly volatile due to high dependence on external flows and high risk sentiment. The COVID-19
pandemic has exacerbated the volatility of capital flows and exerted pressure on the currency.
11 The Balance of Payments reports transactions and financial flows between residents and non-residents. It is also possible that
reserves may be used in transactions with residents which, while not being registered in the Balance of Payments, would still have an
effect on the exchange rate. Therefore, the use of reserves may be greater than is reported here. However, for the purpose of
comparison with other countries, this analysis maintains the BoP definition of use of reserves.
12 The Lira was the worst performing currency amongst currencies of a group of emerging market economies, which are
Malaysia, Indonesia, Thailand, Chile, China, Russia, Mexico, Argentina, Brazil, the Philippines, South Africa, and India.
13
Figure 52: Turkey experienced a sudden stop in 2020 Figure 53: Turkey’s central bank released a massive
amount of international reserves
Sudden stops in selected EMDEs Cumulative change of reserves in 2020, US$bn
0.0 10
Capital flows,. standard errors from mean
-0.5 0
-1.0 -10
-20
-1.5
Turkey -30
-2.0
Brazil -40
-2.5 S Africa
-50
-3.0 Q4-2019 Q1-2020 Q2-2020 Q3-2020 Q4-2020
Q4-19
Q1-20
Q2-20
TUR CHN RUS BRA ZAR
Figure 54: But the lira nevertheless depreciated Figure 55: Portfolio flows recovered in line with other
throughout the year EMs
Cumulative change of NEER in 2020 Weekly portfolio inflows (US$bn)
10% 2.0 30.0
5% 1.5 22.5
0%
1.0 15.0
-5%
0.5 7.5
-10%
0.0 0.0
-15%
-20% -0.5 -7.5
-25% -1.0 -15.0
8/21/20
9/18/20
10/2/20
1/22/21
2/19/21
10/16/20
10/30/20
11/13/20
11/27/20
12/11/20
12/25/20
9/4/20
1/8/21
2/5/21
-30%
Q4-19 Q1-20 Q2-20 Q3-20 Q4-20
TUR CHN RUS BRA ZAR Turkey Tur avg
Other EMs EM avg
Figure 56: Lira volatility has been lower
TL-US$ exchange rate, 20-day average and
implied volatility
8.2 30
8.0 25
7.8
20
7.6
15
7.4
10
7.2
7.0 5
6.8 0
01-Oct-20
15-Oct-20
29-Oct-20
07-Jan-21
21-Jan-21
04-Feb-21
18-Feb-21
12-Nov-20
26-Nov-20
10-Dec-20
24-Dec-20
Exchange rate (LHS) Implied volatility (RHS)
Source: Haver Analytics; Bloomberg, JP Morgan, International Institute of Finance
Note: One-month implied volatility of US$-TRY in the options market
14
Portfolio flows to Turkey then rebounded, influenced by both domestic policy and global liquidity
conditions. Portfolio inflows, which remained negative on average up to the end of October, have moved into
positive territory with an average net inflow of US$300 million each week since then (Figure 55). While this
positive reversal has been supported by Turkey’s tightening of monetary policy, it also reflects a broader return
to EMs. The rebound in portfolio flows has only slightly outperformed the EM average, and this underlines
the ongoing criticality of global market sentiment on EMs for Turkey, in addition to domestic conditions.
Turkey’s sovereign credit default risk spread price, which had reached 535 basis points in October, then almost
halved to 283 basis points by early February, and while the Lira has steadily appreciated against the US$, its
volatility (measured as the range between minimum and maximum daily values over the past 20 days) has
trended downwards (Figure 56).
Merchandize exports and services are now recovering in line with demand in advanced Europe. From
October onwards, merchandize exports had recovered above their average level in 2019, and were 5 percent
higher by January, and the trade deficit was narrowing (Figure 58). This export performance is in the mid-range
of emerging markets in the same neighborhood (Figure 57), and generally better than advanced economies.
However, activity indicators like the Purchasing Manager Index indicate that Eurozone activity remains weak,
with German output growing only slightly (Figure 61). Given that Europe is Turkey’s largest export market,
this weak demand is likely to have dented the export recovery to date. Exports to non-EU countries also show
a strong upward trend (Figure 59) and Turkish PMI suggests that new export orders are losing pace (Figure
62). Services exports are similarly growing, and by December had begun to bring down the current account
deficit (Figure 60).
Figure 57: Exports recovered relatively well… Figure 58: …and the trade balance is narrowing
Merch. Exports (percentage of 2019 average) Merchandize trade aggregates (US$bn, 3-month
average)
1.20 22 6
1.10 20 5
1.00 18 4
0.90 16 3
0.80 14 2
0.70 12 1
0.60 10 0
Apr-20
May-20
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Aug-19
Sep-19
Nov-19
Aug-20
Sep-20
Nov-20
Dec-19
Mar-20
Jun-20
Dec-20
Jun-20
Dec-20
Dec-19
Mar-20
Mar-21
Feb-20
Jul-20
Oct-20
Oct-19
Jan-20
Feb-20
Jul-20
Oct-20
Jan-21
Feb-21
Jan-20
Jan-21
Turkey Bulgaria Poland Romania Trade deficit (RHS) Exports (LHS)
Serbia UK USA Germany Imports (LHS)
Italy Greece
Source: Haver Analytics. Source: TURKSTAT.
15
Figure 59: Export growth to non-EU markets is Figure 60: Net services receipts also strengthening…
stronger
Merch. Exports by Country Aggregate Current account balance & components (SA,
(percentage change in 3m sum) US$bn)
1.30 4.0
3.0
1.20 2.0
1.10 1.0
0.0
1.00 -1.0
0.90 -2.0
-3.0
0.80 -4.0
0.70 -5.0
-6.0
0.60 -7.0
Apr-20
May-20
May-19
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Aug-19
Sep-19
Nov-19
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Jun-19
Dec-19
Mar-20
Jun-20
Dec-20
Jul-20
Jul-19
Jul-20
Jan-20
Feb-20
Oct-20
Jan-21
Feb-21
Oct-19
Jan-20
Feb-20
Oct-20
Jan-21
Feb-21
EU non-EU Goods Services Incomes CAB
Source: Haver Analytics. Source: Central Bank of Turkey.
Figure 61: …but limited by weak demand in Figure 62: …and new demand is weakening
destination markets
Composite PMI Output PMI: New export orders (>50 signals growth)
60 60
58 55
56 50
54
52 45
50 40
48 35
46 30
44 25
42
40 20
15
Aug-20
Sep-20
Nov-20
Dec-20
Jul-20
Oct-20
Jan-21
Feb-21
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Jan-20
Feb-20
Jul-20
Oct-20
Jan-21
Feb-21
Euro area France
Germany UK Monthly 3-month average
>50 signals growth
Source: IHS Markit via Haver Analytics. Source: IHS Markit via Haver Analytics.
Inflation accelerated as the effects of monetary and credit stimulus played out
Inflation accelerated in the second half of 2020 due mainly to depreciation of the Lira, rising
international commodity prices and deteriorating inflation expectations. Consumer price index (CPI)
inflation reached 15 percent - three times the target - and domestic producer price inflation (PPI) hit 25 percent
at the end of 2020 (Figure 65). CPI and PPI further accelerated to 16.2 percent and 31.2 percent in March,
respectively. Inflation has persistently overshot the 5 percent target over the past decade with sizable deviations
in recent years and rising inflation expectations 13 (Figure 63). Turkey recorded the highest deviation of inflation
from its target among many peer countries in 2020 (Figure 64). The diffusion index also reached its highest
level since the August 2018 shock, indicating strong broad-based price pressures (Figure 66). The sharp
depreciation of the lira coupled with high exchange rate pass-through, the credit-driven boost to domestic
demand and deteriorating inflation expectations led to a jump in consumer and domestic producer prices (Box
13 Kose et al (2019) found for a group of emerging and developing economies that inflation targeting, high central bank transparency,
strong trade integration and a low level of public debt are associated with better-anchored long-term inflation expectations.
16
1). Supply constraints and rising international commodity prices also exerted pressure on producer prices in
addition to the effects of exchange rate depreciation. Inflation remained high in 2021Q1.
Figure 63: Deviation from the target and inflation Figure 64: Turkey shows the largest deviation among
expectations move in parallel peer countries
Inflation expectations (12 months ahead) and Inflation deviation from the target 2020 (perc.
22 deviation from the target (%) points)
CHI
17
IDN
12 ROM
POL
7
UKR
2
BRA
IND
Deviation from the target (Realization-Target), perc. points TUR
Inflation Expectations, 12 months ahead. % -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
Sources: Haver Analytics, TURKSTAT, WB Staff estimates.
Figure 65: Inflation resurged in the second half of Figure 66: …with strong broad-based price pressures
2020…
Inflation and Exchange Rate (yoy, % change) Diffusion Index for CPI Inflation
35 50 1.0 0.95 0.7
31.2
30 40
0.9 0.6
25 30 0.75
0.5
20 17.3 20 0.8 0.69
0.4
15 10 0.7 0.15
16.2 0.3
10 0
0.6 0.2
5 -10 0.50 0.29
0.5 0.1
0 -20
The fraction of items above 10% inflation
Exc. Rate (TL/$), RHS CPI Inflation The fraction of items with increasing prices
Core Inflation PPI Inflation The fraction of items above 20% inflation, ra
Sources: Haver Analytics, TURKSTAT, WB Staff estimates.
Food and core goods, particularly durables, were the largest contributors to inflation. Food and core
goods accounted for almost 65 percent of the rise in inflation by the end of the year, while the contribution of
energy was negligible (Figure 67). Among food items, the rise in the price of unprocessed food was notable
(26.3 percent yoy in December). Currency depreciation, seasonal factors, drought and an upward trend in
international food prices were the main drivers of high food inflation. 14 Over most of the past decade, food
price inflation in Turkey has been significantly higher than international food prices. However, by end 2020,
international food prices had reached a six-year high and were exerting upward pressure on domestic food
prices (Figure 68). Rising food prices then had a knock-on effect on prices in hotels, cafes and restaurants
14Considering the importance of reducing food inflation, especially for price stability, the government recently took a step to perform
an early warning function by carrying out a detailed and timely analysis of data on the prices of food and agricultural products. In this
regard a new unit the “Food and Agricultural Product Markets Analysis Division” was established at the CBRT. The recently announced
Economic Reform Package includes measures to address structural issues contributing to price volatility in the agriculture sector,
including ensuring intermediaries operate competitively and efficiently.
17
services, which reached 13 percent. Transportation inflation has also started to pick up. Core goods inflation
rose to 17.2 percent (yoy) at the end of the year, driven mainly by a sharp surge in durable goods inflation (30
percent yoy) on the back of strong credit stimulus.
Figure 67: Driven mostly by core goods and food Figure 68: Rising international food prices add
prices pressure on domestic prices
CPI (yoy change and contribution, %)) Turkey, vs. international food price inflation
20 (yoy % change)
45
15 35
25
10
15
5 5
-5
0
-15
Energy Food&Bev.
Core Goods Services International Turkey-Unprocessed food
CPI Turkey-Processed food
Sources: Haver Analytics, TURKSTAT, WB Staff estimates.
Prices of durable goods, which are more sensitive to exchange rate developments and credit
momentum recorded a sharp rise. The credit-driven demand stimulus was reflected as a big price hike in
some durable goods such as automobiles, furniture and white goods, which are more exposed to credit
momentum and exchange rate developments. Automobile prices recorded 42 percent (yoy) price hikes at the
end of the year, whilst the prices of white goods and household appliances increased by 25-30 percent (yoy)
(Figure 69). The increase in special consumption tax rates on cars in 2020 also contributed to the rise in
automobile prices. Similarly, furniture prices rose by around 20 percent (yoy) as home purchases reached record
levels.
Figure 69: Inflation of some durable goods jumped Figure 70: Consumer consumption patterns have
backed by massive credit expansion changed with the COVID-19 pandemic
Durable Goods Inflation (NSA, yoy %) The Shares in Debit and Credit Card
45 Expenditures (%)
9
35 8 Electronics Various Food
25 Building
7 supplies Clothing
15 6 Insurance
5 5 Petrol Stations
2020
Furnishing
-5 Telecom
4
Airlines,
-15 3 accomodation
and travel
2 Health
agencies
1
Vehicles Furniture
0
Durable goods (exc. gold) Major HH appliances
0 1 2 3 4 5 6 7 8 9
Small electric HH appliances 2019
Sources: Haver Analytics, TURKSTAT, WB Staff estimates. Source: CBRT.
Note: The biggest share belongs to Market and Shopping Centers, which
was at 16.9 percent in 2019. The share increased by more than 4 percentage
points to 21.1 percent in 2020. As it has a high share compared to other
items, it was not presented in the figure.
18
Consumption patterns changed dramatically during the COVID-19 pandemic, understating the actual
rate of consumer inflation. During the pandemic, consumers decreased their spending on transport,
restaurants, hotels, recreational, cultural and sporting events, and clothing, while increasing their spending on
food worldwide (IMF, 2020). 15 Credit and debit card expenditures in Turkey confirm similar trends and show
more spending at shopping centers, on electronics and food and less spending on clothing and fuel. The share
of airlines, accommodation and travel agencies in the total almost halved in 2020 compared to 2019 (Figure
70). These changes in spending patterns were reflected in CPI weights in early 2021. 16 Applying these new
weights to 2020 CPI expenditure groups shows that CPI inflation would have been 0.4 percentage points
higher. 17
Box 1: Drivers of Core inflation-Phillips Curve Estimation
Core inflation accelerated rapidly in 2020, particularly in the second half 2020 to over 14 percent at the end
of 2020. Sharp currency depreciation was among the key drivers of this rise. However, other factors such as
inflation expectations played a role. For a better understanding of inflation dynamics in the recent period
and to decompose core inflation (defined as CPI excluding unprocessed food, alcoholic beverages, and
tobacco) in terms of drivers, a reduced-form Phillips curve is estimated:
πt =α1πt-1 + α2πexp+ α3gapt + α4∆exct + α5∆exct-1 + α6∆pimpt + µt (1)
where πt is quarterly (seasonally adjusted) annualized core (D index-excluding unprocessed food, alcoholic
beverages and tobacco) inflation; πexp is 12 months ahead inflation expectations; gapt is the output gap; exct
is the quarterly average exchange rate (TL/US$) and pimpt is US$ denominated import price. The model
covers the inflation targeting regime period of 2006:1-2020:4.
Table 1: Phillips Curve Estimation Results
Dependent Variable: CPI inflation (quarterly, seasonally adjusted annualized), 2006Q1-2020Q4
Model 1
Lagged inflation 0.32*
(0.084)
Inflation expectations (12 months 0.65*
ahead) (0.100)
Output gap 0.19***
(0.106)
∆ Exchange rate 0.34*
(0.054)
∆ Exchange ratet-1 0.15*
(0.056)
∆ Import Price 0.59*
(0.089)
Dum20191 -0.15
(0.026)
Adjusted R-squared 0.82
Note: Standard errors are in parentheses. *, ** and *** refer to statistical significance at 1, 5 and 10% level.
Constant parameter estimation results suggest that the largest contribution to recent inflation comes from
inflation expectations with a coefficient of 0.65 and exchange rate depreciation with a coefficient (sum of
15 Under-weighting of rising food prices and over-weighting of falling transport prices are the main causes of the underestimation of
inflation. The COVID-19 indexes based on spending changes in CPI imply that CPI underestimated inflation over February-May 2020
for 65 out of 83 individual economies (IMF, 2020).
16 In the basket, the weights of food and beverages, housing and furnishings were increased, whilst the share of services items was
decreased.
17 If we apply 2021 shares to CPI inflation for the period of April-December 2020, inflation rises by 0.1 percentage points.
19
exchange rate coefficients) of around 0.5. This impact is not only felt through imported consumption goods,
but also through the sharp rise of cost of production and its pass through to consumer prices. The coefficient
of import price inflation is around 0.6. Inflation in Turkey has strong inertia with the coefficient of past
inflation at around 0.32. The coefficient of the output gap is estimated to be around 0.20. The estimations
of different samples (2006:1-2020:4) suggest that the coefficients change over time.
To capture the behavioral change in coefficients, a time varying parameter reduced-form Phillips curve is
estimated. The estimation results show that there is a variation in the coefficients of the model. Both the
coefficients of expectations and past inflation have also declined over time, albeit slightly. While the
coefficient of the output gap is found to be relatively stable, the pass-through from the exchange rate, on the
other hand, has risen notably. These findings seem to be consistent with the results of Kara et al. (2017) and
Koc et al. (2021).
Based on the time varying parameter estimation results, core inflation is decomposed into contributions from
its determinants for 2018-2020 on a quarterly basis. The results indicate that one of the main drivers of core
inflation was sharp depreciation of the Lira, particularly in the second half of 2020 due to rising
macroeconomic imbalances triggered by massive credit expansion and a relatively loose monetary stance. In
addition to the rise in prices of imported consumption goods, the pass-through from the rising cost of
production due to deprecation and pandemic-related factors on consumer goods impacted inflation
adversely as well.
Figure 71: Time Varying Parameters of the Model 1
0.35 0.75
0.7 0.47
0.65 0.42
0.3 0.37
0.6
0.32
0.55
0.25 0.27
0.5
0.22
0.45 0.17
0.2 0.4 0.12
2006Q1
2007Q1
2008Q1
2009Q1
2010Q1
2011Q1
2012Q1
2013Q1
2014Q1
2015Q1
2016Q1
2017Q1
2018Q1
2019Q1
2020Q1
0.07
Past Inflation
12 months ahead inflation expectations, ra Exchange rate Lagged Exchange rate
0.3 0.7
0.25
0.6
0.2
0.15 0.5
0.1
0.4
0.05
0 0.3
Output gap Import Price
Sources: Haver Analytics, TURKSTAT, WB Staff estimates.
Note: These coefficients are obtained from time varying parameter estimation with flexible least squares Eviews add-in ‘Tvpuni’ was used. Flexible
least square estimation with Kalman filter produces similar results regarding the coefficients of exchange rate and import price. However, the
coefficient of inflation expectations, past inflation and output gap is found to be relatively stable compared to the above results.
20
The exchange rate pass through from producer prices to consumer prices tends to be relatively low during
periods of weak demand. The demand for product groups differed during the pandemic due to strong credit
stimulus. Thus, as indicated in the recent inflation report of the CBRT, the exchange rate pass-through was
high in product groups where demand was high supported by credit stimulus, such as white goods and
automobiles. On the other hand, depreciation of the TL had a more limited impact on clothing and some
services items due to weak demand.
Strong recovery in 2020H2 reversed the demand driven disinflationary impact on core inflation. Massive
credit expansion led to a big jump in demand for durable goods such as automobiles, electronics and white
goods, and led to demand pressure on inflation. On the other hand, demand for clothing and hardest-hit
services items such as accommodation, travel and entertainment were muted. The contribution of the output
gap to core inflation was negligible in 2020, as the output gap moved to positive territory in the second half
of 2020.
On the import price side, there was no contribution to core inflation, unlike the previous year. This was due
mostly to weak global demand, particularly in the first half of 2020.
Inflation expectations also had a significant impact on the rise of inflation in 2020, being the largest
contributor. This underlines the importance of inflation expectations in controlling inflation. Gulsen and
Kara (2021) found that expectations seem to be more sensitive to exchange rate movements during
depreciation periods. This also confirms that exchange rates play a more important role in driving inflation
expectations, beyond the dimension of pass-through to domestic prices.
Figure 72: Contributions to Core Inflation (annualized, percentage points)
30
25
20
15
10
5
0
-5
-10
Inflation Exp. Past Inflation Exchange Rate (both level and its lag) Output gap Import Price
Sources: Haver Analytics, TURKSTAT, WB Staff estimates.
Corporate and bank balance sheets faced pressure, obscured by forbearance measures
The corporate deleveraging process, initiated in 2018, was interrupted by the COVID-19 pandemic.
From late 2018, corporate debt started falling, from 71.4 percent of GDP in 2018Q2 to 65 percent at the end
of 2019. But the need for liquidity following the COVID-19 disruption saw corporate debt leaping back up to
around 77 percent by 2020Q3. Whilst corporates continued FX deleveraging process, they increased their TL
borrowing which significantly contributed to the rise in total debt ratio. Indeed, among peer countries, Turkey
21
saw the sharpest increase (more than 11 percentage points) after Russia in terms of the corporate financial debt
to GDP ratio in 2020 (Figure 73) 18. Most of this increase stemmed from TL denominated debt. 19
Figure 73: Corporate debt burden remains relatively Figure 74: Net FX position of corporates has declined
high
Credit to Non-Financial Corporations (% share Net FX Position of Corporates (Billion $)
in GDP)
100 15 140
70
80
0
10
60 -70 -157.7
-180.9 -171.8 -166.6 -162.4
40 -140
5 -210
20
-280
0 0 -350
Assets Liabilities Net FX Position
2020Q3 Difference from end-2019 (perc. points), RHS
Sources: BIS and CBRT.
The net FX position of corporates improved during 2020. FX liabilities declined by US$6 billion over 2020
as corporates refinanced with lira debt (Figure 74). Short term domestic FX debt to the domestic banking sector
declined by US$4.4 whilst import payables rose by US$2.8 billion in the same period. FX assets of corporates
on the other hand have increased by around US$20 billion largely through deposits in both domestic and foreign
banks. As a result of these changes, the net FX position improved by more than US$20 billion to US$158
billion (around 22 percent of GDP).
Monetary tightening started in late November 2020 led to rises in the external debt rollover rates of
corporates. TL lending rates, which dropped to below 10 percent at the end of 2020Q2 amid a loose monetary
stance, started to rise with the policy shift in late November 2020. The weighted average interest rates for
commercial loans almost doubled to 19 percent at the end of 2020. This led corporates to decrease their
borrowing in TL terms. This is reflected in rising external debt rollover rates of close to 100 percent for
corporates through the end of the year, supported by a relatively lower cost of borrowing in FX.
Accumulation of debt is evident among listed companies and solvency pressures persist due to high
levels of overall debt. Deleveraging of listed companies was also interrupted in early 2020 and an index of
debt to equity displayed a sharp rise in 2020 (Figure 75). Turkey’s increase in the debt-to-equity ratio was sharper
than the equivalent for the MSCI Emerging Markets Index. 20 Elevated corporate debt coupled with a plunge in
earnings put pressure on the interest coverage ratio (ICR) 21. The ICR fell to levels very close to the sustainability
threshold in the first half of 2020. However, the subsequent surge in demand and rapid rise in inflation led to
a recovery in nominal earnings in the second half of 2020, and thus a rebound in the ICR (Figure 76).
18 The corporate financial debt to GDP ratio of Russia and Chile increased by around 13 percentage points over the same period.
19 Around 33 percent of this increase is due to currency depreciation and the remainder is due to TL loan increase. Tighter macro-
prudential regulations since May 2018 made it more difficult for some relatively small-sized corporates to access FX loans from the
domestic market.
20 The MSCI Emerging Markets is an international equity index that tracks stocks from 24 emerging market countries, including
Turkey. All corporates, both financial and non-financial, are presented for comparison with other emerging market economies.
21 The ratio of earnings before interest and tax to interest expenses.
22
Figure 75: Turkish traded companies increased their Figure 76: Contributing to solvency pressures
leverage during the pandemic
Total Debt to Shareholder Equity Index Interest Coverage Ratio for listed companies
170 3.5
150
3
2.5
130 2
1.5
110
1
90 0.5
0
MSCI Emerging Markets Index
BIST Istanbul 100 Index All listed Non-Financial Corporates
Average (2013Q1-2020Q4) Threshold
Source: CBRT.
The pandemic led to a rise in corporate vulnerability and disrupted cash flows. Reduced earnings
induced by the COVID-19 pandemic have deepened existing corporate solvency and liquidity problems.
Corporate vulnerability – as measured by probability of default – has risen sharply since August 2020, and
remains high compared to selected peer countries (Figure 77). Recent analysis of the effects of COVID-19 on
firms’ liquidity by the CBRT 22 shows that the number of illiquid firms increased in the second quarter of 2020,
declined in the third quarter supported by credit expansion, and increased slightly at the end of 2020 with the
impact of reintroduced restrictions (Figure 78). The negative impact of the pandemic is more evident for smaller
firms. As expected, the services sector has the highest ratio of illiquid firms to total number of firms.
Figure 77: Corporate vulnerability elevated Figure 78: Reflected to a rising number of illiquid
firms
Corporate Vulnerability Index Number of Illiquid Firms by Size (Thousands)
(14-day moving average) 250
90
80
70 200
162
60 144
150 125 129
50
40
30 100
20 49 52
10 50 29 29
8 2 9 2 5 1 5 1
0
0
2020Q1 2020Q2 2020Q3 2020Q4
Turkey Greece Indonesia Micro Small Medium
China Malaysia Thailand Large Total
Source: NUS Credit Research Initiative, WB Staff estimates. Source: CBRT, Inflation Report, 2021:1.
Note: Corporate vulnerability indices (CVI) are value weighted. The value-
weighted CVI sums up the individual probability of defaults with their
market capitalizations as weights.
22 CBRT, Inflation Report 2021-I, Box. 2.2.
23
Despite the impact of COVID-19 on the private sector and individuals, there has not been a
deterioration in reported bank asset quality metrics due to rapid credit growth and forbearance 23. Rapid
loan growth, payment deferrals, and forbearance measures have helped banks keep asset quality metrics
favorable. Official NPL ratios across all banks had declined from 5.4 percent in December 2019 to 4.1 percent
as of February 2021 (Figure 79). Stage 2 loans also declined from 12 percent in 2019 to 10 percent as of 2020
Q3 (Figure 80). Regulatory forbearance measures make it more challenging to assess banks’ true asset quality 24,
which is expected to have declined. The NPL ratio is likely to assume an upward trend in the period ahead due
to a combined effect of phasing out regulatory forbearance measures related to forborne exposures, a slowdown
in credit growth, and loans that were restructured, or that benefitted from postponed installments falling due. 25
Figure 79: Forbearance masks true NPLs… Figure 80: …and Stage 2 loans…
10%
NPL ratios 55 Distressed Assets 15
(February 2021) 50 (2020Q3, %)
8%
45
40
6% 35 10
30
4% 25
20 5
2% State Banks NPL 15
Restructured / Stage 2 Loans
Foreign Banks NPL 10
Domestic Private Banks NPL 5 Stage 2 Loans / Gross Loans (rhs)
0%
0 0
Apr-18
Apr-19
Apr-20
Jul-18
Oct-18
Jul-19
Oct-19
Jul-20
Jan-18
Jan-19
Jan-20
Oct-20
Jan-21
Sep-18
Sep-19
Sep-20
Dec-17
Mar-18
Jun-18
Dec-18
Mar-19
Jun-19
Dec-19
Mar-20
Jun-20
Source: BRSA, Fitch Ratings, Haver, and Staff calculations.
Elevated corporate sector vulnerability due to COVID-19 remains the key risk to banks susceptible to
FX and asset quality shocks. Banks’ ability to restructure corporate debt in the near term is diminished by
significant restructuring following the 2018 currency shock and an environment of higher interest rates. The
loan-to-deposit ratio continued its sustained decline from over 120 percent in mid-2018 to just below 110
percent in December, though the decline stems mostly from reducing FX lending’s share in total loans (Figure
81). The slowdown in credit growth may help banks restore their balance sheets in the medium term, but
ongoing COVID measures will continue to weigh on corporates’ ability to service their new and restructured
debt in a higher interest rate environment.
Banks continued to increase provisions for NPLs, despite forbearance measures in place. While
extended forbearance measures helped alleviate pressures on reporting asset quality deterioration until the end
of 2021Q2, banks have continued to increase IFRS-9 based provisions to build buffers against the expected
rise in NPLs (Figure 82). The reserve coverage ratios for NPLs have increased to 81.5 percent (70.8 percent in
2019 Q4), 72.5 percent (60.8 percent in 2019 Q4), and 72.6 percent (64.1 percent in 2019 Q4) for state banks,
domestic private banks, and foreign banks, respectively. Most of the banks also set aside free provisions for as
an extra measure for NPL increases.
23 Regulatory forbearance on loan classification was introduced in March 2020, and has recently been extended to the end of 2021 Q2.
According to the measures, overdue loans are classified as NPLs after 180 days (previously 90 days) and overdue distressed assets are
classified as Stage 2 loans after 90 days (previously 30 days).
24 FinSAC on Borrower Relief Measures in the ECA Region, cautioned against relaxing regulatory definitions for NPLs and
forborne exposures as this can undermine market discipline and distort the reliability of bank’s financial information. The
paper suggested that policymakers need to prepare and set clear exit strategies.
25 Please also see, CBRT Financial Stability Report - November 2020, Volume 31.
24
While banks entered the COVID-19 crisis with higher capital buffers, built after the turmoil of 2018,
they have also benefitted from forbearance measures. Banks had gradually built up their buffers after the
2018 currency shock, with an increase in the Capital Adequacy Ratio (CAR) from 16 percent in July 2018 to
18.4 percent by December 2019 (Figure 83). CAR declined slightly in 2020 Q1 to 17.9, but would have
deteriorated more were it not for the help of forbearance. 26 State bank’s CAR position declined more rapidly
in the first half of 2020 given high credit growth, but was buttressed by capital injections from TWF to state
banks 27 in May, and sector’s CAR rose to 18.7 percent in December. Risk of slower growth, an increase in
distressed assets, and further TL depreciation could weigh on risk-weighted assets and capital adequacy going
forward.
Figure 81: …but rebalancing continues Figure 82: Banks remain prudent with provisions
Loans to Deposit Ratios 100% NPL Provisions
130 (December 2020, %)
(February 2021)
120
80%
110
100
60%
90 State Bank NPL Reserves
Foreign Bank NPL Reserves
80 Domestic Private Bank NPL Reserves
40%
Dec-08
Dec-09
Dec-10
Dec-11
Dec-12
Dec-13
Dec-14
Dec-15
Dec-16
Dec-17
Dec-18
Dec-19
Dec-20
Apr-17
Apr-18
Apr-19
Apr-20
Jul-17
Oct-17
Jul-18
Oct-18
Jul-19
Oct-19
Jul-20
Oct-20
Jan-17
Jan-18
Jan-19
Jan-20
Jan-21
Figure 83: Capital Adequacy may deteriorate due to Figure 84: Improvement in risk premia encouraged
the rise in NPLs and stage 2 loans banks to tap external finance
20 Capital Adequacy and NPL ratios 6 150 External Debt Roll-over ratios
(%, January 2021) (%, December 2020)
130
5
16
110
4
90
12
Capital Adequacy Ratio 3
70
NPLs / Total Loans (rhs) Corporate Sector Banks
8 2 50
Sep-18
Sep-19
Sep-20
Dec-17
Mar-18
Jun-18
Dec-18
Mar-19
Jun-19
Dec-19
Mar-20
Jun-20
Dec-20
Aug-17
Aug-18
Aug-19
Aug-20
Dec-17
Dec-18
Dec-19
Dec-20
Apr-18
Apr-19
Apr-20
Sources: Haver Analytics, BRSA, CBRT, WB Staff estimates.
26 Capital adequacy related forbearance measures include the use of 2019 year-end FX rates in calculating foreign currency
denominated risk-weighted assets (RWA) and the suspension of mark-to-market losses on securities in capital adequacy calculations
(both due to be retracted at end-1H21), as well as a 0% risk-weighting on government foreign currency denominated bonds. The FX
rate has been changed to a rolling 252-day average in December 2020, which decreased the impact of forbearance significantly. Apart
from forbearance, CGF loans also have a low risk weighting for RWAs.
27 Turkey Wealth Fund (TWF) has contributed TRY 21 billion (US$ 3 billion) to shore up the capital of three public deposit banks.
25
Access to external financing has improved since 2018, but maturity mismatches remain. Banks, along
with corporates, managed to increase their rollover ratios (Figure 84) and access syndicated loans at a reasonable
cost during 2020 amid a chronic shortage of long-term financing (both external and domestic) coupled with
persistent maturity mismatches. As of 2020 Q3, the share of assets having a remaining maturity of less than a
year increased to 47 percent from 44 percent in 2019 Q4. The percentage of liabilities having a remaining
maturity of less than a year also increased to 79 percent from 75 percent. Domestic and global monetary
tightening pose risks to constraining loan tenors, while feeding additional vulnerabilities through heightened
rollover risks for corporates.
The banking sector reduced exposure and market-based measures of risk, and volatility fell. Aided by
credit cooling and a return to more orthodox monetary and economic policies, state banks’ FX net positions
improved from November, and were effectively closed by January (Figure 85). However, the banking sector is
still exposed to currency mismatches 28 and refinancing risk, especially from deposit dollarization 29 that remains
high, and that increased over the course of 2020 (Figure 86).
Figure 85: State banks’ FX net open position recovers Figure 86: Deposit dollarization increased over 2020
FX net Position/Regulatory Capital Share of non-Lira deposits
15 (%, February 2021) 60%
58%
5
56%
-5 54%
52%
-15 50%
Sector
State 48%
-25
Foreign 46%
Domestic Private 44%
-35
1/3/20
2/3/20
3/3/20
4/3/20
5/3/20
6/3/20
7/3/20
8/3/20
9/3/20
1/3/21
2/3/21
3/3/21
10/3/20
11/3/20
12/3/20
Nov-18
Jan-19
Nov-19
Jan-20
Nov-20
Jan-21
Jul-18
Jul-19
Jul-20
May-18
Sep-18
May-19
Sep-19
May-20
Sep-20
Mar-18
Mar-19
Mar-20
Source: Haver Analytics, BRSA, CBRT, WB Staff calculations.
C. Policy frameworks were adjusted to restore macroeconomic stability
The fiscal policy framework remained strong, despite COVID-19 pressures
The 2020 fiscal deficit was moderate as targeted stimulus expenditure was offset by expenditure
reprioritization and rising tax revenues later in the year. Turkey’s fiscal position remained in noticeably
good shape with a budget deficit of 3.4 percent, lower than many countries around the world and less than the
revised 30 budgeted deficit of 4.3 percent. Total expenditure realization was in line with the revised estimate, and
10 percent higher than the original budget estimate for 2020 (Figure 87). Revenue exceeded the projection by
6 percent (Figure 88). Almost two thirds of the extra revenue stemmed from indirect taxes, following the boost
to demand from the credit stimulus. Turkey’s limited fiscal deterioration was accompanied by a massive credit
stimulus, which may be reflected on the budget in future periods if and when contingent liabilities are realized.
Current transfers accounted for almost half of the expenditure increase beyond the budget target.
28 The banking sector is offsetting its on-balance sheet FX position with off-balance sheet transactions, mostly with currency swaps,
including the swaps with the CBRT. While off balance sheet transactions may limit the exposure, they cannot solely protect the
balance sheets against any external FX shocks.
29 BRSA data shows that the share of FX deposit in total deposits is 55 percent as of April 2, 2021.
30 This reflects the budget deficit cited in the New Economic Program for 2021, published in September 2020. This expenditure
estimate accounts for in-year budget adjustments that were already made early in 2020 in response to COVID-19.
26
Expenditure increased by almost one percent of GDP, driven by COVID-19 related transfers to
households and firms and rising interest costs. Total expenditures as a share of GDP rose by 0.9 percentage
points in 2020, driven largely by current transfers. The authorities adjusted spending on capital and goods and
services to help create fiscal space for public transfers, particularly for vulnerable households affected by the
downturn. Most central government transfers are for health, retirement and social aid expenses, which
accelerated rapidly in 2020. In fact, current transfers accounted for almost 60 percent of the increase in
expenditures (Figure 90). Employment support programs (short-term work allowance, unpaid leave, etc.) are
financed through the unemployment insurance fund, and so do not appear directly on the central government
budget. Interest costs rose to 2.7 percent of GDP in 2020, due to a sustained increase in borrowing
requirements (Figure 91).
Figure 87: Total expenditure was in line with NEP Figure 88: Revenue outturn was higher amid a surge in
projection indirect taxes
Expenditure Outturn/Budget Target (2020) Revenue Outturn/Budget Target (2020)
2.5 2.24 1.6 1.42
1.4 1.19
2.0 1.64 1.08 1.03 1.06
1.2 0.97 0.94 0.93 0.92
1.5 1.28 1.0
1.10 1.10 1.10 1.02 1.00 0.96 0.99
0.8 0.55
1.0
0.6
0.5 0.4
0.2
0.0 0.0
Source: New Economic Program (2021-2023), MoTF and WB staff calculations.
Tax collections increased in 2020, as indirect taxes jumped later in the year. Direct tax collections as a
share of GDP dropped by 0.4 percentage points due to cyclical factors and tax relief measures. However, this
decline was fully offset by an increase in indirect taxes. The rapid recovery in domestic demand led taxes on
goods and services to rise by almost 40 percent (yoy) in real terms in 2020H2 (Figure 92). Taxes on vehicle
purchases increased sharply on strong demand and a rate increase. A big jump in imports in line with strong
domestic demand led tax on international trade to rise by around 28 percent (yoy) in real terms in 2020H2.
Overall, indirect taxes as a share of GDP rose by 1.3 percentage points in 2020.
Table 2: COVID-19 measures introduced by the Ministry of Family, Labor and Social Services
Number of Amount of
individuals/ assistance
households reached provided (TL m)
Social Support Program Phase 1 2,111,254 2,111
Phase 2 2,316,010 2,316
Phase 3 2,003,582 2,004
Short Term Work Allowance Workers 3,756,584 27,666
Unpaid Leave Subsidy Workers 2,471,134 8,266
Unemployment Insurance Individuals 994,608 5,082
Normalization support Workers 3,183,435 3,206
Total 50,651
Source: Ministry of Family, Labor and Social Services, accessed on February 13, 2021.
27
Most COVID-19 related government support to households through employment measures continued
through late 2020 and into 2021. The measures to support workers on unpaid leave, provide a short-term
work allowance to furloughed workers, and prohibit the layoff of formal workers were first extended into
2021Q1. Short-term work allowance support and the prohibition of the layoff of workers were extended further
until mid-May of 2021. The number of workers receiving short-term work allowance had reached 3.76 million
by mid-February 2021 (Table 2). For around 2.7 million workers not eligible for short-term work allowance,
but who were put on unpaid leave, the government provided a smaller benefit. Unemployment benefit
payments were continued for people who lost their jobs prior to the COVID-19 pandemic. The government
also introduced a new support program 31 in August 2020, under which the government pays firms’ social
security premiums for workers who benefitted from the short-term work allowance, or who were sent on
unpaid leave. So far, 3.2 million workers have benefitted from the normalization support. While these measures
were still in place as of early 2021, the social support program (one-off support for vulnerable households) was
ended in early 2021. However, households can still receive application-based social assistances in the context
of pandemic social support program’s third phase.
Table 3: Fiscal operations (CG, percent of GDP)
2017 2018 2019 2020
Revenue 20.1 20.2 20.3 20.4
Direct Tax 5.6 6.2 6.0 5.6
Indirect Tax 11.5 10.4 9.6 10.9
Other 3.0 3.6 4.7 3.9
Expenditure 21.6 22.1 23.1 23.8
Comp. of Employees & Soc. Sec. Contr. 6.0 6.3 6.8 6.7
G&S Purchases 2.0 1.9 2.0 1.9
Current Transfers 8.6 8.6 9.3 9.9
Capital Transfers & Expenses 2.7 2.8 2.2 2.1
Interest 1.8 2.0 2.3 2.7
Other 0.4 0.6 0.6 0.6
Overall Balance -1.5 -1.9 -2.9 -3.4
Primary Balance 0.3 0.0 -0.6 -0.8
Memo items
GDP growth ( yoy) 7.5 3.0 0.9 1.8
Inflation ( yoy) 11.1 16.3 15.2 12.3
Sources: Haver Analytics, WB Staff estimates.
Government debt increased, and a growing share of the rising budget deficit was financed by external
borrowing. A larger budget deficit and a depreciating lira led to an increase in overall central government debt
from 31 to 36 percent of GDP (Figure 94). The deficit was increasingly financed through external debt,
especially as the cost of domestic borrowing jumped 10 percentage points (Figure 93). The share of gross
borrowing that was external had increased from 20 percent in 2020Q2 to 60 percent by the end of 2020. In
addition to central government debt, general government debt amounted to an extra 4.1 percent of GDP as of
the third quarter of 2020, while publicly-guaranteed debt amounted to an estimated additional 4.6 percent of
GDP at end 2020.
31 This program also provides similar incentives for the employment of people aged below 25 and above 50.
28
Figure 89: Rising fiscal imbalances… Figure 90: …driven largely by rising current
transfers…
CG fiscal balance and primary balance (12- Real growth in CG expenditure
month rolling, share of GDP) 15%
0.0% 10%
-0.5% 5%
-1.0% 0%
-1.5%
-5%
-2.0%
-2.5% -10%
-3.0% -15%
-3.5%
Interest Other
Capital Expenses G&S Purchases
CG Overall balance CG Primary balance Personnel Current Transfers
CG Total Expenses (real)
Figure 91: …and interest expenditures Figure 92: But an uptick in tax on G&S contained the
deterioration
CG Interest Expenses Real growth in CG Revenue
18% 2.8% 25%
17% 2.7% 15%
2.6%
16% 5%
2.5%
15% -5%
2.4%
14% -15%
2.3%
13% 2.2% -25%
12% 2.1%
Income Tax Property Tax
Domestic tax on G&S Tax on Int. Trade
Tax/Interest (%) Interest/GDP (%), ra Property Income interest
Other CG Total Revenue (real)
Figure 93: Deficit increasingly financed through Figure 94: Contributing an increase in central
external debt government debt
Budget financing: TL vs. FX borrowing (gross, 45% Central government debt (share of GDP)
% share)
90% 25% 40% 35.9%
35% 30.8%
70% 20% 30% 27.9%
25%
50% 15%
20%
30% 10% 15%
10%
10% 5% 5%
0%
Q1-18
Q2-18
Q3-18
Q4-18
Q1-19
Q2-19
Q3-19
Q4-19
Q1-20
Q2-20
Q3-20
Q4-20
FX Gross Borrowing
TL Gross Borrowing
Domestic Debt External Debt Total Debt
Cost of Domestic Borrowing (%),ra
Sources: Haver Analytics, MOTF, WB Staff estimates.
29
Exceptional regulation aimed at supporting domestic lending was withdrawn
Various exceptional measures aimed at boosting domestic lending and banking activity were repealed
or relaxed in the second half of 2020. The BRSA, which had imposed limits on domestic banks’ holdings of
Lira swaps with offshore banks in the aftermath of the 2018 currency shock had again tightened between
February and April, curtailing access to this market, although banks retain access to similar swaps in the
domestic market. These limits were relaxed between September and November 2020, especially for longer-
maturity swaps, enabling an expanded use of these offshore markets (Figure 95). Starting in August 2019, the
reserves required to be held at the Central Bank had been linked to a bank’s real lending growth rate, effectively
enabling banks with higher loan growth to maintain a lower reserve ratio. In August 2020, these ‘discounted’
reserve ratios were increased substantially (by 200bps, less for some longer lira maturity term deposits and more
for gold assets), for both lira and FX assets, and reserves held at the CBRT grew sharply in the second half of
the year (Figure 96). The BRSA’s new “Asset Ratio”, introduced in April 2020, had also been instrumental in
raising banks’ lending, holding of Treasury debt and CBRT swaps, by mandating a certain level of domestic
assets relative to deposits (Figure 97). As the policy framework was tightened, the Asset Ratio was first relaxed
by allowing the inclusion of domestic FX loans to other banks in the ratio, and then abolished all together by
the end of 2020.
Figure 95: Offshore swap restrictions were relaxed Figure 96: Required Reserves were increased
Volume of swaps against Lira with offshore Required Reserves held at CBRT
banks (US$bn) 100 60
3.0
RRs on non-TL liabilities (US$ bn)
RRs on TL liabilities (TL bn)
90 55
2.5
80 50
2.0 70 45
1.5 60 40
50 35
1.0
40 30
0.5 30 25
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Feb-20
Jul-20
Jan-20
Oct-20
Jan-21
Feb-21
0.0
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Dec-20
Feb-20
Jul-20
Oct-20
Feb-21
Jan-20
Jan-21
TL liabilities (LHS) FX liabities (RHS)
Figure 97: Asset Ratio repeal eased pressure on banks Figure 98: Real interest rates negative and lowest
to lend amongst EMs
Approximate "Asset Ratio" for deposit banking 13 Real policy interest rate (percentage)
sector (TL bn) 11
1.02
9
1.00 7
5
0.98
3
0.96 1
A.R. relaxed -1
0.94
-3
A.R. -5
0.92
A.R. introduced repealed
Jan-19
Mar-19
May-19
Jul-19
Sep-19
Nov-19
Jan-20
Mar-20
May-20
Jul-20
Sep-20
Nov-20
Jan-21
Mar-21
0.90
Mar-2020
Jun-2020
Dec-2020
Jul-2020
Feb-2020
Oct-2020
Jan-2020
Apr-2020
May-2020
Jan-2021
Aug-2020
Sep-2020
Nov-2020
Max Min Turkey
Sources: Haver Analytics, World Bank Staff estimates.
Note: Turkey compared to the range of 12 other EMs; real post rate is ‘ex post’ as inflation is discounted by realized year-on-year consumer price
inflation; using the average cost of Central Bank funding for Turkey does not materially alter the trend.
30
Supporting monetary policy tightening as vulnerabilities grew
Turkey saw one of the sharpest monetary expansions amongst EMs in 2020. The CBRT took steps to
ensure liquidity in the financial system and to help firms manage the COVID shock with targeted liquidity
facilities, lowering real policy interest rates into deeply negative territory (Figure 98). Both FX and local currency
denominated deposits in the banking system rose sharply. As the lira depreciated, FX-denominated deposits
grew (Figure 99). With the lockdown hitting consumption and savings rising, lira deposits rose strongly. Overall,
broad money supply growth had peaked at nearly 50 percent year-on-year by October (Figure 99). Also, the
Primary Dealer banks were enabled to sell the government domestic debt securities that they have bought from
the Unemployment Insurance Fund to the CBRT for a temporary period. This level of monetary expansion
stood out amongst all EMs aside from Argentina (Figure 100).
Figure 99: Both FX and TL deposits grew strongly Figure 100: Money supply among the fastest amongst
EMs
Growth of M2, by component (yoy, %) 50 Growth of broad money, M2 (yoy, %)
50%
40
40%
30% 30
20% 20
10%
10
0%
0
Dec-19
Jan-20
Feb-20
Mar-20
Apr-20
May-20
Jun-20
Jul-20
Aug-20
Sep-20
Oct-20
Nov-20
Jan-18
Mar-18
May-18
Jul-18
Sep-18
Nov-18
Jan-19
Mar-19
May-19
Jul-19
Sep-19
Nov-19
Jan-20
Mar-20
May-20
Jul-20
Sep-20
Nov-20
TL deposits FX deposits Currency M2
Turkey and range of 20 other EMs (excluding ARG)
Sources: CBRT, World Bank Staff estimates. Sources: Haver Analytics, World Bank Staff estimates.
Since August 2020, the Central Bank began tightening monetary policy, gaining pace after the
appointment of a new Governor in November. Initially, the Central Bank used its interest rate corridor,
imposing quantitative limits on different CBRT funding windows to raise the average funding rate from below
8 percent to about 10.5 percent by late September, with a policy rate rise of 200 basis points on September 24
providing space for further tightening. With these measures, the CBRT’s effective rate had risen to 13.5 percent
by the end of October. Shortly after the appointment of the new Governor of the Central Bank in early
November, the CBRT raised by 875 basis points to 19 percent (Figure 101).
The Central Bank completed this move by reverting to a simpler policy framework and increasing
information and forward guidance. In addition to changes in interest rate policy, the Central Bank increased
the clarity and information on its forward-looking policy and started publishing additional data of market
interest. In early August the window via which primary market dealers were able to access TL liquidity below
the policy rate was closed. In a significant change from recent practice, the CBRT also reverted to the use of a
sole policy rate in November, no longer actively managing liquidity using an ‘interest rate corridor’. The CBRT
also increased the level of forward guidance in Monetary Policy Committee Statements and began publishing
detailed high-frequency data on CBRT derivative transactions (Figure 102).
31
Figure 101: The Central Bank’s effective interest rate more than doubled between Sept-Dec 2020
Policy and headline interest rates
20%
18%
16%
14%
12%
Policy rate
10%
CBRT effective rate
8%
Swap rate
6%
06.Aug.20
12.Aug.20
18.Aug.20
24.Aug.20
28.Aug.20
03.Sep.20
09.Sep.20
15.Sep.20
21.Sep.20
25.Sep.20
04.Nov.20
10.Nov.20
16.Nov.20
20.Nov.20
26.Nov.20
02.Dec.20
08.Dec.20
14.Dec.20
18.Dec.20
24.Dec.20
30.Dec.20
02.Mar.21
08.Mar.21
12.Mar.21
18.Mar.21
31.Jul.20
01.Oct.20
07.Oct.20
13.Oct.20
19.Oct.20
23.Oct.20
29.Oct.20
05.Jan.21
11.Jan.21
15.Jan.21
21.Jan.21
27.Jan.21
02.Feb.21
08.Feb.21
12.Feb.21
18.Feb.21
24.Feb.21
Sources: CBRT, Haver Analytics.
Figure 102: Monetary and financial policy reversals began in August and gained pace towards end-2020
Timeline of monetary developments
Traditional repo auctions resumed
06.Jan CBRT publishes high-freq. data on
New CBRT Governor appointed
07.Aug Overnight borrowing limits halved
Late liquidity rate raised 150bps
11.Nov New CBRT Governor appointed
09.Dec Monetary & Ex. Rate Policy for
Reversion to single policy rate
13.Aug Primary dealer window closed
Policy rate raised by 200bps
15.Mar Policy rate raised by 200bps
FX res. req. raised 300bps
12.Jan swap & futures transactions
Policy rate raised 200bps
22.Jan Weekly data published on
Abolish res. req discount
17.Nov Policy rate raised 425bps
19.Feb Res. Req on TL deposits
2021 announced
30.Oct Interbank & overnight
for higher lending
windows suspended
CBRT derivative
transactions
raised
09.Mar ERP launched
05.Nov
23.Nov
27.Nov
04.Sep
10.Sep
16.Sep
22.Sep
28.Sep
03.Aug
19.Aug
25.Aug
31.Aug
02.Oct
08.Oct
14.Oct
20.Oct
26.Oct
03.Mar
19.Mar
03.Dec
15.Dec
21.Dec
25.Dec
31.Dec
18.Jan
28.Jan
03.Feb
09.Feb
15.Feb
25.Feb
Sources: CBRT, BRSA.
Monetary tightening reverted the negative to zero real yields seen in mid-2020 and commercial lending
rates moved up. Based on both the policy rate and widely used instruments such as the one-year Treasury
bond, yields adjusted for year-ahead expected inflation were negative during the middle of 2020. The real policy
rate by this measure is now a relatively strong seven percent (Figure 103). Real expected yields (not adjusted for
risk) on Treasury bonds moved into positive real territory in August and have risen to six percent so far in 2021
(Figure 104). TL lending rates, which had dropped below 10 percent by mid-2020, started to rise in August
2020. The rates increased even further with the policy shift in late November 2020. The weighted average
interest rates for commercial loans almost doubled to 19 percent at the end of 2020 (Figure 105). This led
corporates to decrease their borrowing in TL terms.
32
Figure 103: Policy rates moved out of negative real Figure 104: As did Treasury bond yields
territory
CBRT policy rate and expected inflation Treasury bond yield and expected inflation
20% 20%
15% 9% 15%
5% 6% 7% 3% 4% 4% 4% 6%
1% 1% 0% 2% 2% 2%
10% 1% 10%
5% 5%
0% 0% -1% -1% -1% -1% -2% 0% 0%
-1% -1% 0%
-5% -5%
Aug 20
Nov 20
Sep 20
Mar 20
Jun 20
Dec 20
Mar 21
Jul 20
Oct 20
Jan 21
Feb 21
Apr 20
May 20
Aug 20
Nov 20
Sep 20
Mar 20
Jun 20
Dec 20
Mar 21
Jul 20
Oct 20
Jan 21
Feb 21
Apr 20
May 20
Expected inflation Real interest rate Expected inflation Real yield
Source: Haver Analytics, Central Bank of Turkey and World Bank Staff Source: Haver Analytics, Central Bank of Turkey and World Bank Staff
estimates. estimates.
Notes: Based on end of period CBRT policy (one-week repo) rate and one- Notes: Based on average monthly one-year Treasury bond yields and one-
year ahead expected inflation from CBRT Survey of Expectations year ahead expected inflation from CBRT Survey of Expectations
Figure 105: TL commercial lending rates have Figure 106: Gross international reserves were built up
increased
Weighted Average Interest Rate for Cumulative change in gross intl reserves since
Commercial Loans (%) Nov 2020 (US$bn)
12
30
10
25
8
20
6
15
4
10
2
5
0
0
-2
01-Jan-21
08-Jan-21
15-Jan-21
22-Jan-21
29-Jan-21
05-Feb-21
12-Feb-21
19-Feb-21
26-Feb-21
06-Nov-20
13-Nov-20
20-Nov-20
27-Nov-20
04-Dec-20
11-Dec-20
18-Dec-20
25-Dec-20
05-Mar-21
12-Mar-21
Lira Euro U.S. Dollar
Source: CBRT.
Policy adjustments initially supported an improvement in central bank and corporate sector external
positions. The CBRT has been gradually rebuilding international reserves, which stood US$10 billion higher
by mid-February 2021 compared to their level as at end October 2020 (Figure 106). However, with capital
inflows to Turkey drying up after mid-February as rising inflation reduced real returns, and as investor sentiment
globally became less positive towards emerging markets, this progress in building reserves has stalled.
33
D. Millions of people left behind after a year of turmoil
Firms have been severely affected
The World Bank has conducted a comparable international business survey that shows the impact of
COVID-19 on businesses in Turkey relative to other countries. The World Bank, in coordination with the
Union of Chambers and Commodity Exchanges of Turkey, implemented the Business Pulse Survey (BPS) in
Turkey between mid-June and mid-July 2020, for which 1,508 companies were surveyed. At a similar time, the
BPS took place in more than 60 countries around the world which were experiencing the first wave of the
pandemic. This section compares the results for Turkey with a set of regional and emerging market comparator
countries 32.
Firms in Turkey were affected more severely, on average, than firms in other European and Central
Asian countries. In Turkey, 69 percent of them were fully open before and during the survey period. 19
percent of the surveyed firms were partially open, 10 percent of establishments were temporarily closed, and 1
percent of firms reported they were permanently closed. Averaged over firms in all sectors, firms reported a
drop in sales of 37 percent, greater than the average for EU and for non-EU ECA comparators, although below
the 52 percent drop in the selection of non-ECA comparator countries.
By contrast to comparator countries where accommodation experienced the largest decrease in sales,
in Turkey, the logistics sector reported being the hardest hit. In Turkey, food services and accommodation
also reported being severely affected, but transportation and storage reported a colossal 75 percent fall in sales
(Figure 107). The financial sector reported a large decrease in sales in Turkey as in other countries.
Figure 107: Logistics hardest hit in Turkey Figure 108: Turkish firms cut wages and provided
leave with pay instead of firing employees
Average decrease in sales (percent) Change in employment conditions (average
percent of employees)
90 35
80 30
70
60 25
50 20
40
30 15
20 10
10
0 5
0
Finance, real estate
Other services
Agri. fishing mining
Accommodation
Food services
ICT
Const. & utilities
Wholesale & retail
Manufacturing
Transpor. & storage
Wage cut
Hours cut
Leave w/o pay
Leave w/ pay
Turkey EU Non-EU ECA Others Turkey ECA-EU ECA-nonEU Others
Source: Business Pulse Survey, World Bank Group Source: Business Pulse Survey, World Bank Group
Female-headed firms were more likely to close during the pandemic in Turkey, being concentrated in
the accommodation sector. Female-headed firms 33 tend to be much smaller than male-headed ones; a far
more pronounced difference than seen in comparators. Female-headed firms experienced on average a similar
impact on sales and a similar share of firms received support. However, the share of firms that declared being
fully operational is significantly lower for female-headed firms in Turkey; a discrepancy not seen in comparators.
32 This section builds on the un-benchmarked analysis of Turkey’s PULSE survey in the August 2020 Turkey Economic Monitor.
Regional comparators are: Albania, Georgia, Uzbekistan, Tajikistan, Russia, Kyrgyzstan, Bulgaria, Croatia, Hungary, Poland, Romania,
and Italy; and other comparators are: Brazil, India, Pakistan, Nigeria and Morocco.
33 Defined as the gender of the owner, or the primary manager of the respondent firm.
34
Such a large difference is likely related to the fact that the share of female-headed firms is by far the highest in
the accommodation sector, with that share being much higher than in comparators. Moreover, Turkey has a
lower share of female-headed firms than comparators in the wholesale and retail trade sector and in the
agriculture, fishing and mining sectors, which were less hard hit by COVID-19.
Related to Turkey’s moratorium on layoffs, formal employment changed less, but workers were more
likely to be put on paid leave. In April 2020, the Turkish government introduced a rule that temporarily
prohibited redundancies. The BPS shows that on average, the number of employees did not decrease in any
sectors in Turkey except agriculture, fishing and mining where a large portion of labor is seasonal with
temporary contracts. By comparison, there was a considerable decrease in employment in comparator countries.
Turkey does not differ significantly from the other countries in the sample in terms of the percentage of
employees given unpaid leave. But firms in Turkey introduced paid leave more often, with the average firm
issuing paid leave to 20 percent of its employees, likely supported by use of the government’s short-term work
allowance (Figure 108).
Furthermore, Turkish firms stand out for implementing hour reductions. In Turkey, 11 percent of
employees received wage cuts, which is in the mid-range of comparator groups. However, Turkish firms were
far more likely to reduce employee’s working hours. On average, firms in Turkey reduced the workhours of 31
percent of their employees, approximately double the proportion in comparator groups. It seems that to cope
with the moratorium on redundancies, firms in Turkey responded by seeking to lower the costs of their
workforce during the COVID-19 shock by either putting employees on leave or else reducing their hours.
The share of firms in Turkey that received public support during the pandemic is in the mid-range of
comparators. The BPS collects information on whether a respondent firm received any type of public support
designed to mitigate the adverse effects of the pandemic. Turkey is in the mid-range, with mostly EU countries
registering a higher proportion of supported firms (Figure 109). However, this might also be influenced by the
timing of the onset of COVID-19, with some countries being impacted later than Turkey, and may not have
fully rolled out support by the time of the survey.
Figure 109: More than a third of firms received public support
Receipt of public support (percent of firms)
80
70
60
50
40
30
20
10
0
ITA
POL
PAK
BRA
KGZ
MAR
ROU
NGA
HUN
RUS
UZB
GEO
TJK
TUR
ALB
HRV
BGR
Source: Business Pulse Survey, World Bank Group
35
Job vulnerabilities have widened labor market disparities
The pandemic led to a large-scale exit from the labor market, which has both a short- and long-term
impact. By May 2020 the economy had lost 2.6 million jobs (9.2 percent of total employment) relative to the
level observed a year before. While labor markets staged a partial recovery over the year, employment remained
down by 3.9 percent in November 2020 compared to a year earlier. The contraction of employment in the non-
agriculture sector was 3.2 percent, whilst job losses were more evident in the agriculture sector with a 7.4
percent contraction compared to a year earlier. The bulk of job losses were in the service, where many jobs are
informal. Informal jobs continue to prevail amongst approximately 30 percent of the Turkish population,
particularly in the agricultural sector. The labor force fell sharply to 30.6 million people in June 2020, but then
reversed some of its losses and rose to 31.1 million in November thanks to the easing of restrictions in 2020H2.
However, the labor force remains below the pre-pandemic level (around 1 million below the level of end-2019),
with more than 50 percent of the losses borne by women. This translated to a drop in labor force participation
to below 50 percent, close to its 2013 level.
Pre-existing labor market vulnerabilities were exacerbated, with female, lower-income, informal and
unskilled workers the worst-affected. Initial labor market losses were unevenly distributed across the
population. Most of the employment contraction was felt by informal workers, the lower-skilled, and women
and youth34 – groups that have traditionally experienced more limited employability and job quality. The job
recovery over the latter part of 2020 thus benefitted much less those groups that were hit the hardest. For
instance, as of November 2020, employment among unskilled workers remained 9.5 percent lower than the
year before, whereas it had reverted to pre-pandemic levels for medium- and highly-skilled workers. A similar
unequal pattern in the employment recovery is observed for women, youth and informal workers (Figure 110
to Figure 113).
Similar disparities can be seen in exits from the labor force altogether. Youth (ages 15-24) have borne
the largest labor force participation losses, driven largely by women and semi-skilled workers. The rate of youth
not in employment, education or training (NEET) has continued to climb since the COVID-19 outbreak
(Figure 114). A total of 363 thousand young people left the labor force in the 12 months to November 2020,
adding to significant youth employment losses since 2018. The share of NEET young people has increased by
nearly 2 million people relative to a year ago, with the rate being considerably higher among females than males
(34 percent versus 20 percent). At all education levels, around 60 percent of NEET are female, and the bulk of
NEET youth have less than a high school education (Figure 115).
Figure 110: Informal workers… Figure 111: lower skilled…
(Year-on-year percentage change in employment by groups)
Formal Informal Unskilled (< High School)
10%
Medium (High/Vocational School)
5% High (Higher Education)
8.1%
0% 3.6% 3.2% 9%
2.1% 4.2%
3.4% 2.5% 2.9%
% change (y-o-y)
-5% -1.6%-1.9% 4%
% change (y-o-y)
-10% -6.1% -1%
-3.4%
-15% -6%
-12.9%
-7.4%
-20% -15.7% -9.5%
-11%
-25% -21.2%
-23.8% -16% -15.6%
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Jul-20
Jan-20
Feb-20
Oct-20
-30%
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Feb-20
Jul-20
Oct-20
Jan-20
34 The definition of ‘youth’ accords with the Turkish national definition of ages 15 to 24, except where stated.
36
Figure 112: female… Figure 113 and young workers saw greatest job losses
Male Female Young (Ages 15-24) Adult (Ages 25+)
4%
1.6%
0.6%
2%
-1.4% 1% -2.0%
0% -3.2%
% change (y-o-y)
% change (y-o-y)
-2.7% -4% -1.2%
-2% -7.2%-6.6%
-4% -2.2%
-9% -6.2%
-6% -7.9%-7.6% -8.4%
-5.2%
-8% -6.5% -14%
-10%
-19%
-12% -10.5%
-20.4%
-21.2%
-11.8% -24%
-14%
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Feb-20
Jul-20
Oct-20
Jan-20
Apr-20
May-20
Aug-20
Sep-20
Nov-20
Mar-20
Jun-20
Feb-20
Jul-20
Oct-20
Jan-20
Source: World Bank calculations using monthly reports from the Labor Force Survey.
Figure 114: Youth NEET rates very high for women Figure 115: across all levels of education
Youth NEET rate Youth NEET rate by education level and
gender (percentage of population in group)
40 40%
35 35%
30%
30 25%
25 20%
15%
20 10%
15 5%
0%
10
Male
Female
2014 2015 2016 2017 2018 2019 2020
(Nov)
Male Female Less than high school High school Higher education
Source: TURKSTAT.
Figure 116: Youth LFPR has fallen Figure 117: and is far below comparators for women
Youth labor force participation rate Labor force participation rate by gender, 2019-
90
2020
60 80
55 70
60
Percent (%)
50
50
45 40
40 30
35 20
30 10
25 0
20
15
2020 F
2020 M
2020 M
2020 A
2020 S
2020 O
2020 N
2014
2015
2016
2017
2018
2019
2020 J
2020 A
2020 J
2020 J
Female Male Female Male
Source: TURKSTAT. Source: World Development Indicators, World Bank Staff estimates
37
Female labor force participation continues to lag that of males. Labor force participation losses over the
period of November 2019 to November 2020 reversed nearly three years of progress. The labor force
participation rate (LFPR) decreased to 49.3 percent from 52.5 percent a year earlier, and females lag
considerably behind, particularly amongst youth (Figure 116). The LFPR remains highest among the most
highly educated, and lowest among the low- to mid-skilled. Female labor force participation, at 30.6 percent,
continues to be less than half that of males, at 68.4 percent. Female LFPR is considerably lower than comparable
regional averages such as Central Europe and the Balkans (45.2 percent) and Latin America and the Caribbean
(41.5 percent) and far below the level of high-income countries (Figure 117).
Females exiting the labor force continued to do so largely due to family roles. Exit from the labor force
was dominated by household responsibilities, accounting for 31 percent and driven by women; 46 percent of
women who leave the labor force cited this as the factor, compared to zero percent among men. As shown in
Box 2, the vast majority of women report increased responsibilities at home during COVID-19, while less than
half of men reported the same. This represents 9.8 million work-able women; were these women to work, this
would represent an increase of nearly 30 percent of the labor force. Discouragement and retirement showed an
increase relative to pre-COVID levels, notably among women.
The unemployed are also more likely to be women and first-time job seekers. Official unemployment
rates do not currently provide an accurate picture of labor market dynamics as there have been large-scale labor
force exits over the past year 35, while the redundancy moratorium means that people cannot be laid off.
Nevertheless, a large gender gap is clear from the data. Unemployment stands at 15 percent among women
compared to 12.2 percent among men, and 25 percent among youth. Younger and first-time job seekers tend
to lack the breadth of skills and networks needed to adapt or transfer between jobs and sectors. Data from the
formal sector demonstrates younger workers aged 15-24 years were disproportionately more likely to drop out
from the job search compared to workers aged 25 years or above (Figure 118). In addition, COVID-19 job
protections do not apply to workers in the informal sector who, prior to COVID-19, comprised over thirty
percent of the workforce. The lack of employment support to these workers accentuates labor market and
welfare segmentation which unemployment rates do not show.
Figure 118: More youth dropping out of labor force
Change in formal sector job applications by age
group
16%
13%14%
12%
9%
Annual percentage change in job
3% 4% 3%
applications (%)
2016 2017 2018 2019 2020
-16%
15-24 years 25 years or more -24%
Source: World Bank staff using Kariyer.net 36 data.
35 In March 2021, labor market statistics were updated to include supplementary indicators in addition to employment and
unemployment, in line with the 19th International Conference of Labour Statisticians (ICLS) Resolutions of the International Labour
Organization (ILO) and the relevant EU regulations. Data publication was also changed from three-month moving averages to
independent monthly estimates. These datasets were not available at the time of preparation, and currently available breakdowns with
the new datasets are not sufficient to complete the analyses presented in this section, which is why an analysis based on the previously
published datasets are presented in this section.
36 This database represents a significant formal sector job search engine in Turkey, used to monitor formal sector job vacancies by the
Central Bank of Turkey and other institutions. In 2019-2020, monthly job openings ranged between 80,000-100,000. The bulk of job
38
Box 2: Constraints to Women’s Economic Empowerment and Covid-19
Unpaid care work constrains women’s economic opportunities. Women in Turkey have traditionally
been the primary caregivers for children and elderly family members, and are responsible for most household
tasks. They do five times as much unpaid care work than men and cite household and family duties as their
main reason for not joining the labor force. This is exacerbated by an insufficient supply of quality childcare
affordable enough to justify joining the labor force when compared to expected earnings. During the
pandemic, care centers have limited their activities or closed altogether, and, although both men and women’s
unpaid work at home has increased, women have reported a far greater increase compared to men, raising
further their constraints on participating in the labor force according to a recent survey by UN Women 37.
Women also face constraints on entrepreneurship. Women in Turkey are less likely to be entrepreneurs than
men (22 percent of men and 9 percent of women are self-employed), partly because of social and cultural
constraints, but also because of barriers in access to finance, and skills gaps.
Figure 119: Women report more home duties due
to COVID-19
Reporting increased hours due to COVID (%)
Cleaning & maintaining dwelling
Cooking & serving meals
Shopping for households
Play, talking, reading w children
Caring for children
Teaching children
Assisting older, sick or elderly…
Collecting water, fuel
0 20 40 60 80
Women Men
Sources: Turkstat; Cinar, K. & Kose, T., 2018. The Determinants of Women’s Empowerment in Turkey: A Multilevel Analysis. South European
Society and Politics. Vol. 23.; World Bank, 2018. Turkey Country Gender Assessment; World Bank. 2015. Supply and Demand for Child Care
Services in Turkey. A Mixed Methods Study; Kalaylıoğlu, Y, et. al. 2020. The Economic and Social Impact of COVID-19 on Women and Men:
Rapid Gender Assessment of COVID-19 Implications in Turkey. UN Women Turkey Office.; Tuzun, Ipek & Takay, Bahar, 2017. Patterns of
female entrepreneurial activities in Turkey. Gender in Management: An International Journal, Vol. 32 Issue: 3; World Bank. 2009. Female Labor
Force Participation in Turkey, Trends, Determinants and Policy Framework. Report number 48508-TR. ; Cebeci, T. 2014. Performance of Female
Employers in Turkey.; Kizilaslan, N., and M. Karaomer.
The first year of COVID-19 has also deepened labor market challenges faced over the mid- to long-
term, including learning and skills needed for a changing economy. The long-term impact on the labor
market is both tied to the impact of the adult labor force, young new labor market entrants, and school-age
students at risk of learning losses. While Turkey’s overall human capital index (HCI) prior to COVID was on
par with most middle-income countries, challenges were already apparent. Learning losses are expected to lead
to HCI and growth losses 38, particularly for poor and low-income households, unless targeted measures are
taken to safeguard skills. On average, for example, one year of school closures in Turkey has been estimated
leading to a 40-point decrease in reading and mathematics PISA scores, which could reverse recent gains and
openings tend to be found in manufacturing (30 percent), followed by wholesale and retail trade (11 percent), information and
communication technology (11 percent) and the remainder in formal sector services such as accommodation, finance, real estate,
administrative, and health services (roughly 5-8 percent per sector).
37 UN Women, with the support of SAM Research and Consulting, conducted a survey of 1,508 men and women between 19-25
April, 2020. The survey was conducted on the basis of Computer-Assisted Telephone interviewing. It was based on a multi-stage
random sample frame, evenly split between men and women, and was statistically-significantly representative in terms of socio-
economic development and regions.
38 World Bank (2020). Human Capital Project Report, Turkey Snapshot. Washington DC: World Bank.
39
harm future productivity 39. The impact is expected to be worse for low-income households and non-native
Turkish speakers. Turkey’s share of NEET, at 26 percent, is higher than other countries with a comparable
HCI. Female labor force participation, at 34 percent, is also far lower than other countries that have a similar
HCI amongst females (Figure 117). Given these risks to learning, the pandemic has potential long-term scarring
effects on the labor market unless adequately addressed today.
Figure 120: More youth are NEET Figure 121: And far fewer women in the labor force in
Turkey
Source: World Bank staff using Turkstat data and (for female labor force comparisons) and World Development Indicators data.
Note: Data for NEET and LFPR relates to 2019, HCI 2020. Share of youth not in education, employment or training (NEET) is the proportion of young
people who are not in education, employment, or training to the population of the corresponding age group. The age group can be youth (ages 15 to 24);
persons ages 15 to 29; or both age groups, depending on national definitions.
Elevated inflation hit lower income households
Considering the persistent high inflation in Turkey, it is important to understand how inflation affects
low-income households. If not offset by proportional increases in income, inflation can diminish the
purchasing power of individuals, undermining their ability to maintain the same level of consumption. Inflation,
while conventionally presented in average terms, can in fact affect different households differentially. Inflation
can be regressive if prices rise relatively more for items that are consumed more by households from the bottom
of the income distribution. The regressive nature of price changes can be more pronounced in periods of high
inflation as has been the case in Turkey since 2017, as inflation has remained in double digits. In particular, the
period following the exchange rate instability of the summer of 2018 witnessed the highest episode of inflation
since the early 2000s. In this context, it is important to examine the nature of inflation in Turkey over recent
years.
Varying consumption patterns across income groups in Turkey suggest that households may indeed
by affected differently by inflation. Figure 122 shows how expenditure groups have varying shares in the
overall consumption across expenditure deciles. People from the bottom of the distribution would be hit the
hardest when prices of food and housing rise, while better-off households will experience relatively higher
inflation when prices of transportation increase, for instance.
39World Bank (2020). Turkey Safe Schooling and Distance Education Project Appraisal Document, Annex 4 COVID-19 Learning
Loss Assessment. Report No: PAD3962. Washington DC: World Bank.
40
Figure 122: Consumption patterns vary greatly across income groups
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 2 3 4 5 6 7 8 9 10
Consumption deciles
Food and non-alcoholic beverages Clothing and footwear Housing, water, electricity, etc.
Furnishings, household equipment, etc. Transport Hotels, cafes and restaurants
Miscellaneous goods and services Other Expenditure Groups
Source: Household Budget Survey 2018, Turkstat; World Bank Staff estimates
Analysis using highly disaggregated consumption and price data shows that the high inflation episode
following the 2018 currency shock hurt the poor significantly more. The Turkish Household Income
Consumption Expenditure Survey (HICES) and price data from the Turkish Statistical Institute (TURKSTAT)
can be used to gauge the distributional effects of inflation. Decile-specific inflation rates were computed using
household-level consumption baskets and disaggregated price information for 12 expenditure groups 40. Data
for 2019 reveals that households from the bottom of the distribution bore higher price rises 41. While overall
CPI 42 increased by 15.2 percent, the level of inflation experienced by households from the lowest decile reached
16 percent, almost two percentage points above the inflation borne by the top decile (Figure 123). A similar
gap occurred between the Bottom 40 and Top 60 inflation rates, at 15.7 percent and 14.9 percent, respectively. 43
While most expenditure groups saw significant price increases, inflation was higher for expenditures on food
and non-alcoholic beverages (19.5 percent) than for transportation (9.8 percent).
Figure 123: Lower income households experienced Figure 124: High inflation depressed incomes of the
higher inflation poor
Esimated annual inflation in 2019 Real Income for first decile and average
(2011=100)
16.5 150
16.0
140
16.0 15.7
15.5 15.5
15.2 130
15.5 15.2
15.0 14.9 120
15.0 14.7 110
14.5 14.3 100
90
14.0 2011 2012 2013 2014 2015 2016 2017 2018 2019
1 2 3 4 5 6 7 8 9 10
Decile Overall 1st Average
Source: World Bank calculations using data from Turkstat.
40 This analysis diverges from the official methodology used to calculate consumer price inflation, which uses prices in the current year
and consumption weights from the previous year. The rationale for this is that, given significant changes to consumption patterns,
applying current weightings can reveal the sensitivity of results to this assumption.
41 A similar inflation incidence analysis could not be carried for 2020, since household-level expenditure data was not available.
However, an examination of the actual changes in prices across expenditure groups for 2020 suggests that the effects of inflation were
borne more evenly across all income groups.
42 On a 12-month average basis.
43 The differences between the bottom, middle and top of the distribution are statistically significant at the 95% confidence interval.
41
Over 200,000 additional people could be counted as poor due to the regressive effects of inflation in
2019. The poverty line is updated for each year using the overall change in the CPI. This methodology can
underestimate poverty by a non-trivial amount if the bottom of the distribution experiences higher inflation.
Adjusting the threshold for 2019 using decile-specific inflation suggests that 209,000 additional individuals
could be deemed poor (an increase in the poverty rate of 2.2 percent) compared to the standard calculations
that rely on the average CPI to adjust for annual changes in prices. In 2015, a year when the price gap between
low-income and the overall CPI was even larger, the number of poor was underestimated by 342,000 individuals
(4.2 percent). In recent years, the inflation elasticity of poverty has been close to 1 – in other words – an increase
in prices not offset by income leads to a proportionally similar increase in the poverty rate, as can be seen in
Figure 124.
And millions more are now in poverty
A key feature of Turkey’s growth over the past two decades has been inclusiveness, but this is now
under threat. The incidence of poverty44 fell by 77 percent between 2003 and 2018, from 37 to 8.5 percent.
But this impressive poverty reduction was set back by the 2018 downturn, which caused a loss of over a million
jobs, created significant upward pressures on prices and raised poverty to 10.2 percent in 2019. This translated
into nearly 1.5 million additional poor, a total of 8.4 million nationally, erasing almost all gains achieved in the
three years preceding the economic turmoil (Figure 125). In fact, the social and economic burden of the 2018
economic turmoil fell mostly on lower income households. While consumption shrank by 0.7 percent between
2018 and 2019 for the median Turkish family, households from the bottom 20 percent experienced a much
worse contraction that ranged from 5.5 to 14.6 percent.
Figure 125: The poverty rate is estimated to reach its highest level since 2012
Absolute poverty rate and number of poor (consumption-based poverty)
40% 30,000,000
35%
37.0% 25,000,000
30%
Number of poor
20,000,000
Poverty rate
25%
10,170,932
27.0%
24.6%
8,419,849
20% 15,000,000
6,939,264
21.9%
15% 19.1%
10,000,000
15.7% 15.7% 15.2%
10% 13.1% 12.3% 12.2%
10.7% 10.0% 10.7%
9.5% 8.7% 8.5% 10.2%
5,000,000
5%
0% 0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Source: World Bank calculations using data from Turkstat
Note: Poverty is measured as the proportion of people with per capita consumption of below $5.5 a day 2011 PPP. Source: World Bank calculations
using HBS 2003-2019.
Following shortly after the last shock, the impact of COVID-19 is expected to increase poverty even
further. Simulation analysis of the effects of Covid-19 on household incomes (as opposed to modeling the
impacts on consumption which require further assumptions) 45 reveals the scale of the potential poverty effects
44 Measured as the share of households with consumption levels below the World Bank’s poverty line for Upper-Middle Income
Countries ($5.50 per person per day in constant 2011 purchasing power parity prices)
45 The simulations were performed using income-based, rather than consumption-based poverty measures since the former approach
does not require making assumptions about the magnitude of the marginal propensity of households to consume across the
distribution, which is a parameter that likely changes during crises, as well as about the ability to smooth consumption through the use
of savings and other buffers. Moreover, poverty trends and levels in Turkey are highly comparable when using income or
consumption as the measure of household welfare.
42
of this shock. 46 The poverty rate may have increased by as much as 2.1 percentage points by the end of 2020,
equivalent to pushing another 1.6 million people below the poverty line (Figure 126). As a result of the two
back-to-back shocks, Turkey could have entered 2021 with over 3 million new poor, totaling 10 million
nationally, an increase of 40 percent compared to the number of poor in 2018. The high effects of Covid-19
on poverty in spite of overall positive growth are explained by the stark inequalities in the burden of the
employment contraction. Consequently, the bottom 40 accounted for 60 percent of the job cuts. In contrast,
most individuals from the upper half of the distribution kept their jobs, while some groups (for instance, those
from the top decile) even enjoyed net job gains (Figure 129). Had the government not acted swiftly, simulations
show that the impact on poverty could have been even worse – almost three times as much. In this respect, the
social emergency package – introduced shortly after the first round of lockdowns – which included social
transfers, unemployment insurance benefits and unpaid leave subsidies – was important in avoiding even worse
effects of the COVID-19 pandemic.
Figure 126: Poverty can increase by up to 1.6 pp in 2020, corresponding to 1.6 million new poor
Simulated impacts on income-based poverty from a six-month income shock
20 17.4
16 13.6 13.4 13.2 13.2
11.1
Percentage
12
8
4 2.1
0
Counterfactual in Drop in income Short-term Emergency Social Emergency Social Emergency Social Net change
2020 due to all potential allowance + Assistance (Phase Assistance (Phase Assistance (Phase
job losses Suppport to 1) 2) 3)
unpaid leave
No Shock Shock Shock + Policy Responses Covid-19 impact
Source: World Bank calculations using HBS-2019 and monthly reports from the Labor Force Survey
Note: Poverty is measured as the proportion of people with per capita consumption of below $5.5 a day 2011 PPP. Source: World Bank
calculations using HBS 2003-2019.
The economic turmoil translated into downward economic mobility for other income groups and
contributed to keeping inequality high. One of the most remarkable accomplishments of the broad-based
process of economic progress in Turkey has been the growth of the middle class, more than doubling since the
early 2000s and making it the largest socioeconomic group in the society. Yet, the shocks of the past three years
stalled that expansion, shrinking the middle class by 4 percentage points (77 percent of that explained by the
COVID-19 crisis alone), namely 3.3 million individuals who passed to a lower income group. At the same time,
many non-poor, but vulnerable households (comprising 26.7 percent of the population prior to COVID-19)
have likely exhausted their buffers and could be one shock away from sliding into poverty (Figure 128).
Following a trend set in earlier years, the uneven burden of the 2018/19 economic turmoil and the 2020
pandemic on low-income households has furthered income and consumption inequality. The Gini coefficient
in 2019 reached 0.42 (up from 0.41 in 2017), and is expected to stay high in 2020. Other indicators also point
to a worsening of inequality.
46 Poverty in Turkey is measured using the Household Budget Survey (HBS), compiled annually by TUIK. Due to the mobility
restrictions imposed to curb the spread of Covid-19, HBS was not collected in 2020. This, in addition to long lags between data collection
and the public release of the survey data, makes it necessary to rely on micro-simulation, or imputation methods to estimate the effects
of Covid-19 on poverty in 2020.
43
Box 3: The Effects of Covid-19 on Global Poverty
Covid-19 and its associated health and economic crisis have affected every corner of the globe, partly
reversing the steady gains in economic progress and poverty reduction achieved over the past two decades
in most developing and emerging economies. The latest World Bank projections suggest that the Covid-19
pandemic will result in a substantial increase in poverty in much of the world. Based on the January 2021
forecast from the Global Economic Prospects, it is estimated that up to 124 million people globally would
be pushed into extreme poverty in 2020 (considering the $1.90 per person per day in constant 2011
purchasing power parity poverty line), with South Asia and Sub-Saharan Africa each contributing roughly
two-fifths of the increase. At the $5.5 poverty line for Upper-Middle Income Countries, the relevant
yardstick for Turkey, global poverty is projected to increase by 210 million people, with more than half of
the new poor in South Asia, with Europe and Central Asia accounting for a small share (2.4 percent,
equivalent to over 5 million extra poor).
Figure 127 shows the annual change in the number of global extreme poor from 1992 to 2020. The increase
in poverty caused by Covid-19 is not only significant in size, but also unprecedented in the last two decades
to 2020. Before Covid-19, the only setback in global poverty reduction had occurred during the Asian
financial crisis of 1997 and 1998, which increased extreme poverty by 18 million in 1997 and by a further 47
million in 1998, with about half of the increase in absolute numbers due to Covid-19. The increase in poverty
may not be short-lived. While the numbers for 2021 remain highly uncertain and contingent upon the
evolution of the pandemic and the vaccine roll out, nowcasting of poverty up to 2021 shows that the
estimated Covid-19 induced poor is set to rise to between 143 and 163 million in 2021.
Figure 127: Covid-19 may increase poverty globally by up 124 million in 2020
Source: Lakner et al. (2020) (updated), PovcalNet, Global Economic Prospects. Note: Projections for years 2018-2021 are based on updated
estimates of Lakner et al. (2020). For 2020, we show both (a) the number of people that were expected to move out of extreme poverty had the
COVID-19 pandemic not happened (Pre-COVID-19 counterfactual scenario, gray bar) and (b) the number of people who are pushed into
poverty under the COVID-19-baseline scenario (blue bar) or the COVID-19-downside scenario (blue + orange bar). The “new poor” induced
by COVID-19 is the sum of (a) and (b). For instance, under the COVID-19-baseline scenario in 2020, the “new poor” is equal to 88 + 31 = 119
million. Similarly, under the COVID-19-downside scenario in 2020, the “new poor” is equal to 88 + 31 + 5 = 124 million. We thank colleagues
at USAID for the inspiration behind this figure.
44
Figure 128: Job losses at the peak of the COVID-19 Figure 129: More than 3 million people dropped out of
crisis affected chiefly the bottom deciles the middle-class since the 2018 currency shock
500,000 20%
Job losses Percentage 70 Changes in income groups 63.2 3.0
60.1
15.5%
400,000 16.5% 60 2.1 2.0
15%
Cahnge in percentage points
13.8%
12.2% 50 1.1 1.0
Percentage of total job losses
300,000 12.3%
11.4%
Percentage
10% 40 0.0
Net job osses
200,000 9.8%
7.5% 30 25.6 26.7 -1.0
100,000
4.9% 5% 20 -2.0
11.1 13.2
0 -3.1
10 -3.0
1 2 3 4 5 6 7 8 9 10
0%
-100,000 0 -4.0
Poverty rate Vulnerability rate Middle class rate
-3.9%
-200,000 Income Deciles -5% Counterfactual Covid-19 scenario Net change
Note: Net change in employment between April 2019 and April 2020. Note: “Poor” denotes the share of people with per capita consumption of
Negative numbers correspond to net job gains. Source: World Bank below $5.5 per capita a day in 2011 PPP, “Vulnerable” denotes the share of
calculations using TurkStat data. people with per capita income of above $5.5 and below $11 per capita a day
2011 PPP and “Middle Class” denotes the share of people with per capita
income of above $11 per capita a day 2011 PPP. The change in the share of
the income group in the population (orange circle) is measured on the right
vertical axis. Source: World Bank calculations.
45
II. LOOKING AHEAD
Coming after such a challenging year, Turkey is expected to see robust growth in 2021. However,
the outlook is beset with uncertainty, and will require prudently tight monetary policy, supportive
fiscal and financial sector policies, and control over COVID-19. Turkey has made good progress
in vaccination roll-out, but the road ahead is long and new COVID outbreaks may undermine
recovery in 2021. While the policy framework is presently geared towards stabilization, the
replacement of the Central Bank Governor has unsettled markets and raised questions over its
future direction. The baseline assumption of this report is that monetary policy will remain suitably
responsive throughout 2021 in order to tame inflation and boost investor confidence. But the report
also recognizes that there is greater uncertainty around this baseline assumption. There remains
space for accommodative fiscal policy to facilitate a full recovery in 2021, although increasing
contingent liabilities and high exposure to external market risk suggest that fiscal expansion should
be moderate and time limited. Strengthening insolvency and debt resolution frameworks will be
important to de-risk the corporate and financial sectors. Turkey can raise potential growth by
putting in place measures to incentivize a decarbonization transformation early, which would also
be supported by the development of longer-term and sustainable finance sources with diversified
financial markets. Reducing trade barriers and increasing market access for Turkey-based firms
will also help support growth in the medium-term. Finally, seismic shocks to labor markets are
likely to have long-lasting effects, and active and targeted social and labor market policies will be
increasingly important to ensure people – especially women and the youth of today – do not miss
the opportunity to fulfil their potential as productive members of society.
A. The outlook for growth is tentative in a risky environment
Respectable projected growth in 2021 reflects the gradually waning effects of COVID-19 and a low
GDP base in 2020. The economy is projected to expand by 5 percent in 2021, followed by 4.5 percent growth 47
in the medium term (Figure 130). Effectively all growth achieved in 2021 is driven by the unusually low base
of activity in 2020, and a more typical level of economic activity this year will yield high year-on-year growth.
Growth is expected to be weak in the first half due to the impact of restrictions, relatively weak external demand,
and a tight monetary stance, whilst GDP in the second quarter will be far higher than the previous year when
the COVID-19 pandemic brought the economy to a near-standstill. These projections assume that tentative
economic recovery is not undermined by the need to implement strict COVID-19 control measures for
extended periods to deal with new outbreaks in Turkey or major export markets.
Growth in 2021 is expected to be driven largely by a rebound in exports. The composition of growth is
projected to shift towards external demand as global activity picks up. Exports are projected to account for
more than two-thirds of the growth in 2021. This will be driven mostly by goods exports, but an expected
partial recovery in tourism will also contribute. The current account deficit is expected to fall to 3.7 percent of
GDP in 2021, with a recovery in tourism, exports and recovery in gold trade balance. Private consumption and
investment growth will decelerate from the peak reached in late 2020 and make modest contributions to growth.
Public consumption is projected to increase by 2.4 percent in 2021.
Weak labor markets and high costs of borrowing will preclude strong private consumption growth in
the near term. Inflation remains relatively high, yet large numbers of workers have yet to recover to their pre-
COVID income level, whereby the scope for consumption growth is limited. Relaxation of restrictions is likely
to support moderate growth in services consumption. And yet the high cost of borrowing and frontloaded
purchases of durable goods in 2020H2 are expected to constrain non-services consumption demand,
47 All GDP growth references refer to GDP in real (constant price) terms, unless stated.
46
particularly in 2021H1. Uncertainty regarding the continuance of the government’s employment support
measures beyond 2021H1 constitutes a downside risk for private consumption.
Figure 130: Exports will be the main driver of growth Figure 131: Debt overhang tends to be a drag on
in 2021 investment.
10% Contribution to Growth Corporate Leverage and Investment Growth
8%
6% 74 25
4% 72 20
2% 70 15
0% 68 10
-2% 66 5
64 0
-4%
62 -5
-6% 60 -10
-8% 58 -15
2019
2020
2021
2022
2023
Stocks Imports Corporate Leverage (Debt/Assets)
Exports Investment Total Investment (real, % change), ra
Government Spending Private Consumption
Figure 132: Rebound in EMDE investment Figure 133: Weak investment can cause long-lasting
insufficient to reverse losses during 2020. damage to economic potential.
Percent
18 Investment Growth Potential Output Growth Decomposition
7 7%
6 3.6% 6%
12 4.2%
5 5%
6 4 4%
3 3%
0 2 2%
Advanced
economies 1 1%
-6 EMDEs 0 0%
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
EMDEs excl.
-12 China Potential Capital Potential Employment
2000 2005 2010 2015 2020 Potential TFP Potential Growth, ra
Sources: Haver Analytics, TURKSTAT, CBRT Sectoral Accounts, Global Economic Prospects January 2021, WB Staff estimates.
Efforts to lower inflation will be set back by heightened uncertainty and exchange rate depreciation,
highlighting the importance of continued commitment to prudent monetary policy. Amid uncertainty
over the future direction of monetary policy, market-priced inflation expectations immediately rose from 14.4
to 18.4 percent (Figure 142). Sharp depreciation, high inflation expectations and rising international food and
energy prices along with a surge in production costs associated with the pandemic are likely to add pressure on
prices. Inflation is projected to accelerate to 15.5 percent on average in 2021. Avoiding premature monetary
easing is important to contain pressure on the Lira and stabilize inflation expectations, particularly in 2021.
Tight credit conditions and debt overhang will be a drag on investment in 2021. Low-cost credit
availability has often been associated with bursts of growth in consumption and investment in Turkey over
recent years. Yet the CBRT’s forward guidance on monetary policy this year, and the need for the banking
sector to restore its balance sheets following last year’s credit boom, means that there is no room for such
47
policy. 48 Corporate leverage increased in 2020 to contain the adverse impacts of the pandemic, and is likely to
pressure investment (Figure 134). The outlook for external inflows is uncertain, especially given the possibility
of a tightening of liquidity in emerging markets as advanced economies reflate, and of elevated policy
uncertainty in Turkey. On balance, investment is not expected to see much growth in 2020, except for some
recovery in the fourth quarter, supported by public investment. The pace of investment growth will be
important for potential economic growth going forward (Figure 136).
Box 4: Global Outlook
As 2020 has so clearly and painfully evidenced, epidemiological developments can have a huge
impact on the global economy. While there is reason to expect that COVID-19 will increasingly be
controlled and immunized against over the course of this year, economic projections are sensitive to setbacks,
such as the development of more virulent or fatal strains or delays in the roll-out of major vaccination
programs.
Global GDP is expected to expand 4 percent in 2021, predicated on good pandemic management and
effective vaccination in many countries, continued monetary policy accommodation and diminishing fiscal
support. After this year’s pickup, global growth is envisioned to moderate in 2022 to 3.8 percent. Advanced
economies are projected to recover, with growth reaching 3.3 percent and 3.5 percent in 2021 and 2022,
respectively. Aggregate EMDE growth is envisioned firming to 5 percent in 2021 and moderating to 4.2
percent in 2022, although much of this recovery is due to China’s rebound. Excluding China, the recovery
across EMDEs is anticipated to be far more muted, averaging at 3.5 percent in 2021-22, as the pandemic’s
lingering effects continue to weigh on consumption and investment.
Figure 134: Global growth to rebound in 2021 Figure 135: But still be much lower than the no-
COVID baseline
GDP Growth (in percent) GDP loss against no-COVID-19 baseline
10
8 0.0
6 -1.0
4 -2.0
2 -3.0
0 -4.0
-2
-5.0
-4
-6.0
-6
-7.0
World EMDEs AEs
World Advanced Economies 2020 2021 2022
EMDEs Turkey
Source: World Bank, Global Economic Prospects, January 2021.
The level of global GDP in 2021 is forecast to be 5.3 percent below pre-pandemic projections — or
about $4.7 trillion, and by 2022, still 4.4 percent below pre-pandemic projections. The gap in EMDEs is
nearly twice as large as in advanced economies as output remains dampened by lingering risk aversion on the
demand side and the effects of diminished physical and human capital accumulation on labor productivity.
48 Investment plunged particularly sharply in EMDEs excluding China as the pandemic took hold. The collapse in investment was much
sharper in large EMDEs (excluding China) than in large advanced economies. Turkey was exceptional in recording positive growth in
investment during the pandemic, supported by strong credit stimulus. The speed of investment will vary among the EMDE group, but
is expected to be weak overall (Figure 133).
48
Most commodity prices rebounded in the
second half of last year, but the pickup in oil prices Figure 136: COVID-19 estimated to have damaged
potential growth
lagged the broader recovery in commodity prices
due to the prolonged impact of the pandemic on Estimated impact of COVID-19 on global
global oil demand. Oil demand fell 9 percent last potential growth
3.5
year—the steepest one-year decline on record—
and prices average $41/bbl – 34 percent down on 3.0
2019. Oil prices are forecast to remain close to 2.5
current levels and average at $44/bbl in 2021 before
2.0
rising to $50/bbl in 2022.
Pre-COVID
1.5
COVID
Even if the pandemic is brought under control, 1.0
its effect on potential growth could be longer
0.5
lasting than expected. Debt burdens have
increased for both corporates and governments. 0.0
2000-09 2010-19 2020-29
This follows a decade in which global debt had
already risen to a record high of 230 percent of Source: World Bank, Global Economic Prospects, January 2021.
GDP by 2019. High debt levels leave borrowers vulnerable to a sudden change in investor risk appetite. This
is especially true for riskier borrowers and EMDEs dependent on capital inflows to finance large fiscal and
external current account deficits. Banks’ capital buffers are also under pressure due to falling profitability and
asset quality deterioration. Defaults have already surged in the hardest-hit sectors and countries, and rising
credit downgrades point to further strains in the future. These developments reduce the resilience of financial
systems, particularly in countries with weaker banking systems, or without the policy space to provide
sufficient support to stressed financial institutions.
Figure 137: Government debt is on the rise Figure 138: Including in EMDEs
Debt in EMDEs
Government debt
140 180
120
100
120
80
60
40
60
20
0
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
0
1970 1980 1990 2000 2010 2020
World Advanced economies EMDEs
Government Private Total
Source: World Bank, Global Economic Prospects, January 2021.
Even once the pandemic has subsided, the global economic landscape is unlikely to return to its
previous state. The pandemic will leave lasting scars on productivity, including through its effect on the
accumulation of physical and human capital, which will exacerbate the downward trend in potential growth.
Source: World Bank, Global Economic Prospects, January 2021.
There is uncertainty over the baseline forecast, with both upside and downside risks On the downside,
recent normalization could trigger a third wave of infections in Turkey that could necessitate a new lockdown
and further dent domestic demand and production capacity, exacerbate uncertainty, and erode policy space. A
similar flaring up of COVID-19 in important partner markets for Turkey, such as the EU, would also lower
growth for Turkey via reduced exports and a possible new wave of supply chain disruptions. Premature
49
monetary easing, further loss of investor confidence and rise in macro financial vulnerabilities would also
undermine growth and delay orderly recovery. On the upside, the recovery in the global economy could be
faster than projected with effective and orderly vaccination developments. Turkey’s experience to date provides
important lessons on responding quickly and in a targeted and timely manner, which should help limit the
likelihood of a downside scenario.
B. There are major uncertainties which could materially affect the path of growth this
year
COVID-19 is a clear and present danger until the vaccination program nears completion
COVID-19 remains a risk while vaccinations do not provide full coverage. While it is likely that the first
stage of vaccinations can be completed in the second quarter of 2021, there is a wider range of uncertainty over
the timing of completion of future phases. Extrapolating from the vaccine rollout beyond the first six weeks in
the UK, USA and Turkey, a range of scenarios suggest vaccination of health workers and the over 65s could
be completed between mid-March and mid-April 2021. The second phase (see Box 5) targets a larger group of
people – close to 20 million – of over 50s, public sector personnel and selected other groups. Depending on
the cumulative progress in vaccinations, as well as the availability of vaccine supply, completing the vaccinations
of this group might feasibly be achieved as early as late May, or else take up to September (Figure 139).
Figure 139: There remains a long way to reaching near-full vaccination rates
Vaccination roll out scenarios (millions of vaccinations)
140
120
100
Phase 3
80
60
40 Phase 2
20
Phase 1
0
6-Apr-21
20-Jul-21
4-Jan-22
12-Jul-22
12-Jan-21
23-Feb-21
12-Oct-21
15-Feb-22
27-Apr-21
18-May-21
2-Nov-21
25-Jan-22
19-Apr-22
10-May-22
31-May-22
2-Aug-22
8-Jun-21
10-Aug-21
31-Aug-21
21-Sep-21
23-Nov-21
8-Mar-22
2-Feb-21
16-Mar-21
29-Jun-21
14-Dec-21
29-Mar-22
21-Jun-22
Actual Med High
Source: Ourworldindata.org, World Bank Staff estimates.
Note: Numbers of vaccinations based on population estimates of targeted groups, two vaccines per person, and 100 percent coverage in group.
Box 5: Turkey’s vaccination program to date
Turkey’s phased vaccination program has begun. Turkey’s timely purchasing and phased-in supply of
the Sinovac’s CoronaVac (100 million doses) 49 and Pfizer-BioNTech (4.5 million doses) 50 COVID-19
vaccines enabled a nationwide vaccination campaign to start in early January 2021. The first phase targets
were the healthcare workforce, people above the age of 65 years, individuals with chronic illnesses, and front-
line workers in critical sectors (See Table 3). By March 20, a total of 13 million doses had been administered,
equivalent to around 15 doses per 100 people, with 5 million people having received their second dose
(Figure 140). The speed of Turkey’s vaccination program compares favorably with other large countries.
Turkey’s vaccination program has proceeded at approximately the same speed as the UK and USA programs
49 https://tr.euronews.com/2020/11/26/turkiye-covid-19-a-kars-50-milyon-doz-sinovac-as-s-icin-cin-ile-anlast
50 https://twitter.com/drfahrettinkoca/status/1359542068654317570
50
(Figure 141), and at end-February, Turkey has the fourth highest vaccination rate amongst countries with
populations greater than 10 million, and is the best performing large developing country worldwide.
Figure 140: Vaccinations have proceeded quickly Figure 141: In line with some of the fastest
worldwide
Vaccinations (millions of people) Total vaccinations per 100 people since first
9 50
vaccination
8 45
7 40
6 35
30
5 25
4 20
3 15
2 10
1 5
0 0
10-Feb
14-Feb
18-Feb
22-Feb
26-Feb
13-Jan
17-Jan
21-Jan
25-Jan
29-Jan
2-Mar
6-Mar
10-Mar
14-Mar
18-Mar
2-Feb
6-Feb
10
20
30
40
50
60
70
80
90
100
-
UK United States Chile
First dose Fully vaccinated Morocco Greece Poland
Romania Portugal Turkey
Source: Ourworldindata.org
Table 4: Population targeting for Turkey’s vaccination program
Stage Group Rank Subgroup
1 A – Healthcare Workforce A
B – Elderly people in nursing homes and B
people working in nursing homes
C - People 65 years of age and older C1 People 90 years of age and older
C2 People of the age group between 85-89
C3 People of the age group between 80-84
C4 People of the age group between 75-79
C5 People of the age group between 70-74
C6 People of the age group between 65-69
2 A – Priority Sectors for Service Delivery A1 Ministry of Defense
A2 Ministry of Interior
A3 People with Critical Jobs
A4 Police and Private Security Staff
A5 Ministry of Justice
A6 People in prison
A7 Education Sector (teachers and instructors)
A8 People working in the food sector
A9 People working in the transportation sector
B - People of the age group 50-64 B1 People of the age group between 60-64
B2 People of the age group between 55-59
B3 People of the age group between 50-54
3 A – People with Chronic Conditions A1 People of the age group between 40-49
A2 People of the age group between 30-39
A43 People of the age group between 18-29
B – Other Groups B1 People of the age group between 40-49
B2 People of the age group between 30-39
B3 People of the age group between 18-29
4 People who have not been vaccinated
Source: The Ministry of Health.
51
Turkey faces pronounced risk of disruptive external adjustment
The risk of disruptive external adjustments has been heightened by the market response to recent
events. As noted, the dismissal of the Central Bank Governor after four months in the role initially triggered a
very sharp adjustments in Turkish financial and asset markets. Measures of sovereign risk, such as the 5-year
credit default swap spread, experienced their sharpest one day rise on record (Figure 143). The Turkish Lira
experienced a 9.3 percent depreciation in one day and the Borsa Istanbul saw its largest one-day losses. While
clear communication of monetary policy and the CBRT’s commitment to price stability since then helped calm
markets somewhat, the sensitivity markets have shown to signals of a course change in monetary policy, and
Turkey’s continued, sizeable external financing requirements, suggest there is a high risk of further market
disruption and external pressures on the Lira.
Figure 142: Market-based inflation expectations Figure 143: …and indicators of risk spiked
increased sharply…
Turkey breakeven inflation 2 Year (%) Bond Yields and 5Y CDS Spreads
21 30 750
19 25 650
17 20 550
15 15 450
13
10 350
11
9 5 250
7 0 150
Mar-18
Jun-18
Mar-19
Jun-19
Mar-20
Jun-20
Mar-21
Dec-18
Dec-19
Dec-20
Sep-18
Sep-19
Sep-20
5
1-Jul-2020
15-Jul-2020
29-Jul-2020
10-Mar-2021
24-Mar-2021
12-Aug-2020
26-Aug-2020
4-Nov-2020
9-Sep-2020
18-Nov-2020
2-Dec-2020
16-Dec-2020
30-Dec-2020
23-Sep-2020
7-Oct-2020
21-Oct-2020
13-Jan-2021
27-Jan-2021
10-Feb-2021
24-Feb-2021
2-yr bond yield 10-yr bond yield
5-yr CDS, RHS
Source: Central Bank of Turkey, Turkish Ministry of Treasury and Finance, Source: IMF, World Bank estimates.
World Bank staff estimates and calculations.
Notes: Excludes trade credits and foreign exchange deposits.
Recent market volatility heightens stress and risks in the corporate and banking sector. A sharp fall in
the Lira and renewed uncertainties exert severe stress in the corporate and banking sectors. In common with
many other countries, corporates have recently increased their overall debt leverage during the COVID-19
pandemic and their balance sheets have deteriorated significantly. A new round of permanent depreciation is
likely to bring further cash flow and debt repayment problems in a fragile economic recovery environment. The
corporate sector, which has around US$$18.5 billion of external debt (excluding trade credits) maturing in one
year 51, struggles with both high financing costs and currency shock at a time of worsening global financing
conditions and high uncertainty regarding the pace of recovery. The high frequency corporate vulnerability
index -as measured by probability of default- has already shown recent signs of heightened corporate distress.
Those signs will in turn stoke concerns over the ability of banks to withstand a possible rise in distressed assets.
Capital buffers are likely to be eroded by a falling Lira, and the securities portfolio could be negatively affected
by rising bond yields.
Turkey may be exposed to further deteriorations in global liquidity. External financing shocks could hit
Turkey hard, and a possible unwinding of quantitative easing in advanced economies could lead to a new period
of turbulence. Inflation expectations are beginning to rise in the US after a long time, with the latest US$1.9
51The corporate sector has a short-term FX surplus of US$25 billion. This is supportive of firms’ capacity to handle short
term currency shocks.
52
trillion fiscal stimulus package projected to close the output gap later this year. With these developments, US
Treasury bonds, and other advanced market bond yields have begun to appreciably rise. An intensification of
market speculation over an end to very loose monetary policy in advanced economies could cause destabilizing
movements in global liquidity away from emerging markets such as Turkey, akin to the taper tantrum episode
in 2013. As in 2013, this could pressure the central banks of emerging-market economies to raise interest rates
to prevent significant weakening of their currencies and a build-up of inflationary pressures. Even when
excluding foreign exchange currency, the deposits of non-residents and trade credits, the monthly GEFR is
estimated at around US$7bn in 2021 (Figure 144). Under normal market conditions, a large proportion of debt
maturing is rolled over, whereby the net external financing requirement is considerably lower than this.
However, should market sentiment deteriorate during high external repayment periods, Turkey may face
external volatility. The IMF’s “Assessing Reserves Adequacy” metric for Turkey is built up based on possible
reserve needs in the event of different shocks, suggesting that in the event of a sudden stop, a large amount of
reserves may be required (Figure 145).
Figure 144: External financing need remains Figure 145: And reserve adequacy metrics suggest a
substantial substantial need in the event of a shock
Estimated gross external financing req. (US$m) Components of Turkey's "Assessing Reserve
12,000 Adequacy" metric
140
10,000
8,000 120
Non-res.
6,000 100 capital risk,
39
4,000
80
2,000
0 60 Sudden
stop, 51
Apr-21
May-21
Aug-21
Sep-21
Nov-21
Mar-21
Jun-21
Dec-21
Jul-21
Jan-21
Feb-21
Oct-21
40
Res. capital
Non-interest CAB Fin. corp. 20 risk, 20
Non-fin. corp. Government ToT risk, 12
Public banks 0
Source: Central Bank of Turkey, Turkish Ministry of Treasury and Finance, Source: IMF, World Bank estimates
World Bank staff estimates and calculations
Notes: Excludes trade credits and foreign exchange deposits.
Building and maintaining buffers to the extent possible is the best insurance against external
financing shocks. At a macroeconomic level, the CBRT will be able to more comfortably exercise its mandate
of smoothing volatility in the exchange rate under any circumstances with a higher level of international
reserves. Efforts to build reserves so far are yielding results and should be continued to the extent possible. As
it stands, reserves net of predetermined and contingent short-term drains are at negative US$45bn (Figure 146),
US$81bn below their level at the end of 2019. As can be seen in the GEFR (Figure 144), substantial external
payments arise for the government, especially in March. To this end, it is prudent that the Treasury has
expanded its FX deposits at the CBRT, which as of end February 2021, stood at nearly US$18bn, and which is
more than is required to settle external financing requirements in 2021. Banks continue to maintain adequate
FX liquid assets to cover short-term FX liabilities falling due over the coming months. Short-term liabilities of
the banking sector as of 2020 Q4 amounted to US$ 86 billion (Figure 147). Still, the majority of banks’ FX
buffers are held at the Central Bank and constitute a large proportion on the Central Bank FX reserves. In the
event of a future shock there will continue to be the risk of potentially competing demand for FX between the
Central Bank and commercial banks. This would, again, be best addressed by the Central Bank raising the level
of its own reserves.
53
Figure 146: International reserves have fallen well below Figure 147: Banks maintain liquidity to meet ST FX
the level of short-term possible drains liabilities
International reserves and short-term drains FX Liquidity of the Banking System
(US$ billion) (US$ million, December 2020)
160 120,000
Accounts Reserve Options
140 payable & other 9,006 Mechanism (Jan 2021)
120 12m public 90,000 Required Reserves in
debt service 43,490.86 FX
100 Other
80 FX required 60,000 FX receviables from
reserves Banks
Gold 19,784.43 86,015
60
30,000 FX receviables from
40 29,687.66 CBRT
Swaps
20 FX FX Cash
0 7,193.22
- Banks' FX FX Liability due
Assets Short-term drains Liquid Assets in 1 year
Source: Central Bank of Turkey. Source: Central Bank of Turkey.
Notes: Data as of March 5, 2021. Short-term drains consist of
predetermined and contingent drains within 12 months.
C. Policy should be geared towards lowering risks in the near-term
As Turkey recovers from a succession of shocks, policy should focus on reducing the risk of further
volatility. While Turkey’s economy has recovered rapidly from both COVID-19 and previous shocks, it
remains susceptible to further, damaging bouts of volatility. Even at the cost of some foregone growth in the
short-term, policy settings should be tilted towards stabilization, until risks have been lowered and buffers
increased. This would include stabilizing inflation around a credible policy anchor, building external reserves to
comfortably withstand the external volatility Turkey will likely continue to face, resolving corporate distress,
cleaning bank balance sheets and maintaining a prudent level of public debt, as well as accounting for contingent
liabilities and possible debt service shocks.
Box 6: Turkey’s new Economic Reform Package
The President of the Republic of Turkey launched a new Economic Reform Package (ERP) on March 12,
2021. The overarching focus on the ERP is on the “establishment and continuity of macroeconomic
stability”. The program is structured on two pillars, “macroeconomic policies” and structural policies” which
each contain five focus areas each. While the ERP is focused on reforms within the remit of the central
government, in order to be successful in achieving these high-level objectives, it will be necessary that
monetary policy remains geared towards price and macroeconomic stability.
The ERP also contains a commitment for the program to be monitored on a quarterly basis. Successful
coordination and implementation of a large number of potentially complex reform initiatives will be crucial
to maximizing the positive impact of the program, and therefore a focus on implementation is positive.
Related to that, several reforms represent important, but selective steps of broader reform programs, and
complementary reform agendas could be considered as part of the implementation of the ERP. The
following discussion summarizes several important reform areas under the ERP, but is not an exhaustive
discussion:
Almost one third of the specific reforms enumerated under the ERP are within the area of public finance,
making this a substantial focus for the reform in the coming period. The publication of a new quarterly
54
Public Finance Report offers the opportunity to convey regular updates on the progress towards fiscal policy
objectives and in-year changes in fiscal policy. Several reforms are committed to in the areas of public
procurement and coordinated budgetary and debt management of central and general government
institutions. These reforms offer the potential to improve coordination, although in several areas, the
authorities may face tradeoffs such as between domestic preference and quality assurance, centralized
budgetary oversight and appropriate levels of operational flexibility for certain agencies. In these cases, a
careful approach to reform, which may involve the implementation of novel approaches (such as net
budgeting for certain budgetary units) would be called for. There is also a significant reform agenda in the
area of revenue administration, with several reforms targeted at improving the customer experience, and
reducing the compliance burden of the taxpayer. In addition to reforms to standardize and improve the
efficiency of tax compliance, these measures could bring about the double dividend of improving the
business environment, while also increasing the efficiency of the government’s revenue collection.
In support of the CBRT’s clear focus on price stability, the government focuses on supporting structural
reforms. These consist of a number of measures aimed at addressing market failures in agricultural value
chains that contribute to higher and more volatile prices, such as supporting a competitive and well-regulated
agricultural intermediation market, improving cold chain supply chains and facilitating more flexible
agriculture practices such as contract farming. In addition, the plan focuses on coordinated and evidence-
based monitoring of price stability risks, including by establishing a price stability committee and an early-
warning system to monitor emerging inflationary risks in a coordinated manner across government.
The ERP has a number of measured aimed at supporting financial sector development, including measures
to facilitate the issuance of corporate bonds, the regulation of green bonds, and measures to improve the
institutional framework for participation finance. These are all positive steps towards an important objective
of financial sector diversification, which will be aided by the broader focus of the ERP of macroeconomic
stability. The ERP also proposed a new regulatory framework for NPL venture capital funds. The impact of
both this and financial sector diversification policies would be enhanced as part of a more holistic program
of reform in these areas, with several supporting measures not explicitly addressed in the ERP.
The ERP supports structural policies across several areas related to investment, competition policy,
consumer rights and industrial policy. In these areas, measures to consolidate and clarify the set of investment
incentives, introduce an investment dispute mechanism and develop private sector focused sectoral
coordination bodies in important areas like ICT and the health industry should help encourage domestic and
foreign investment alike. Proposed improvements to retail trade regulations and fair-trading oversight are
also positive steps.
Under the banner of the current account, the ERP contains measures to export growth in target markets,
improvements to the oil and gas investment climate and a restructuring of EXIM bank to better support its
objectives.
Employment-related reforms include measures to improve the attractiveness of vocational training and
internship, as well as better coordination of employment incentive programs. A major new credit program
will offer CGF loans linked to the number of new staff that firms hire and retain for a period of 2 years.
The ERP contains a high-level commitment to green transformation and several measures aimed at
supporting such a change. This includes the development of green bonds, green organized industrial zones,
planning for circular economy reform, development of electric vehicle infrastructure and support for green
technology R&D.
Source: Presidency of the Republic of Turkey (www.tccb.gov.tr ); World Bank Staff.
55
Tighter monetary policy should be maintained until inflation is under control
Anchoring inflation expectations through the continuation of a credible and transparent monetary
policy stance is critical for the disinflationary process going forward. Empirical studies have found that
persistent and sizable upside breaches of the inflation targets have weakened the anchoring power of the targets
through time in Turkey. 52 This is also evident from a recent survey conducted by the CBRT in which market
participants ranked the inflation target as the least important variable in driving their forecasts of inflation 53. In
the light of these findings, a continued commitment to the current policy stance of appropriately tight monetary
policy to lower inflation in the short-term, and a conventional, transparent policy framework would help better
anchor inflation expectations, support policy credibility, and achieve disinflation.
The existing monetary policy framework is appropriately geared toward reduced inflationary pressure
and needs to be maintained. However, the disinflation path will be set back to the extent that there are price
shocks to the economy. The sharp Lira depreciation in March that followed the replacement of the Central
Bank Governor is one such example. The new Governor has appropriately committed to a continuation of
existing policies to achieve a permanent inflation reduction. Avoiding premature monetary easing will remain
critical to avoid excessive volatility in the capital markets, address dollarization, anchor inflation expectations
and contain pressure on the Lira. The Economic Reform Package (ERP) launched in March 2021 provides a
useful anchor to support economic recovery and stability but needs a sound macroeconomic policy framework
to achieve desired results. Steps to consolidate Central Bank independence as part of the authorities’ ongoing
reform program would also support the objective of the ERP.
Fiscal policy plays an important role in securing the recovery, but increases in debt should be
limited
Fiscal policy will remain important in supporting the recovery from COVID-19 in 2021. Increased
government spending has supported both expanded health system capacity and COVID-19 monitoring and
prevention (e.g. through social distancing and mobility restrictions), which will continue to be important over
the course of 2021, along with an increased need for support to the nationwide vaccination program. Support
to enterprises was mainly through deferral of social security premium payments and tax payments. In parallel,
fiscal policy helped buffer liquidity support through the banking system by expanding the Credit Guarantee
Fund, which does not affect the short-term fiscal stance, but is a contingent liability (Figure 148). Most of this
support has continued in early 2021 to contain the pandemic. Given the lack of room for monetary policy to
loosen, Turkey can continue to maintain responsive and flexible fiscal policy to manage the difficult recovery
ahead so as to avert significant social and economic costs. Supporting vulnerable households and mitigating
employment losses are of utmost importance.
Fiscal consolidation is expected to be gradual to contain the ongoing adverse impacts of the
pandemic. Despite some erosion in the last year, Turkey still has fiscal space to support the economy.
Premature sharp fiscal consolidation in 2021 might hurt vulnerable households and corporates and delay the
orderly recovery. The general government deficit for 2020 is projected to be 3.5 percent of GDP as the need
for additional support to cushion the economic and social impact of the pandemic will continue, particularly in
2020H1. However, fiscal support needs to be timely, targeted and time bound. A gradual fiscal consolidation is
expected to take place in 2022 and 2023, as envisaged in the government’s New Economic Program. The
consolidation will come mainly from public transfers, which have grown sharply over the past couple of years.
Public investment, which was cut sharply to create more room for transfers in recent downturns, is assumed to
recover and support medium term growth. Public investment54 will be important for raising potential growth
52 Gulsen and Kara (2021)
53 Feedback results from survey respondents show that the top six variables driving the inflation forecasts of professionals are exchange
rates, inflation outturn, monetary policy stance, oil prices, economic activity, and the near-term historical average of inflation.
54 This requires careful prioritization and selection of investment projects, and further improvement in PIM efficiency to ensure good
project outcomes and a high growth impact.
56
going forward, as it will take time for private investment to pick up. As temporary tax reductions and
government spending on public transfers are withdrawn, the government deficit is expected to fall back to
around 2.6 percent in 2023.
Figure 148: Support to health, households, and Figure 149: Primary expenditure rise to be driven
businesses was provided mainly by current transfers
80
Announced fiscal measures (cumulative score 2021 Central Government Budget Targets
based on importance of individual measures) (% change compared to 2020 realization)
60
Cumulative score
Interest
40 Lending
Capital Transfers
20 Capital Expenses
Current Transfers
0
16-Mar-20 16-Jun-20 16-Sep-20 16-Dec-20 Purschase of G&S
Personnel
Primary
Direct tax relief Indirect tax relief
Total Expenditures
Subsidies to enterprises Subsidies to households
Support to health system Tax Amnesty -45 -25 -5 15 35
Sources: WB Staff based on government announcements. Source: MoTF.
Note: Scores determined based on the magnitude of the announced
measure, ranking between -3 and +3 (+3 being the most impactful in terms
of coverage, number of people affected, and positive externalities).
Ensuring that short-term responses are consistent with fiscal efficiency and sustainability is of course
critical. Short-term measures that lead to too much debt or unaffordable expenditure, for example, could
reverse economic recovery. Measures should be transparent, timely, targeted, and temporary. Risks to fiscal
sustainability can also be mitigated through discretionary measures that can be easily reversed when the effects
of the pandemic shock are over. The authorities’ fiscal response to date seems to have been largely consistent
with these principles. In the latest New Economic Program (2021-2023), the government envisages a gradual
consolidation to help support a recovery. Only capital transfers and purchase of goods and services were set to
decline in 2021, whilst almost 40 percent of the primary expenditure increase in 2021 comes from the rise in
current transfers. Along with the ramp up in borrowing and rising interest rates, the burden on interest
expenditures in the budget is envisaged to rise in 2021 (Figure 149).
Recently announced fiscal reform of the government (Box 6) would support rationalization of public
expenditures, boost accountability and transparency and mitigate fiscal risks. One of the main focus
areas of the recently announced reform agenda is building a stronger and more resilient public finance system.
Implementation of a new quarterly Public Finance Report is a welcomed step which would help to promote
transparent and open practices of fiscal management. The report could cover an appraisal of factors impacting
fiscal outcomes, including the influence of policy changes, as well as quantified, forward-looking assessment of
fiscal risks. A robust legal framework for PPPs would be another important step and could be complemented
by regular reporting on PPP developments. Other efforts such as simplification of incentive schemes, a review
of tax exemptions, and SOE reform could also support the rationalization of public expenditures.
Rising public debt across advanced economies and historically low interest rates have led some to
speculate that countries may be able to abandon previously held prudential debt limits. Debt to GDP
in the U.S. has risen from 77 percent in 2000 to 163 percent in 2020 Q3, while real interest rates declined from
4.3 percent in 2000 to -0.1 percent in 2020 55. The decline in interest rates across advanced economies has been
driven by massive central bank liquidity injections since the 2008 Global Financial Crisis and persistently low
55 Summers, L., Furman, J. 2020. “A Reconsideration of Fiscal Policy in the Era of Low Interest Rates”. Brookings Institution.
57
inflation rates. While high debt and low interest rates are associated with “secular stagnation”, there has been a
recent tendency towards a new fiscal policy framework. A recent IMF 56study advises a reconsideration of public
finance rules in light of a persistent high debt environment. Summers et al. (2020) notes that although low
interest rates have brought challenges for countercyclical monetary policy, they also create some opportunities
for the economies in terms of improved debt sustainability and greater fiscal space for public investments.
Debt levels in Turkey, as with most emerging markets, remain relatively low, although the interest
burden is high. Selected major EM 57 economies’ central government debt to GDP ratio averaged at 38.1
percent in 2000, which has risen to 46 percent for a selection of major EMs, well below the OECD average of
81 percent in 2019 (Figure 150). Turkey’s central government debt to GDP ratio was 30.8 percent in 2019, the
lowest amongst selected emerging market economies. However, Turkey’s central government interest payments
to GDP ratio was 2.3 percent, placing it in the midrange and equal to the average for selected EM countries
(Figure 151). By comparison, interest payments to GDP is only 0.27 percent for the US economy, and just 0.04
percent for Germany.
Figure 150: Turkey has relatively low public debt… Figure 151: …with a relatively higher interest burden
Central Government Debt to GDP (in percent, Interest payments to GDP (in percent, 2019)
2019)
4.5
160 4
140 3.5
120 OECD 3
Average 2.3
100 2.5
80 2
60 1.5
40 30.8 1
20 0.5
0 0
Source: Haver Analytics, OECD. Source: Haver Analytics.
Turkey has a relatively higher proportion of external debt, and its external financing costs are also
higher. Turkey’s public external debt to GDP ratio was 13.3 percent, compared to 17.5 percent for domestic
debt to GDP in 2019 (Figure 152). This 43 percent share of external debt is one of the highest among major
EMs, with only Indonesia being higher. While some emerging market economies such as Poland and Hungary
were able to access external funding with very low spreads, thanks to their commitment to EU economic
integration; Argentina, South Africa, Turkey and Mexico’s spreads were above 400 bps at end 2020, exceeding
the EMBI global average of 357 bps (Figure 153). Higher country risk pricing for these EMs is likely a key
factor explaining why these countries do not benefit in the same way from loose global liquidity conditions.
56 International Monetary Fund, 2021. “Fiscal Monitor: Policies for the Recovery”. IMF, October 2021.
57 Turkey, Romania, Mexico, Indonesia, Poland, South Africa, Hungary, and Argentina.
58
Figure 152: Debt breakdown differs among EMs Figure 153: Conditions for external funding diverge
70 Breakdown of public debt (in percent of GDP, Eurobond spread for Emerging Market
60 2019) 1600 Economies (2020Q4, in bps)
50 1400
40 1200
30 1000
EMBI
800 Global
20
600 Average
10
400
0 200
0
Argentina Turkey South Mexico Hungary Poland
Africa
Domestic debt to GDP External debt to GDP
Source: Haver Analytics. Source: JP Morgan/ Haver Analytics.
Turkey’s domestic bond yields and EMBI sovereign spreads have been trending up since 2013.
Turkey’s inflation-adjusted cost of borrowing has increased significantly between 2013 and 2019. 10-year
domestic bond yields had risen from 6.5 percent in early 2013 to 21 percent by the end of 2018, while sovereign
spreads followed a similar pattern (Figure 154). Between 2012 and early 2020, the correlation between domestic
bond yields and sovereign spreads was 0.9 (Figure 155). The relationship has weakened during the COVID-19
crisis due to the liquidity crunch, as well as the government policies in Turkey that helped reduce the cost of
domestic borrowing 58.
Figure 154: Volatile trend in bond yields Figure 155: High correlation with CPI expectations
23 Inflation expectations, bond yields and 800
Turkey's domestic bond yields (in percent) and
29 900 20
sovereign spread
EMBI sovereign spread (bps)
800 600
24 17
700
600 14 400
19
500 11
400 200
14 8
300
9 200 5 0
2012 - Jan
2013 - Jan
2014 - Jan
2015 - Jan
2016 - Jan
2017 - Jan
2018 - Jan
2019 - Jan
2020 - Jan
2021 - Jan
2012 - Jul
2013 - Jul
2014 - Jul
2015 - Jul
2016 - Jul
2017 - Jul
2018 - Jul
2019 - Jul
2020 - Jul
100
4 0
CPI in 12 months 10-yr bond yields
2-yr bonds 5-yr bonds
10-yr bonds Sovereign spread (right axis) Sovereign spread (right axis)
Source: Haver Analytics. Source: Haver Analytics.
Turkey’s public debt is sensitive to interest rate shocks. Under a scenario where the real effective interest
rate is assumed to increase by 500 basis points, the government’s borrowing cost increases to over 16 percent
and the debt-to-GDP ratio remains above 37 percent over the medium term (Figure 156). In a different context,
where the interest expenditure was assumed to rise by 2 standard deviations, Turkey’s interest expenditures to
GDP ratio immediately jumps from 2.3 percent to 4.8 percent from 2019 to 2020, while Mexico’s interest
58 See Turkey Economic Monitor, August 2020 for details.
59
expenditure responds to the same shock with a lower magnitude, despite its debt (Figure 157). On the other
hand, the debt of advanced economies such as Germany and the US does not respond significantly to the
shock.
Despite low public debt to GDP, Turkey is especially exposed to sharp increases in debt servicing
costs. Continued high inflation expectations, as well as the volatile interest expenditure pattern raise sensitivities
over fiscal policy in Turkey. Turkey and several other EMs may not be able to benefit from extremely loose
global liquidity conditions, and therefore may continue to be constrained by country-specific macrofiscal
factors.
Figure 156: Increased debt burden with a real interest Figure 157: With rising interest burden
rate shock
41 Interest expenditure to GDP and a potential
Nominal Debt-to-GDP 9 shock (in percent)
40
8
39 7
38 6
37 5
36 4
3
35
2
34
1
33 Baseline Real Interest Rate Shock 0
32 2018 2019 2020 2021 2022 2023
2020 2021 2022 2023 2024 2025 Mexico Turkey Indonesia Romania USA Germany
Source: WB Staff calculations. Source: Haver Analytics.
Note: A 500 bps shock was applied to Turkey’s expected 2020 general Note: A 2 standard deviation shock was applied to the 3-year moving
government debt stock. average interest expenditure-to-GDP ratio after 2019.
Overall, Turkey cannot afford the luxury of increasing debt that advanced economies may be able to,
but can increase fiscal space by seeking to lower country risk pricing. Despite the discussion in advanced
economies of lower interest rates for longer, Turkey can ill afford to face a shock with higher debt levels, given
its high susceptibility to shocks, which could lead to a sharp increase in debt costs. Such a scenario might arise
for various reasons including a sharper-than-expected pick-up in the global economy, a domestic shock, or an
increase in risk-aversion towards emerging markets such as Turkey. As a result, Turkey’s current levels of debt
– relatively lower than other EMs – is prudent, and it is too early to consider major expansionary fiscal policy.
Instead, Turkey’s current economic stance, focused on macroeconomic stabilization and gradually raising
sustainable growth prospects can, over time, lead to improved financing terms, which could create the
conditions for prudently raising debt levels in future.
NPL resolution and corporate insolvency frameworks will be crucial this year
Turkey's banking sector played a key role in alleviating the economic impact of COVID-19, but now
the BRSA and banks need to prepare for an increase in NPLs, which will also impact capital positions.
Strong credit growth in 2020 amid weak economic prospects coupled with loan repayment moratoria have led
to a deterioration in underlying asset quality 59, currently masked by regulatory forbearance measures related to
the classification of NPL and Stage 2 loans. A recovery this year could lower the risk of an NPL uptick once
forbearance measures are phased out, although an extension of COVID-19 restrictions, a sharp slowdown in
59 As noted below, banks have set aside extra provisions for possible increases in NPLs.
60
credit growth and rising interest rates following monetary tightening would increase the risk of asset quality
deterioration and, eventually limit banks’ support to economic activity (see Box 7). The high share of Stage 2
and restructured loans reveals underlying risks for NPL developments; however, conservative provisioning
against expected credit losses exemplifies banks’ prudent approach towards preparing for a surge in NPLs. An
increase in distressed loans, heightened bond yields and continued Lira volatility will also pressure capital
positions and eventually loan growth, especially once capital-related forbearance measures expire. Regulators
and banks should be vigilant.
Box 7: Preventing an expected surge in NPLs from becoming a drag on economic growth
The COVID-19 crisis is one of the worst economic shocks in history. While it did not start as a financial
crisis, signs are pointing towards increasing financial sector vulnerabilities globally. In particular once
unprecedented regulatory forbearance measures are phased out, financial institutions will need to deal with
a rise in non-performing loans (NPLs). The longer the pandemic lasts, the more businesses and consumers
will struggle and the stronger the impact on repayment challenges will become, ultimately undermining
economic growth. Preventing a prolonged episode of low growth and lackluster financial sector performance
across the world requires a swift and comprehensive response, broadly comprising four measures:
• First, a robust regulatory and supervisory framework is needed so that banks properly identify
NPLs and provision for credit losses. Banks have played a vital role in responding to the pandemic,
providing credit to corporates and households struggling with a drop in liquidity. While this credit
expansion makes banks more vulnerable to an increase in NPLs given fragile economic prospects, it
should not be an excuse for undermining existing standards and diverting from international norms. In
cases where internationally agreed standards on the definition of NPLs were relaxed, the time has come
for an orderly exit from these measures. The same holds for broader borrower relief measures. Credible
regulation and supervision are needed to instill confidence and prevent the build-up of more serious
risks. Lessons from the Global Financial Crisis show that a wait-and-see approach only makes a bad
situation worse when it comes to NPLs.
• Second, banks need to get operationally ready for resolving rising volumes of NPLs. This requires
functional workout units in banks, internal policies for the management and resolution of NPLs, and
methodologies for assessing the viability of distressed borrowers. The latter is vital for ensuring that
borrowers considered non-viable do not benefit from constant loan restructurings, but are properly dealt
with through the insolvency framework. Otherwise, inevitable credit losses just end up being delayed,
while locking up resources and constraining the channeling of new credit to deserving companies.
• Third, the COVID-19 crisis will test existing insolvency frameworks. While unviable borrowers
need to be dealt with expeditiously, potentially viable, but distressed borrowers need an opportunity to
be rehabilitated. Encouraging and improving out-of-court workouts for these borrowers is a top priority.
At the same time, legal frameworks need to enable debt reduction, and institutions need to be able to
deal with the to-be-expected increase in proceedings. Increasing digital procedures and simplifying
administrative processes can contribute significantly.
• Finally, a nationwide NPL reduction strategy can be an important component in lowering NPLs
in a holistic and well-coordinated manner. Active participation of the private and public sector is
critical in these efforts.
Source: “COVID-19 and Non-Performing Loan Resolution in the European and Central Asian region”. FinSAC Policy Note, December 2020
http://pubdocs.worldbank.org/en/460131608647127680/FinSAC-COVID-19-and-NPL-Policy-Note-Dec2020.pdf
Safeguarding banking sector stability calls for a transparent and well-communicated exit strategy from
forbearance measures. Any extensions under improved health and economic circumstances exacerbate
expectations that the borrower relief measures constitute a "new normal", making it increasingly difficult to
61
revert to the status quo pre-COVID-19. Public communication about these preconditions for revoking borrower
relief measures is important in managing expectations and ensuring that the temporary nature of the measures
is well understood by borrowers and banks alike.
Effective insolvency and corporate debt restructuring frameworks will be key elements of the COVID-
19 recovery. Turkey was successful in preventing viable firms from being prematurely pushed to insolvency
due to COVID-19, but now a growing number of firms may need an insolvency process to survive (Table 5).
An enabling environment for corporate and consumer debt restructuring and an effective out-of-court debt
restructuring framework are important to prevent a surge in insolvency filings, value-destroying liquidations,
and asset fire sales. This can help preserve employment and reduce pressure on bank balance sheets, which
would otherwise be affected by a surge in NPLs. SMEs, accounting for three-quarters of employment, will be
particularly hard hit by this COVID-19 crisis. It is vital to ensure the smooth functioning of workouts and debt-
restructuring mechanisms. At the same time, it will be critical to assess firm viability and identify zombie firms 60,
to free up capital for lending in future periods.
Table 5: Options to respond to the COVID-19 challenges impacting solvency of firms and individuals?
Key Challenges Critical Responses
Phase 1 Preventing viable firms Implementing one or more extraordinary measures for a limited
from prematurely being period of time:
pushed into insolvency • Increasing the barriers to creditor-initiated insolvency filings;
• Suspending the director's duty to file and associated liability;
• Ensuring complementarities with debt repayment emergency
measures.
Phase 2 Responding to the Ensuring the smooth functioning of workouts and debt restructuring
increased number of mechanisms:
firms that will not • Establishing informal out-of-court, or hybrid workout frameworks;
survive this crisis • Facilitating business rescue through bridge financing;
without going through • Extending procedural deadlines for a limited period of time;
insolvency
• Suspending the requirement to proceed to liquidation if the business
activity of the debtor has stopped while undergoing reorganization;
• Encouraging e-filings, virtual court hearings and out-of-court
solutions in insolvency cases.
Phase 3 Addressing individual • Implementing modern consumer bankruptcy frameworks, which are
financial distress still non-existent or outdated in many Emerging Markets and
resulting from the crisis Developing Economies;
• Ensuring there are flexible options for debt rescheduling and
repayment plans;
• Enabling a debt forgiveness mechanism or discharge is important in
facilitating a fresh start.
Source: "COVID-19 Outbreak: Implications on Corporate and Individual Insolvency"
http://pubdocs.worldbank.org/en/912121588018942884/COVID-19-Outbreak-Implications-on-Corporate-and-Individual-Insolvency.pdf
60 Zombies are firms that earn just enough money to continue operating and servicing debt, but are unable to pay off their debt, thus
diverting resources away from healthy, viable firms. Adalet McGowan et al (2017): “The walking dead: zombie firms and productivity
performance in OECD countries,” and Ricardo Caballero et al (2008) “Zombie Lending and Depressed Restructuring in Japan.”
62
Box 8: Conditional Counterfactual Analysis of NPLs and CARs in the absence of government actions
The COVID-19 shock could have led to a significant increase in NPLs, while lowering CARs
throughout 2020 without the authorities' swift action. This box illustrates how NPLs and CARs would
have evolved in the absence of forbearance measures, loan payment deferrals and facilitated restructurings by
comparing counterfactual outcomes to those observed in light of government and banking sector actions. The
model, based on statistically significant impulse-responses and employing conditional counterfactual analysis,
shows that NPLs could reach much higher levels within a short period if faced by demand and FX shocks. The
highest NPL increase in relative terms is predicted for state banks, despite their low NPL base, but domestic
private and foreign banks would also face noticeable increases. The decline in CAR, on the other hand, is
predicted to be relatively limited.
Turkey's banking sector played a key role in tackling the COVID-19 crisis, but faced underlying asset
quality deterioration, which is eased by regulatory forbearance measures to support borrowers and
banks in weathering the pandemic fallout. The Turkish authorities introduced extensive regulatory
forbearance measures to protect financial stability and viable borrowers from March 2020 onwards, shielding
banks' balance sheets. Higher economic growth in 2021 could lower an NPL uptick risk once forbearance
measures are phased out after 2021Q2 61. Still, an extension of COVID-19 restrictions, tightened monetary and
credit conditions, and recently heightened financial volatility may contribute to further asset quality
deterioration. The impact of a rise in NPLs has been highlighted in previous editions of the TEM 62, finding
that asset quality shocks significantly decrease GDP and reduce CAR, leading banks to reduce loan supply and
increase lending rates. But with corporate sector vulnerability elevated due to COVID-19 and a higher interest
rate environment, the introduced regulatory forbearance measures and heavy loan restructurings make it
difficult to assess to what extent such shocks have already been realized in 2020, and what the risk to banks'
financial positions is in the upcoming period.
With this in mind, we first empirically identify the impact of COVID-19 on the NPLs and CARs of the
domestic private, state, and foreign banks using publicly available group-level data 63. Second, by using
the realized economic indicators in 2020, we employ conditional counterfactual analysis to understand the
possible trajectories of NPLs and CARs in the absence of policy actions.
Estimating the Effect of Demand and Exchange Rate Shocks on NPLs and CARs
The starting point of our analysis is to employ a standard Bayesian Vector Auto Regression (BVAR)
model, a popular tool for forecasting, policy, and scenario analysis of possible multivariate time series
models, as they do not suffer from an endogeneity problem (Ogunc, 2019). It also allows dynamic
interactions between the variables and works well with relatively short data samples. We set up a structural
BVAR with sign restrictions using the Bayesian Estimation, Analysis and Regression (BEAR) toolbox of
Dieppe et al. (2016).
Two major shocks, namely demand and exchange rate shocks, appear to affect NPLs and CAR
adversely during the COVID period. Following the related literature, we employ the variables and sign
restrictions listed in Table A. Conventionally, most of the time series econometric literature estimates the
impact of demand or currency shocks based on a recursive identification scheme. However, such an
identification scheme may have some disadvantages. First, selecting the correct ordering of the variables is
difficult to establish. Different ordering of the variables in the model may lead to different impulse response
functions. Second, such an identification scheme is not based on economic theory, but rather, on the speed at
61 Central Bank of the Republic of Turkey. 2020. “IV.1 Credit Developments and Credit Risk”, Financial Stability Report, November
2020, pg 48.
62 https://openknowledge.worldbank.org/bitstream/handle/10986/32634/Turkey-Economic-Monitor-Charting-a-New-
Course.pdf?sequence=1&isAllowed=y
63 BRSA and CBRT
63
which different variables are assumed to adjust to different shocks. To address these issues, the sign restriction
identification scheme, which is also more intuitive, is widely implemented via the methodology employed in
Rubio-Ramirez et al. (2010) 64. By using this identification scheme, we assume that adverse demand shocks are
generated through the GDP changes. Those shocks are assumed to reduce GDP, increase NPLs and reduce
CARs. Similarly, a foreign exchange shock is generated through increasing FX volatility, which increases NPLs 65,
reduces loan supply, raises interest rates, and decreases CAR.
Table 6: Sign restrictions enabling the identification of demand and currency shocks on NPLs and CARs
Implied FX Real Interest
GDP Loans* NPL CAR
Volatility** Rate***
Demand shock - + -
Exchange rate shock - + + - +
*(Deviation from the trend), (**1-month, US$/TRY, quarterly max-min difference), ***(Weighted average bank loan interest rates
minus inflation)
The model is estimated at a quarterly frequency over the period of 2010Q4 to 2019Q4. GDPs are in year-
on-year percentage change. FX shock is estimated by using the implied volatility of TL against US$. Real interest
rate, NPL ratio and CAR are included in year-on-year differences. Loans are taken as the difference from its
trend and entered in the model as year-on-year differences.
The figure below shows the responses of the NPL ratio and CAR to a one standard deviation shock to
demand and exchange rate. It can be seen from the figure that a negative demand shock significantly increases
NPLs for all banks on impact and causes a decline in CARs. The effect is sustained for at least 4 quarters
thereafter. Following the depreciation of the Turkish Lira, similar results were observed, suggesting that a
sizeable share of FX-denominated balance sheet items leads to an increase in NPLs and decline in CARs.
Demand and FX shocks have a relatively greater adverse effect on the state banks’ NPLs. Although the shape
of responses to demand and FX shocks appears similar, 66 their magnitudes and uncertainty bands differ, and
currency shocks to NPLs are relatively more persistent. In addition, uncertainty surrounding demand shocks is
greater than for currency shocks. Both shocks reveal similar sign restrictions, but responses show that the
exchange rate plays a critical role in shaping the expectations of economic agents and the overall performance
of the economy.
64 We assume a Minnesota prior distribution to obtain the posterior estimates. The structural model is identified with sign restrictions
implemented using Arias, Rubio-Ramirez, and Waggoner (2014) methodology.
65 Impact of the FX vulnerability on short term debt repayment is expected to be negative. On the other hand, NPL ratio stands for
total non-performing loans over gross loans. Depreciation will create denominator effect, as well. Thus, the overall impact might not
be clear. However, TL credits is around 2.4 trillion TL and FX credits are US$167 billion (1.2 trillion TL, as of February 2021).
Therefore, this denominator effect would be small. Moreover, FX shocks may not work only through the loan channel. Its impact on
the economic activity and on the borrowers is assumed to be negative and may diminish the payment abilities of the borrowers.
66 The general similarity of responses can be due to the similarity in transmission mechanism of exchange rate shocks and demand
shocks. Depreciation of the exchange rate eventually leads to the lack of demand for imported goods and domestic goods produced
with a share of higher imported intermediate goods. Ogunc et al. (2018) suggests that the exchange rate not only serves as a means of
storing value and a hedge against inflation, but also plays a critical role in determining the general prospects for medium and long-
term economic performance.
64
Figure 158: Impulse-Responses to negative demand and currency volatility shocks
Domestic Private Banks State Banks Foreign Banks
0.4 0.2 0.5
0.3 0.4
0.15
0.2 0.3
Demand 0.1 0.2
0.1
Shock to 0.05 0.1
NPL 0 0
-0.1 0
-0.1
-0.2 -0.05 -0.2
1 3 5 7 9 1113151719 1 3 5 7 9 1113151719 1 3 5 7 9 1113151719
0.2 0.2 0.2
0.1 0.1
0 0
0
-0.2 -0.1
Demand -0.1
-0.2
Shock to -0.2 -0.4 -0.3
CAR -0.3 -0.4
-0.6
-0.4 -0.5
-0.5 -0.8 -0.6
1 3 5 7 9 1113151719 1 3 5 7 9 1113151719 1 3 5 7 9 1113151719
0.2
0.4 0.5
0.3 0.4
0.15
0.2 0.3
Currency 0.2
0.1 0.1
Shock to 0.1
NPLs 0 0
-0.1 0.05
-0.1
-0.2 -0.2
0
1 3 5 7 9 1113151719 1 3 5 7 9 1113151719
1 3 5 7 9 11 13 15 17 19
0.2 0.2 0.2
0 0 0
Currency -0.2
-0.2 -0.2
Shock to -0.4
CARs -0.4 -0.4
-0.6
-0.6 -0.8 -0.6
1 3 5 7 9 1113151719 1 3 5 7 9 1113151719 1 3 5 7 9 1113151719
Note: Figures show responses of NPLs and CARs to a one standard deviation negative demand shock and a one standard deviation currency shock.
The Y-axis shows the percentage point differences to the previous year. The X-axis shows quarters. While the solid line shows the baseline estimation,
dashed lines show the 85 percent confidence bands for the BVAR estimation.
Conditional Counterfactual Analysis of NPLs and CARs
We use the structural BVAR model findings and restrictions presented above to empirically estimate
how NPLs and CARs would have evolved in the absence of forbearance measures, loan payment
deferrals and facilitated restructurings by comparing counterfactual outcomes to those observed
under the government and banking sector actions. Hence, we conduct a conditional counterfactual analysis
for the 2020Q1-2020Q4 period. Realized GDP, implied US$/TRY volatility, loan growth deviation from its
trend, and real lending rates are used as conditions for the 2020Q1- 2020Q4 period. We use a 95-percent
65
confidence level for this exercise. Although the magnitude of the results may change depending on the different
factorization schemes under BVAR, or any different specification in other methodologies, they are intuitively
parallel to the literature and market assessments on the trajectory of the NPLs and CARs.
Figure 159: Counterfactual NPL ratio: Private banks Figure 160: Counterfactual CAR: Private Banks
12% 22
10% Forbearance Forbearance
20 Lower bound
Lower bound
8% Baseline
Baseline 18 Upper bound
6% Upper bound
16
4%
14
2%
0% 12
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
2020.Q4
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
Figure 161: Counterfactual NPL ratio: State banks 2020.Q4 Figure 162: Counterfactual CAR: State Banks
14% 22
12% Forbearance Forbearance
20 Lower bound
10% Lower bound Baseline
18 Upper bound
8% Baseline
6% Upper bound
16
4%
14
2%
0% 12
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
2020.Q4
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
2020.Q4
Figure 163: Counterfactual NPL ratio: Foreign banks Figure 164: Counterfactual CAR: Foreign Banks
14% 22
12% Forbearance
20
10% Lower bound
Baseline 18
8%
Upper bound
6% 16 Forbearance
Lower bound
4% Baseline
14 Upper bound
2%
0% 12
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
2020.Q4
2017.Q1
2017.Q2
2017.Q3
2017.Q4
2018.Q1
2018.Q2
2018.Q3
2018.Q4
2019.Q1
2019.Q2
2019.Q3
2019.Q4
2020.Q1
2020.Q2
2020.Q3
2020.Q4
Estimation results assume that NPLs could escalate sharply due to weakened economic activity and
currency depreciation without forbearance measures, loan deferrals and restructurings, given 2020
GDP, FX volatility and loan growth. As of 2020Q4, the NPL ratio would have been 2 times higher than the
reported figures in state banks. Baseline results show that NPL ratios could have reached 9.7, 5.7 and 10 percent
for domestic private, state and foreign banks, respectively, at the end of 2020. Those figures are also close to
the reported Stage-II loan ratios of the aforementioned banking groups.
66
The decline in CARs would have been relatively limited. Indeed, historical data also shows that variation
in CARs among bank groups is relatively low against adverse shocks indicating that banks are managing shocks
well by deleveraging, decreasing risky loan books, loan restructuring and increasing their capital buffers. CAR
estimates imply a relatively limited decline, to 17.4, 17.3 and 18.4 percent for private, state and foreign banks,
respectively in the baseline forecast for the same period. The worst-case forecasts point to 15.3, 14.5, and 16.6
percent CAR ratios for the same period. Estimates showing a relatively limited decline in CAR are also similar
to the banks’ recent CAR announcements after deducting the forbearance impact they share with the public.
Overall, our conditional counterfactual estimations show that forbearance measures, loan payment
deferrals and loan restructurings helped banks to manage pressure on both NPL levels and CARs. In
fact, BRSA data 67 on loan classification forbearance, suggests that the impact of the forbearance of NPL past
due day extension is limited to 70 bps on the banking sector’s NPL ratio as of February 2021.
However, the impact of the demand and FX shocks may be sustained. Therefore, the banking sector
requires a transparent and well-communicated exit strategy from forbearance measures in the second half of
2021 to preserve the credibility of the financial system. Strengthening NPL resolution, insolvency, and out-of-
court corporate debt restructuring frameworks with an effective corporate viability assessment is also critical in
order to protect corporates and the banks from spillovers.
References
Arias, J. E., Rubio-Ramirez, J. F., and Waggoner, D. F. 2014. Inference Based on SVARs Identified with Sign and Zero Restrictions: Theory and
Applications. Dynare Working Papers 30, CEPREMAP.
Dieppe, A., Legrand, R. and van Roye, B. (2016). "The BEAR Toolbox", ECB Working Paper 1934, July 2016
Huljak, I., Martin, R., Moccero, D. and Pancaro, C. (2020), Do non-performing loans matter for bank lending and the business cycle in euro area
countries, ECB Working Paper No 2411, May 2020
Kanngiesser, D., I., Martin, R., Maurin, L., and Moccero, D. (2017), Estimating the impact of shocks to bank capital in the euro area, ECB Working
Paper No 2077, June 2017
Ogunc, F. 2019. " A Bayesian VAR approach to short-term inflation forecasting", CBRT Working Paper, 19/25, Central Bank of the Republic of Turkey
Us, Vuslat. (2017), Dynamics of non-performing loans in the Turkish banking sector by an ownership breakdown: The impact of the global crisis,
Finance Research Letters, Volume 20, 2017, Pages 109-117
World Bank. 2019. "Turkey Economic Monitor: Charting a New Course" Washington, DC: World Bank
D. As a foundation for higher medium-term inclusive growth
If the perquisite of improved economic stability is in place, a structural reform agenda can help to
ensure raised, inclusive growth in the medium-term. Addressing near term challenges for the economy to
ensure orderly recovery highlighted in the previous section is a key first step. Selected reform programs could
then help mitigate the risk of long-term adverse effects of the pandemic and transition to a sustainable growth
path. Reforms should include targeted measures to support the most vulnerable, encourage labor market
flexibility, facilitate corporate debt relief and increase international economic integration.
The COVID stimulus offers an opportunity for building a more resilient, sustainable economy
The scale of post-COVID recovery programs can have a significant impact on transforming economies
for the better. As the risk of uncontrolled COVID-19 outbreaks gradually abates, particularly with vaccine
rollout in 2021, many countries around the world have now shifted their attention to recovery efforts, with
stimulus packages underway. 68 The International Monetary Fund (IMF) estimates an unprecedented global
fiscal response of US$12 trillion as of October 2020. 69 As of October 2020, the Government of Turkey’s
estimated discretionary fiscal support package was TL646 billion (~ US$ 92.5 billion, or 12.8 percent of GDP).70
67 Unpublished data received via exchange of letters with the authorities.
68 OECD&IEA. 2020. Addressing the COVID-19 and climate crises: Potential economic recovery pathways and their implications for
climate change mitigation, NDCs and broader socio-economic goals. Climate Change Expert Group Paper No.2020(4). France.
69 https://blogs.imf.org/2020/10/14/fiscal-policy-for-an-unprecedented-crisis/
70 https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19#T
67
Given the magnitude of these stimulus packages, they will have a significant impact on a country’s economic
prospects, in addition to environmental and social effects with long-lasting implications in many cases.
The COVID-19 recovery phase offers Turkey an opportunity for rebuilding its economy on a more
resilient, inclusive, and sustainable growth path—often referred to as “Building Back Better”. Fast
paced technological developments and changing consumer preferences are transforming the global trade
landscape; and countries need to adapt to maintain their competitive edge in global markets. This calls for well-
targeted stimulus measures and structural policy reforms to rebuild more competitive, efficient, and resilient
economies, so as to enable more sustainable growth. 71 Turkey remains a resource-intensive economy with
increasing carbon intensity and decreasing trends in material productivity, compared to OECD countries that
are reducing their carbon intensity.
Turkey is already elevating climate and environmental policy, and linking this agenda to the recovery
program could accelerate progress. A strong focus on cross-sectoral policies that stimulate technological
innovations and modernize the economy can help Turkey improve material efficiency, maintain
competitiveness, create jobs, and support overall development priorities. Stimulus used in this way would not
cost more – but rather, be used more efficiently and for more sustainable outcomes.
The path to greener technology adoption in many sectors is quite established, and steps could be
taken to align with the evolving requirements of the EU and other key trade partners. Stimulus resources
could be used to incentivize resource efficiency and also position Turkish firms strategically if such instruments
as the Carbon Border Adjustment Mechanism (CBAM) come to fruition. There are a wide range of instruments
that can be used in combination to stimulate resource efficiency and improved environmental performance.
These instruments can be categorized into three groups (Figure 165): Regulatory instruments that mandate
specific behavior; market-based instruments that act as incentives for particular activities, and information-
based instruments that seek to change behavior through the provision of information.
Figure 165: Range of environmental policy instruments
There are several opportunities for Turkey to utilize its current stimulus package in more targeted
ways. Elements of Turkey’s stimulus package are considered green, including the introduction of a “Green
Tariff” for power generated from renewable sources, support for solar power, and extension of the Renewable
Energy Support Scheme. Turkey is also committed to increasing solar energy production capacity by 1 GW,
while the recently announced ERP includes commitments to support further green reforms (see Box 6). There
71 OECD&IEA. 2020. Addressing the COVID-19 and climate crises: Potential economic recovery pathways and their implications for
climate change mitigation, NDCs and broader socio-economic goals. Climate Change Expert Group Paper No.2020(4). France.
68
are also further opportunities to orient COVID recovery support towards incentivizing green growth. Some of
the short-term measures being undertaken by other countries include: i) bailouts with green strings attached; ii)
nature-based solutions; iii) loans and grants for green investments; iv) green R&D subsidies; and v) subsidies
or tax reductions for green products for environmentally relevant sectors such as agriculture, energy, industry,
transport, and waste. As governments around the world are rolling out their recovery packages, many have
committed to these green stimulus measures.
Box 9: Green COVID-19 recovery measures for various sectors: Country examples
- Spain’s US$13.5 billion commitment for sustainable agriculture and urban development 72
- India’s US$780 million fund for afforestation is expected to create a large number of jobs to
Agriculture support low-income and vulnerable communities 73
- China’s US$4 billion Green Development Fund will support green investments along the
Yangtze River economic belt
Australia’s committed more than US$233 million to hydrogen development - France’s new
‘Relance’ stimulus package of US$6.1 billion for green energy infrastructure and US$7.4 billion
for building renovations for energy efficiency.
- The UK’s US$3.7 billion support for energy efficiency improvements in residential and public
Energy buildings and additional support for wind energy.
South Korea’s US$130 billion stimulus package, providing funding for the Korean ‘New Deal’
which includes funding for renewables, electric and hydrogen vehicles, and energy efficiency in
buildings. - Germany’s funds to coal workers and companies to phase out coal-fired power
plants by 2038, support for renewable electricity, hydrogen, and provision of subsidies to avoid
additional charges on consumers in financing renewable energy expansion. 74,75
- France’s US$7.7 billion bailout deal to Air France having green strings attached – a 50 percent
reduction of emissions by 2030 and a minimum standard of 2 percent renewable fuel. 76
- Germany’s commitment to €1 billion green innovation and industry transformation in the
vehicle sector, €1 billion for customer rebates for EVs to 2025, and €1 billion for a scrappage
Transport scheme for older trucks to support fleet modernization. 77
- The UK’s allocation of US$2.5 billion for public investment in cycling and pedestrian
infrastructure 78 and US$250 million to support green R&D in aerospace.
China’s (US$379 million for EV charging infrastructure and US$14 billion for railway
infrastructure)
- France’s US$1.4 billion for industry decarbonization and US$264 million for circular economy
Manufacturing efforts
- The UK’s US$450 million fund for emissions reductions in heavy industry, including CCS and
clean hydrogen, materials, new technologies, and efficient construction.
72 https://www.miteco.gob.es/es/prensa/ultimas-noticias/el-idae-destina-24-millones-de-euros-en-la-segunda-tanda-de-
convocatorias-para-financiaci%C3%B3n-de-proyectos-renovables-innovadores-en-cinco-comunid/tcm:30-512142
73 India TV (2020). https://www.indiatvnews.com/business/news-nirmala-sitharaman-final-phase-of-announcement-economic-
stimulus-package-11-am-live-updates-617884 , Jagran (2020). https://english.jagran.com/business/economic-package-tranche-2-
mnrega-support-free-foodgrains-for-migrants-s-0000-crore-additional-credit-support-for-farmers-10011841 , Economic Times (2020).
https://bfsi.economictimes.indiatimes.com/news/policy/key-highlights-of-the-finance-ministers-whole-economic-package/75797903
74 Carbon Brief (2020). https://www.carbonbrief.org/coronavirus-tracking-how-the-worlds-green-recovery-plans-aim-to-cut-
emissions
75 Recharge News (2020). https://www.rechargenews.com/transition/germany-lowers-renewables-surcharge-to-cushion-covid-
impact-on-consumers/2-1-894231
76 Routes Online (2020). https://www.routesonline.com/news/29/breaking-news/291047/air-france-told-by-government-to-
drastically-cut-domestic-flying/
77 Reuters (2020). https://de.reuters.com/article/us-germany-autos/germany-to-up-financial-aid-for-cars-sector-government-sources-
idUSKBN27X1S7
78 Forbes (2020).
https://www.forbes.com/sites/carltonreid/2020/05/09/uk-government-boosts-bicycling-and-walking-with-ambitious-2-billion-post-
pandemic-plan/#3a5ce00a3d7c
69
Turkish firms can have doors opened worldwide with a new drive to reduce trade barriers
Reducing barriers to trade in goods and services will likely expand potential for Turkish businesses.
Turkey’s applied ‘most-favored nation’ (MFN) tariffs are relatively high and have a large number of peaks. In
2019, Turkey’s simple average of MFN tariffs stood at 10.0 percent, and at 4.5 percent on a trade-weighted
basis (latest, 2018), putting Turkey in the mid-range of all countries (at 79th out of 138 countries) 79. In addition
to import tariffs, most imported goods also have a standard-rate 18 percent value-added tax levied at the border.
7.2 percent of tariff lines are more than triple the average rate (of 10 percent) and Turkey is amongst 25
countries in the world with the highest number of peaks in its tariff profile. While agricultural tariffs are much
higher, a substantial fraction (9.3 percent) of non-agricultural products also face MFN tariffs in excess of 25
percent.
Imports of a significant number of products from countries not in a customs union or free trade
agreement with Turkey pay “additional duties” over and above the applied MFN rate. A World Bank
analysis finds that the simple average of the additional duties across tariff lines is 16.8 percent, ranging from 0
to 30 percent. According to the current tariff schedule, these are to be applied until September 30, 2021, and
lowered by 10 percentage points thereafter. The additional duties cover an estimated 9 percent of Turkey’s 2019
imports, and do not apply to imports from countries with which Turkey has a free trade agreement. The
additional duties cover 42 percent of textiles, clothing and footwear imports; 10 percent of other manufactured
imports, and about 0.2 percent of food imports. On a geographical basis, they fall most heavily on imports
from South Asia (39 percent) and East Asia (31 percent). The additional duties effectively reverse most of the
“Everything but Arms” duty-free treatment that Turkey would otherwise be compelled to grant to countries
such as Bangladesh, Cambodia, Myanmar, Nepal, and Pakistan. By estimated dollar value of 2019 imports,
additional duties fall most widely on imports from China (US$8.2 billion), India (US$2.7 billion), the United
States (US$1.4 billion), and Japan (US$1.3 billion).
Turkey’s services restrictiveness is relatively high compared to OECD countries. In the OECD’s
Services Trade Restrictiveness Index for 2020, the average regulatory restrictiveness score for Turkey is 0.348,
where 0 is completely unrestricted and 1 is completely restricted. This score shows that Turkey lags behind
most OECD countries, especially in accounting, legal services, air transport, courier services, and broadcasting.
Revitalizing the EU-Turkey Customs Union can help realize greater gains from trade and GVCs
The EU-Turkey Customs Union covers the largest share of Turkey’s trade, in the deepest relationship.
Customs unions are generally closer partnerships than a preferential trade agreement (PTA) offers, because
they involve adopting a single customs area, with no tariff restrictions and a common external tariff. This
arrangement can offer great benefits – without administrative restrictions at the border, goods can move quickly
and freely providing just the kind of efficient, large and diversified economic area that highly-efficient GVC
firms crave.
But the Customs Union remains incomplete, and it lacks mechanisms to ensure a level playing field
with third parties. Many modern PTAs include deep provisions on topics such as services, regulatory
coherence, and competition policy. While Turkey undertook to adopt the EU regulatory acquis when it joined
the Customs Union, there is no systematic mechanism for services liberalization between the EU and Turkey.
With services accounting for more than two-thirds of global GDP, more than three-quarters of foreign
investment in advanced economies, and half of value-added in manufacturing exports, remaining barriers to
services trade and investment are a significant lag on competitiveness.
79 Based on the WTO’s 2020 World Tariff Profiles report and simple-average MFN rates. MFN tariffs on industrial products alone are
considerably lower: The simple average MFN applied rate is around 4.2 percent and trade weighted average MFN applied rate is 1.38
percent. Considering the duty exemptions under inward processing regime and the GSP, average MFN applied rates decreases further.
70
The fact that Turkey does not have PTAs with all countries that the EU does leads to an asymmetry
of market access. Figure 154 shows Turkey’s network of preferential trading agreements (light green) as well
as those countries for which the EU has preferential agreements, but Turkey does not (dark green). For the
second group of countries, the agreements modify the EU’s external tariff, which is also the Customs Union
tariff. Thus, Turkey is compelled to give duty-free treatment for goods from these countries without receiving
reciprocal market access.
Figure 166: Turkey and the EU’s preferential trade agreements in force
Source: Turkish Ministry of Trade, European Commission, World Bank Staff estimates.
Note: Light green denotes where Turkey has a PTA in force, dark green where the EU has a PTA in force, but Turkey does not.
At the same time, new large-scale EU-wide policy initiatives which have the potential to impact
Turkish competitiveness are largely beyond Turkey’s control. The EU Green Deal could impose border
adjustment taxes affecting Turkey’s exports, as well as a wide range of other environmental policies that could
impact the cost of production and the eligibility of Turkish goods to enter the EU market. Similarly, the
proposed Digital Single Market is significantly changing the regulatory environment for the digital economy in
the EU. An important feature of these initiatives is that they are either in the design phase or under construction.
Thus, non-EU firms operating in the EU market – including Turkish firms – face significant policy uncertainty
going forward. The Government of Turkey is currently preparing a roadmap to support competitiveness and
effective economic partnership in light of the European Green Deal.
Going forward, significant structural changes are needed to preserve and extend the benefits of the
Customs Union. These could include, but are not limited to, an active program of deepening existing
agreements between the EU and Turkey to include new subject areas; services liberalization by Turkey;
discussions to resolve outstanding issues in the Customs Union; developing new mechanisms for Turkey to
engage with EU-wide policy initiatives, and proactive responses from the private sector to anticipated changes
in EU regulatory frameworks.
71
A more diversified and greener financial system would support a resilient, sustained recovery
Developing domestic sources of long-term finance would alleviate existing imbalances in the financial
system and raise economic growth. Turkey's bank-centric financial system is exposed to several structural
challenges. Banks account for 90 percent of all financial sector assets, while insurance and pension funds each
constitute only around two percent. The financial system has long suffered a chronic shortage of long-term
finance, especially in local currency. Having access to more Lira-denominated funding would reduce Turkish
banks' and corporates' exposures to short-dated and FX-denominated debt, reduce rollover risks, and lower
financial stability risks. Long-term finance is crucial for firms to invest and grow, and to fund Turkey's large
investment and infrastructure needs.
Capital markets instruments can create momentum for long-term finance development in Turkey.
Changes in the Capital Markets Law in the pre-COVID period were expected to encourage capital market
development. Amendments to the Capital Markets Law that entered into force in February 2020 introduced
global standards to Turkish capital markets laws, and were thus expected to encourage that development. These
amendments introduced establishments such as the noteholders' meeting, security agents and project bonds.
Other modifications included expanding the scope of crowdfunding, limiting the range of significant
transactions of public companies, exit rights, and tender offer bidders.
But a volatile economic environment and its symptoms have acted as a constraint to capital market
and long-term finance development. Macro-financial dynamics are among the most significant barriers to
long-term domestic finance and capital markets development. These include; a challenging macroeconomic
environment that dampens the confidence of foreign and domestic investors; a low national savings rate, high
dollarization, and low sovereign credit ratings, and high local currency government bond yields across the yield
curve and limited financial and capital market literacy. The COVID-19 crisis, as well as the recent financial
turmoil, further amplifies these challenges.
Challenges also remain related to the issuers and investors' side and the regulatory framework. The
most binding constraint is the small size of the institutional investor base. While pension funds have recorded
rapid growth in recent years with the introduction of automatic enrollment, the insurance sector, pension funds
and mutual funds together are only equivalent to six percent of GDP. There is a need to develop financial
instruments for long-term finance, including, but not limited to sustainable and green finance purposes (Box 9:
Greening the financial sector). In promoting the development of the insurance and pension markets, Turkey
Wealth Fund consolidated six state-owned insurance companies under a single insurance and pension company
to restructure the insurance sector and create economies of scale. However, a more holistic approach requires
a regulatory framework to manage institutional investors' portfolio management and investment policies to
increase the saving pool and encourage investing in longer-term, local currency and sustainable finance
instruments.
SME finance is a critical area in need of long-term finance. Large and sophisticated corporates and banks
enjoy good access to domestic and international financial markets. However, SMEs, which account for most
of the corporate sector, continued to face severe challenges in accessing long-term finance. Bank financing has
been the traditional source for SMEs. While the CGF has helped SMEs otherwise struggling to meet the high
collateral requirements imposed by banks, high-interest rates and the short-term nature of bank financing
remain a constraint.
Capital markets could promote SME finance in Turkey mainly through private markets, although
these markets and regulatory frameworks have yet to be developed. The imposition of disclosure and
governance requirements sets a high entry barrier for SMEs for accessing the stock exchange, and thus private
markets including bank finance, private equity, venture capital and other sources are often more attractive for
SME finance. Bond issuance is also challenging for SMEs, given the absence of strong governance structures
and information disclosures. However, SMEs' characteristics render them unsuitable for public markets since
they are not in a position to have the requisite financial disclosure, either because they cannot generate it, or
72
because the cost is too high. Equity financing solutions also require both that SMEs open their capital to outside
shareholders, which many SMEs are reluctant to do because of their family structure, and that investors
demonstrate a much higher risk appetite than many do.
The size of the SME financing gap calls for additional solutions to expand SME financing the world
over. In advanced economies the low interest rate environment of recent years has resulted in lower returns
for the large fixed-income holdings of institutional investors; as a result, their search for alternative investments,
including SME finance instruments has increased. The increased interest of institutional investors in ESG
investment also has a positive effect on SME financing. Financial technology has triggered the emergence of
new capital markets solutions that allow start-ups and SMEs in need of funding to connect more directly with
investors. Financing raised through electronic platforms has been growing at a fast pace over the past few years.
Many emerging and developing economies are seeing their pension funds grow, financial technology is opening
the doors to new mechanisms for market-based financing, and a growing middle class is increasingly investing
in capital markets through mutual funds. 80
Alternative finance solutions are available for SMEs in Turkey, despite their insignificant size
compared to bank finance. Both indirect solutions that offer SMEs refinancing facilities and direct finance
solutions that provide SMEs direct access to capital markets could be financing alternatives. Depending on the
level of market development and the regulatory environment, indirect mechanisms for SME financing
instruments could be plain vanilla issuances by SME lenders, SME loan securitization, and SME structured
notes. Direct debt finance solutions are receivable-based solutions, loan-based solutions, and security offering
solutions. Direct solutions could be private equity and venture capital, equity crowdfunding, and SME offerings
on equity finance. Nevertheless, concerns regarding macroeconomic volatility, governance issues on both
investor and issuer sides in equity finance, and an enabling environment, especially the insolvency regime,
diminish the development of innovative instruments.
Box 10: Greening the Financial Sector in Turkey
Turkey's financial sector and capital markets offer the potential to facilitate a green transition,
mobilizing capital for sustainable and green recovery. Demand for green financing is likely to grow,
particularly with the EU Green Deal coming into force (see page 67). The country can quickly benefit from
the increasing ESG focused interest from domestic and international investors and issuers. A holistic
approach including regulatory framework, a taxonomy, supervision, guidance, strategies, and innovative
instruments is necessary. Among other enablers needed to achieve green growth 81, a supporting and robust,
vibrant financial system will be required with substantial financing needs intermediating through banks,
insurers, and other capital providers. However, the financial system will need accurate, readily available, and
equivalent databases, measurements, methodologies, guidelines, and frameworks to handle climate-related
risks.
Climate change and environmental issues are increasingly recognized as important sources of risk
and opportunity for the financial sector. A growing number of central banks and regulators across the
globe have issued warnings on the impact of climate and environmental risks on their financial institutions
and systems' stability. These risks can materialize through two channels: physical and transition risks. Physical
risks stem from both gradual and abrupt impacts of climate change, including natural disasters relevant to
Turkey, such as droughts, floods, and landslides, affecting economic performance and the value of real or
“World Bank Group. 2020.” Capital Markets and SMEs in Emerging Markets and Developing Economies: Can They Go the
80
Distance?”
World Bank. https://openknowledge.worldbank.org/handle/10986/33373
81 The World Bank defines green growth as that which is efficient in its use of natural resources, clean in that it minimizes pollution
and environmental impacts, and resilient. It accounts for natural hazards and the role of environmental management and natural
capital in preventing physical disasters.
73
financial assets. Transition risks derive from efforts to transition to a resilient low-carbon economy and
improve local environmental conditions, creating significant economic adjustment costs across a broad range
of sectors.
There is global recognition of the emerging role of financial sectors in mobilizing capital for green
objectives, including for those laid out in national climate action plans and strategies. The benefits of aligning
with the European Green Deal, which aims to facilitate public and private investments needed to transition
to a climate-neutral, green, competitive and inclusive economy, further add to the demand for sustainable
finance in Turkey, considering the EU's share in Turkey's external trade.
Green finance remains at an early stage of development in Turkey. Turkey's financial sector plays a
crucial role in delivering on the green agenda and mobilizing capital for climate action. However, climate
change poses risks to financial stability and the soundness of institutions. As the COVID-19 crisis has shown,
major disaster events have direct impacts on the financial sector. They heighten levels of uncertainty
regarding the medium to long‐term macro‐economic outlook and growth, increase public and private sector
financing needs, cause structural changes to the economy, and influence wealth and income distribution.
Climate and environmental crises can have a similar impact and threaten the stability of financial institutions
and systems. With increasing knowledge and awareness of these risks and more frequent occurrence of
extreme weather events, and understanding of transition risks, Turkey's financial sector must be ready and
able to respond to a crisis caused by climate.
Domestic funding for green investments in Turkey depends mostly on the banking sector and faces
many green financing hurdles. The banking sector in Turkey, which is a big part of its financial sector,
which has played a crucial role in the country's growth of renewable energy production and the issuance of
green bonds and mortgages, is well established and has taken measures towards sustainable finance.
However, challenges to such initiatives involve, but are not limited to, unpredictable policy conditions, lack
of knowledge and institutional capability, and a low degree of preparation to promote healthy green
development.
In Turkey, green finance is understood by the financial sector as part of the broader sustainable
finance agenda. The Turkish banks' market-led course to sustainable banking is aligned with national goals
and international principles and good practice. This understanding takes place in a broad perspective within
the framework of the UN Sustainable Development Goals.
The banking associations and financial sector's top players have undertaken voluntary industry-led
initiatives. Turkish financial institutions have incorporated environmental and social issues with a blend of
risk management and loan origination. In 2014, the Banks Association of Turkey (BAT) issued Turkey's
voluntary Sustainability Guidelines for the Banking Sector. The Guidelines were prepared by a BAT working
group on the Role of the Financial Sector in Sustainable Growth, with the participation of 18 banks.
Sustainability guidelines were revised by the BAT in March 2021. In 2019 six leading Turkish banks signed
the UN's Responsible Banking Principles to implement sustainability. Additionally, several financial and
nonfinancial corporates devoted themselves to issuing sustainability reports to raise investor awareness and
resilience against shocks.
Green financial regulations drive capital away from unsustainable investments
Investors and the financial sector need clear and credible guidance for a green economy from the
authorities. The challenges related to funding green investments need a favorable policy and regulatory
structure to be developed, accompanied by the institutional capacity to execute them. Providing best practice
advice on sustainable finance to the banking sector will help contribute to better finance for green
development. The authorities and regulators could maximize their efforts to guide financial sector players on
environmental sustainability. Turkey needs to develop a holistic regulatory framework, guidance, a green
74
labeling taxonomy (possibly aligned with the EU taxonomy 82), strategies, and innovative instruments to
expedite the greening of the financial sector and make it resilient to climate risks. Supervisors or regulators
can be instrumental in enhancing awareness by carrying out an initial assessment of Turkey's financial sector's
exposure to key sectors most exposed to climate risk (e.g., fossil fuels, power, agriculture, automotive, and
steel). In the recent Economic Reform Package, the authorities shared their agenda to strengthen the
ecosystem of green finance. The Capital Markets Board is assigned as the responsible authority to issue green
bond and green sukuk guides in cooperation with the Presidency’s Finance Office and Borsa Istanbul.
Financial institutions, including banks, insurance companies, and pension funds need to be more
aware of the climate risks they face. Climate risk is a new and essential driver of financial risk. Financial
supervisors must push financial institutions to manage this risk adequately. The information that the financial
sector uses is only as vital as the corporate disclosures it is built upon. That is why the quantity, quality, and
consistency of corporate reporting on sustainability-related information must improve. Credible and
comparable information should be readily available for all market participants, shareholders, and other
stakeholders. Reporting83 aligned with the requirements of global accounting standards ensures a level
playing field.
Capital market instruments are essential for sustainable finance to take off as the bank-dominated
financial sector has a limited ability to provide longer-term sustainable finance at an affordable cost
due to its short-term funding structure. Turkish firms are currently dependent on banks for their funding
and to a marginal extent on the capital markets. They only have limited access to private funding markets,
which are very small in Turkey. Turkish capital markets need to supply more diverse types of risk capital
required to fund the green transition. Risk capital typically comes from venture capitalists, private equity
funds, and investment funds.
Source: World Bank, Technical Note “Unlocking Green Finance in Turkey”, forthcoming
Addressing the youth, especially female youth, employment challenge critical for recovery
Both demand and supply-side constraints need to be addressed in Turkey to improve labor market
entry and continuity for women and youth. Among multiple factors to be addressed, labor market dynamics
reveal three specific constraints that should be addressed over the short- to medium-term. First, on the demand
side, the growth of firms and their hiring practices suggest that job growth has not kept up with labor force
growth. The mix of labor- versus capital-intensive activities and lagging productivity in job-creating sectors in
agriculture and services have hampered job growth for low- and high-skilled workers 84. Second, on the supply-
side, in certain sectors such as manufacturing, industry and, increasingly, services particularly with the onset of
COVID-19, firms report a lack of adequate skills among applicants. Nearly 20 percent of firms find an
inadequately skilled workforce is a major constraint to doing business. 85 The need for a digitally skilled
workforce that responds to the demand of firms as the economy becomes greener and more knowledge-based
is key. Finally, at the cross-roads between the demand- and supply-side, informality has remained relatively
high at approximately 30 percent, suggesting that incentives within the labor market for firms to continue
employing workers informally and for undeclared self-employment to persist remain high. This may be due
82 https://ec.europa.eu/info/business-economy-euro/banking-and-finance/sustainable-finance/eu-taxonomy-sustainable-
activities_en
83 The Capital Markets Board (CMB) amended the Corporate Governance Communiqué to ensure that public companies take
concrete steps to ensure sustainability. The Amendment entered into force on 1 October 2020. The CMB has also published the
“Sustainability Principles Compliance Framework” which contains the principles that public companies should follow. However, there
is no obligation to follow the ESG principles recently set out in the Corporate Governance Communiqué. But, public companies
must issue publicly available annual reports indicating as to whether they comply with these principles, or the reasons for failing to do
so.
84 World Bank (2019). Country Economic Memorandum: Firm Productivity and Economic Growth in Turkey
85 WBG Enterprise Survey 2019.
75
the perceived rate of return of labor costs in Turkey. The hike in minimum wage in 2016, for example,
hampered firm growth notably among small and medium enterprises (SME) 86.
Policies to protect workers and households most negatively impacted by shocks have been expanded,
yet remain modest in the face of COVID-19 challenges, particularly for women and youth. The choice
of instruments is key to striking a balance between supporting immediate needs and building long-term
resilience. While countries respond in different ways, in-kind and cash public support, either direct to workers
and households, or via firms, will likely be needed over the mid-term. Currently, Turkey’s main social and labor
expenditure is equivalent to around 12 percent of GDP 87. By comparison with Turkey’s major labor market
programs, unemployment benefits amount to only 0.7 percent of GDP, active labor market programs (ALMPs)
0.4 percent, and employment subsidies 0.5 percent of GDP, although these figures may not account for
expansion in 2020. In the OECD, the drop in economic growth during 2007-2009 of nearly 7 percentage points
was met by an increase in social expenditures of nearly 3 percentage points of GDP on average in the OECD
(17.7 percent of GDP to 20.7 percent of GDP).
In Turkey, a range of labor market policies and active labor market programs are available to protect
vulnerable workers in the formal sector, but effects vary. Key labor market programs include
unemployment benefits, ALMPs and wage subsidies. Wage subsidies, also known as employment incentives in
Turkey, have tended to expand following shocks, such as the post-2008 and post-2018 periods. At least thirteen
employment subsidies operate in Turkey, targeting different populations and firms, with a range of different
parameters regarding duration and benefits. While some of these programs were introduced before 2008,
several were introduced following the GFC. Among the registered unemployed, subsidies to cover wages and
social security contributions are afforded to apprentices, interns and trainees not covered by full-time job
contracts, amounting to an estimated 1.5 million individuals as of first quarter 2020. Of the employment
incentives afforded for full-time jobs, one main subsidy scheme predominates, namely Scheme 5510 (or the
Five-Points Scheme), benefiting nearly 70 percent of all firms receiving Social Security Institution (SGK)
employment subsidies. Employment subsidies are primarily financed through the Unemployment Insurance
Fund and the Ministry of Treasury and Finance. The four largest schemes reached a total of over 1.5 million
firms (out of an estimated 3.5 million active SMEs) 88 and 9 million workers in 2019 (out of 28 million) 89.
COVID-19-associated expansions of employment subsidies in terms of wage protection and social
security premia have further broadened the scope of beneficiary firms and workers. Although
employment subsidies exist to promote formal job creation, the way they are designed may not respond to new
economic realities. Their eligibility, benefit structure, duration, level and coordination with other labor market
institutions, social assistance and social insurance requires revisiting. Previous work has shown that employment
subsidies lead formal employment in small firms to increase through the formalization of existing informal jobs,
and that impacts tend to be larger in sectors such as construction and manufacturing 90. Targeting the most
vulnerable workers, notably first-time job seekers and women, would further improve efficiency and
employment impacts, although the duration of impacts depends on productivity gains and labor costs over the
mid-term.
Training-based active labor market programs covered 15 percent of the registered unemployed in
Turkey 91 in 2019, coverage which has generally increased anti-cyclically since 2007. In 2019, ALMPs
86 World Bank (Bossavie, L; Acar, A; Makovec M) (2019). Do Firms Exit the Formal Economy after a Minimum Wage Hike? World
Bank Policy Research Working Paper No. 8749. Washington DC: World Bank.
87 TurkStat 2019; Republic of Turkey Social Security Institution (SGK) (2019). Organizational Profile and Overview of the Social
Security System in Turkey. Ankara: SGK. Republic of Turkey and World Bank (2017).
88 Union of Chambers and Commodity Exchanges of Turkey,
https://www.tobb.org.tr/KobiArastirma/Sayfalar/Eng/SMEsinTurkey.php.
89 World Bank (forthcoming), Evaluation of employment subsidy schemes, from progress reviews for June 2019 and February 2020.
90 Betcherman et al., 2020; World Bank, forthcoming.
91 Registered unemployed defined as those registered with the Turkish national employment agency, ISKUR. On average, ISKUR
data captures approximately 80 percent of the total unemployed as estimated through national labor force surveys conducted by the
Turkish national statistics institute, TUIK.
76
accounted for nearly 568,000 beneficiaries (15 percent of the unemployed), compared to 1.013 million recipients
of unemployment benefits (26 percent of the unemployed) (Figure 167). As of 2020, of the 423,133 beneficiaries
newly enrolled in ALMPs, the national On-the-Job Training Program (OJT) remained the dominant choice (80
percent), with the Vocational and Technical Courses Program (VT) accounting for 20 percent. Among the
nearly 1,400 VT courses on offer, clothing and textiles was the most common occupational skill in demand,
accounting for nearly one out of three beneficiaries (26 percent). Among approximately 34,000 OJT programs
on offer, one in four beneficiaries were in sales or retail occupations, followed by clothing and textile-related
occupations (14 percent), with the remainder split across trades (metallurgy, furniture), hospitality, and other
services.
ALMPs tend to cover younger, less-skilled workers and serve as a pathway to re-skilling and
facilitating the transition to new jobs and sectors. Most ALMP beneficiaries are young adults under the age
of 34 years (77 percent), heavily concentrated among 20-24-year-olds (33 percent). The majority of ALMP
beneficiaries hold a primary or secondary education nearly split equally by educational level overall (74 percent),
although vocational courses are skewed towards primary-schooled workers rather than secondary (51 versus
27, respectively). By gender, while no major differences are seen overall among on-the-job training, vocational
course enrollment is skewed towards females relative to males (69 versus 31 percent, respectively) (Figure 169).
These patterns have been generally constant over time.
The impact of ALMPs is tied to how responsive they are to the demand by firms for certain
occupational skills, and shifts driven by COVID-19 will heighten the need for demand-driven training.
Administrative and online job vacancy data highlights the need for social as well as technical skills in the formal
sector and across regions 92. The demand for skills may have shifted pre- and post-COVID with for instance,
the need for service sector workers declining. Occupations including routine tasks (e.g. machine operators, call
center clerks and product graders) and occupations requiring non-routine manual tasks (such as customer
service) have historically been in high demand. Increasingly, IT-related and social skills (such as software
knowledge, communication and teamwork skills) and professionalism (discipline, time management) are in high
demand, particularly in regions with higher economic activity. As greater attention is given to building back
better and the green economy as part of the COVID-recovery, targeting skills retraining at youth and first-time
job seekers, particularly females, may be especially cost-effective.
Figure 167: Coverage rate of ALMPs has not been Figure 168: ...although unemployment benefits have
growing… been scaling up
(a) Coverage of labor programs among unemployed (b) Unemployment benefit applicants versus recipients,
2007-2020
70.0% 2500000
Share of registered unemployed (%)
60.0%
2000000
No. of workers
50.0%
1500000
40.0%
30.0% 1000000
20.0%
500000
10.0%
0.0% 0
% ALMP % UB % Wage Subs Applicants Beneficiaries
Source: World Bank staff using ISKUR data. Note: ALMP = active labor market programs; UB = unemployment insurance benefits; Wage Subs = wage
subsidies for apprenticeships, interns and trainees (excludes full-time job contract incentives).
92 Turkish Employment Agency; Makovec, Nas Ozen and Demirel, forthcoming.
77
Unemployment benefits also tend to go to less skilled workers and have been increasing over time.
Unemployment benefits are still relatively new in Turkey and coverage has been growing rapidly since their
introduction in 2007, reaching 1 million beneficiaries in 2019 (Figure 168), with coverage hovering at 26 percent
of all registered unemployed. Over 2020, coverage decreased by nearly half to 509 thousand, as labor force exits
increased and the redundancy moratorium took effect. Unemployment benefits remain concentrated among
workers with primary education (50 percent of beneficiaries), followed by secondary (27 percent), suggesting
the program appears to be relatively progressive by educational level. Most beneficiaries tend to be younger at
25-44 years (over 70 percent), with 20 percent aged between 45-54 years, and the remainder older. As a policy
instrument, unemployment benefits have gradually assumed a greater focus in Turkey.
Figure 169: On-the-job training more used by men, but Figure 170: And unemployment benefits are primarily
more women access vocational training claimed by men
ALMP by type and gender, 2020 140000 (d) Unemployment benefit beneficiaries by gender and
educational level, 2020
120000
250,000 100000
200,000 80000
150,000 60000
100,000 40000
20000
50,000
0
0
Male Female Male Female Male Female
Vocational On-the-Job All ALMPs
Training Courses Training Programs
Illiterate Literate Primary edn
Secondary Edn Associate Degree License
Male Female
Post Graduate
Source: World Bank staff using ISKUR data.
Moving forward, policy options for realizing the full potential of Turkey’s human capital in terms of
jobs and welfare requires a range of both short- and long-term actions, including a more integrated,
whole-of-government approach. Recommended policy options regarding jobs will depend on the specific
policy constraint and aims sought, whether projecting jobs, stimulating earnings and benefits, compensating
income losses, or fostering greater job creation. While emergency one-off, universal measures were useful early
on, as impacts manifest, more systemic reforms may be more beneficial and cost-effective. First, regarding the
demand for labor over the short-to medium-term, consolidation across Turkey’s employment subsidies,
unemployment insurance and ALMPs in terms of eligibility, targeting and monitoring will help better integrate
women, youth, low-income workers, informal workers and the non-working poor into emerging, productive
sectors. Second, ALMP financing can be strengthened through performance-based reforms to stimulate skills
and on-the-job training for low- and semi-skilled vulnerable workers, and respond to new demand in growing
sectors such as the green, digital and the digitized-economy. Finally, strengthening the school-to-work transition
over the mid- to long-term can start at an early age through a coordinated approach to reforming the education
system teaching models and linkages to ALMPs and firm growth. Measures that incentivize twenty-first century
skills and widen eligibility to on-the-job training among secondary, vocational, and higher education students
can prove effective. Harnessing such a whole-of-government vision will be an important pathway to boosting
long-term job prospects in Turkey.
78
Annex 1: Medium-Term Outlook, Nominal
Key Macroeconomic Indicators
2018 2019 2020 2021 2022 2023
Population (mid-year, million) 81.4 82.6 83.4 84.4 85.4 86.4
GDP (current US$, billion) 797.2 760.8 717.0 720.0 776.3 825.6
GDP per capita (current US$) 9792 9213 8599 9170 9089 9555
CPI (annual average, in percent) 16.3 15.2 12.3 15.5 12.0 10.0
Real Economy TL Billion, unless otherwise indicated
Real GDP 1756.1 1772.2 1803.4 1892.6 1978.4 2067.7
Private Consumption 1025.5 1041.5 1074.7 1102.1 1130.3 1162.6
Government Consumption 245.6 256.4 262.3 268.3 274.1 280.0
Gross Fixed Capital Formation 501.6 439.4 468.1 476.7 514.7 551.6
Net Exports 43.1 84.8 -12.4 34.9 48.6 63.8
Fiscal Accounts TL Billion, unless otherwise indicated
Total Revenues 1238.5 1429.9 1673.0 1942.8 2198.2 2461.4
Total Expenditures 1327.1 1561.3 1871.7 2153.2 2407.6 2665.9
General Government Balance -88.6 -131.4 -198.7 -210.5 -209.4 -204.5
Government Debt Stock 1134.0 1404.6 2001.0 2423.6 2702.1 2983.9
Primary Balance -9.3 -23.1 -51.9 -23.2 -1.4 5.8
Monetary Policy TL Billion, unless otherwise indicated
Broad Money (M3) 1994.7 2554.0 3421.5 - - -
Credit Growth (FX-adjusted, eop, y-o-y)* 1.0 8.5 37.5 - - -
Average Funding Rate (annual average, in percent) 17.7 20.7 10.5 - - -
Gross Reserves (in US$ Billion) 93.0 105.7 93.3 - - -
o/w Gold Reserves 20.1 27.1 43.2 - - -
External Sector US$ Billion, unless otherwise indicated
Current Account Balance -21.7 6.8 -36.8 -26.7 -31.0 -32.8
Trade Balance (in goods and services) -10.6 18.8 -28.3 -14.4 -19.4 -20.8
Exports 237.5 245.8 203.1 234.6 255.3 277.9
Imports 248.1 227.1 231.4 249.1 274.7 298.7
Net Foreign Direct Investment 9.2 6.3 4.7 6.4 7.8 8.6
Source: TURKSTAT, CBRT, Strategy and Budget Presidency, WB Staff calculations.
*FX-adjusted credit growth is calculated using end-year average exchange rate value of US$/TRY since 2010 for adjustment.
Annex 2: Medium-Term Outlook, Percent of GDP
Key Macroeconomic Indicators
2018 2019 2020 2021 2022 2023
Real Economy Annual percentage change, unless otherwise indicated
Real GDP 3.0 0.9 1.8 5.0 4.5 4.5
Private Consumption 0.5 1.6 3.2 2.5 2.6 2.9
Government Consumption 6.6 4.4 2.3 2.3 2.2 1.8
Gross Fixed Capital Formation -0.3 -12.4 6.5 1.8 8.0 7.2
Exports 9.0 4.9 -15.4 19.5 9.0 8.5
Imports -6.4 -5.3 7.4 7.0 6.5 6.0
Fiscal Accounts Percent of GDP, unless otherwise indicated
Total Revenues 33.0 33.1 33.1 32.5 32.2 31.7
Total Expenditures 35.3 36.1 37.1 36.0 35.2 34.4
General Government Balance -2.4 -3.0 -3.9 -3.5 -3.1 -2.6
Government Debt Stock 30.2 32.5 39.6 40.6 39.6 38.4
Primary Balance -0.2 -0.5 -1.0 -0.4 0.0 0.1
Monetary Policy Percent of GDP, unless otherwise indicated
CPI (annual average, in percent) 16.3 15.2 12.3 15.5 12.0 10.0
Broad Money (M3) 53.1 59.1 67.8 - - -
Gross Reserves 11.9 13.9 13.0 - - -
In months of merchandise imports c.i.f. 5.1 6.4 5.4 - - -
Percent of short-term external debt 79.2 86.0 67.3 - - -
External Sector Percent of GDP, unless otherwise indicated
Current Account balance -2.7 0.9 -5.1 -3.7 -4.0 -4.0
Trade Balance (in goods and services) -1.3 2.5 -4.0 -2.0 -2.5 -2.5
Exports 29.8 32.3 28.3 32.6 32.9 33.7
Imports 31.1 29.8 32.3 34.6 35.4 36.2
Net Foreign Direct Investment 1.2 0.8 0.7 0.9 1.0 1.0
Source: TURKSTAT, CBRT, Strategy and Budget Presidency, WB Staff calculations.
Annex 3: Gross Domestic Product, Production Approach
Gross Domestic Product: Production Approach
2016 2017 2018 2019 2020
GDP (current, TL billion) 2626.6 3133.7 3758.3 4320.2 5047.9
Agriculture 161.3 189.2 217.1 277.5 333.3
Industry 514.9 646.8 837.3 941.5 1130.8
Construction 222.7 266.1 267.6 233.3 272.3
Services 1418.0 1677.7 2047.1 2439.7 2758.4
GDP (constant prices, TL billion) 1586.6 1705.7 1756.1 1772.2 1803.4
Agriculture 101.3 106.3 108.5 112.6 117.9
Industry 312.7 341.8 346.4 343.1 350.1
Construction 117.3 128.3 125.9 115.0 111.0
Services 870.3 936.1 980.8 1008.8 1019.4
Real GDP Growth (in percent) 3.3 7.5 3.0 0.9 1.8
Agriculture -2.6 4.9 2.1 3.7 4.8
Industry 4.4 9.3 1.3 -1.0 2.0
Construction 5.1 9.4 -1.9 -8.6 -3.5
Services 3.4 7.6 4.8 2.8 1.1
GDP (constant prices, share)
Agriculture 6.4 6.2 6.2 6.4 6.5
Industry 19.7 20.0 19.7 19.4 19.4
Construction 7.4 7.5 7.2 6.5 6.2
Services 54.9 54.9 55.9 56.9 56.5
Source: TURKSTAT, WB Staff calculations.
Annex 4: Gross Domestic Product, Expenditure Approach
Gross Domestic Product: Expenditure Approach
2016 2017 2018 2019 2020
GDP (current, TL billion) 2626.6 3133.7 3758.3 4320.2 5047.9
Private Consumption 1560.4 1836.6 2111.0 2457.0 2864.5
Government Consumption 387.0 450.6 552.0 670.8 769.5
Gross Fixed Capital Formation 764.5 935.6 1114.7 1117.6 1372.8
o/w Construction 424.4 536.0 643.5 580.9 628.2
o/w Machinery and Equipment 279.8 323.5 379.4 417.2 576.5
Net Exports -56.8 -115.4 -5.2 121.0 -189.6
Change in Inventories -28.6 26.2 -14.2 -46.2 230.8
GDP (constant prices, TL billion) 1586.6 1705.7 1756.1 1772.2 1803.4
Private Consumption 963.4 1020.5 1025.9 1041.9 1075.2
Government Consumption 219.5 230.5 245.6 256.4 262.3
Gross Fixed Capital Formation 464.6 503.0 501.6 439.4 468.1
o/w Construction 247.7 278.5 285.2 233.1 216.7
o/w Machinery and Equipment 182.0 184.0 173.0 159.1 192.7
Net Exports -22.8 -19.0 43.1 84.8 -12.4
Change in Inventories -38.0 -29.3 -60.0 -50.2 10.2
Real GDP Growth (in percent) 3.3 7.5 3.0 0.9 1.8
Private Consumption 3.8 5.9 0.5 1.6 3.2
Government Consumption 9.5 5.0 6.6 4.4 2.3
Gross Fixed Capital Formation 2.2 8.3 -0.3 -12.4 6.5
o/w Construction 3.0 12.4 2.4 -18.3 -7.1
o/w Machinery and Equipment 0.2 1.1 -6.0 -8.0 21.1
Exports -1.7 12.4 9.0 4.9 -15.4
Imports 3.0 10.6 -6.4 -5.3 7.4
Change in Inventories -9.4 -22.9 104.8 -16.5 -120.3
GDP (constant prices, share)
Private Consumption 60.7 59.8 58.4 58.8 59.6
Government Consumption 13.8 13.5 14.0 14.5 14.5
Gross Fixed Capital Formation 29.3 29.5 28.6 24.8 26.0
o/w Construction 15.6 16.3 16.2 13.2 12.0
o/w Machinery and Equipment 11.5 10.8 9.9 9.0 10.7
Exports 22.2 23.2 24.6 25.6 21.2
Imports 23.7 24.3 22.1 20.8 21.9
Change in Inventories -2.4 -1.7 -3.4 -2.8 0.6
Source: TURKSTAT, WB Staff calculations.
Annex 5: Prices
Consumer and Producer Prices: End of period y-o-y, percentage change
2016 2017 2018 2019 2020
CPI (All items) 8.5 11.9 20.3 11.8 14.6
CPI (Food and non-alc. Beverages) 5.7 13.8 25.1 10.9 20.6
CPI (Core C) 7.5 12.3 19.5 9.8 14.3
Alcoholic beverages, tobacco 31.6 2.9 2.4 43.1 0.7
Clothing and footwear 4.0 11.5 14.8 4.5 -0.3
Housing & Energy 6.4 9.6 23.7 9.9 9.6
Health 9.7 11.9 16.7 13.6 16.7
Transport 12.4 18.2 16.0 12.2 21.1
Communication 3.2 1.4 9.6 3.2 5.7
Recreation and culture 5.9 8.4 20.9 7.0 10.4
Education 9.5 10.5 10.2 14.5 6.8
Restaurants and Hotels 8.6 11.5 19.8 13.2 12.7
Miscellaneous goods and services 11.1 12.8 28.8 13.6 28.1
PPI (All items) 9.9 15.5 33.6 7.4 25.1
Consumer and Producer Prices: Annual average, percentage change
2016 2017 2018 2019 2020
CPI (All items) 7.8 11.1 16.3 15.2 12.3
CPI (Food and non-alc. Beverages) 5.8 12.7 18.0 19.5 13.9
CPI (Core C) 8.5 10.1 16.5 13.4 11.2
Alcoholic beverages, tobacco 18.1 15.4 1.5 24.4 16.4
Clothing and footwear 7.4 7.1 13.6 5.6 5.7
Housing & Energy 6.6 8.0 15.8 13.4 12.4
Health 9.6 12.4 12.4 17.1 14.4
Transport 7.4 16.8 21.8 9.8 12.2
Communication 2.8 2.7 4.6 6.4 4.3
Recreation and culture 7.1 9.8 12.9 14.5 7.0
Education 8.2 10.0 10.5 13.5 10.5
Restaurants and Hotels 10.2 10.3 15.1 17.6 12.1
Miscellaneous goods and services 11.3 12.3 19.9 22.0 22.6
PPI (All items) 4.3 15.8 27.0 17.6 12.2
Source: TURKSTAT, WB Staff calculations.
Annex 6: Balance of Payments
Balance of Payments Statistics
2015 2016 2017 2018 2019 2020
US$ Billion, unless otherwise indicated
Current Account -27.3 -27.0 -40.8 -21.7 6.8 -36.8
Trade Balance -19.0 -19.4 -32.2 -10.6 18.8 -28.3
Exports of goods 154.9 152.6 169.2 178.9 182.2 168.4
Imports of goods 203.9 192.6 227.8 219.6 199.0 206.3
Services Balance 30.0 20.5 26.3 30.2 35.5 9.5
Primary Income Balance -9.7 -9.2 -11.1 -11.9 -12.8 -8.7
Secondary Income Balance 1.4 1.5 2.5 0.8 0.8 0.3
Capital Account 0.0 0.0 0.0 0.1 0.0 0.0
Financial Account -21.2 -21.7 -47.0 -10.9 1.3 -39.8
Direct Investment -14.2 -10.7 -8.3 -9.2 -6.3 -4.6
Portfolio Investment 15.3 -6.6 -24.3 3.1 1.4 5.5
Other Investment -10.6 -5.3 -6.2 5.6 -0.1 -8.7
Net Errors & Omissions 6.1 5.3 -6.2 10.8 -5.5 -2.9
Reserve Assets -11.8 0.8 -8.2 -10.4 6.3 -31.9
Overall Balance -11.8 0.8 -8.2 -10.4 6.3 -31.9
memo item:
Energy Balance -33.5 -24.1 -32.9 -37.8 -33.3 -24.2
Gold Balance 4.0 1.8 -10.0 -8.7 -9.3 -22.4
Percent of GDP, unless otherwise indicated
Current Account -3.2 -3.1 -4.8 -2.7 0.9 -5.1
Trade Balance -2.2 -2.2 -3.8 -1.3 2.5 -4.0
Exports of goods 17.9 17.6 19.7 22.4 24.0 23.5
Imports of goods 23.5 22.2 26.5 27.6 26.2 28.8
Services Balance 3.5 2.4 3.1 3.8 4.7 1.3
Primary Income Balance -1.1 -1.1 -1.3 -1.5 -1.7 -1.2
Secondary Income Balance 0.2 0.2 0.3 0.1 0.1 0.0
Capital Account 0.0 0.0 0.0 0.0 0.0 0.0
Financial Account -2.4 -2.5 -5.5 -1.4 0.2 -5.6
Direct Investment -1.6 -1.2 -1.0 -1.2 -0.8 -0.6
Portfolio Investment 1.8 -0.8 -2.8 0.4 0.2 0.8
Other Investment -1.2 -0.6 -0.7 0.7 0.0 -1.2
Net Errors & Omissions 0.7 0.6 -0.7 1.4 -0.7 -0.4
Reserve Assets -1.4 0.1 -1.0 -1.3 0.8 -4.4
Overall Balance -1.4 0.1 -1.0 -1.3 0.8 -4.4
memo item:
Energy Balance -3.9 -2.8 -3.8 -4.7 -4.4 -3.4
Gold Balance 0.5 0.2 -1.2 -1.1 -1.2 -3.1
Source: CBRT, WB Staff calculations.
Annex 7: Monetary Survey
Monetary Survey (TL Billion, , end of period)
2016 2017 2018 2019 2020 2021-Feb
Total Assets 1894.5 2225.0 2717.9 3348.0 4474.7 4535.4
Net Foreign Assets -42.3 -79.6 -3.1 187.8 82.5 93.9
Foreign Assets 561.8 631.2 876.0 1065.7 1194.5 1223.2
Monetary Authorities 380.3 417.1 499.1 638.6 698.4 698.7
Deposit Money Banks 181.5 214.0 376.9 427.1 496.1 524.5
Foreign Liabilities 604.1 710.8 879.1 878.0 1112.0 1129.4
Monetary Authorities 10.5 11.6 21.7 24.8 41.7 50.8
Deposit Money Banks 593.6 699.2 857.4 853.2 1070.2 1078.6
Domestic Credits 1936.8 2304.6 2721.1 3160.2 4392.2 4441.6
Net Claims on Central Government 174.5 178.2 289.3 420.9 734.2 761.7
Claims on private sector 1687.0 2025.9 2307.3 2585.0 3472.8 3493.7
Total Liabilities 1894.5 2225.0 2717.9 3348.0 4474.7 4535.4
Money 270.1 297.4 290.2 392.2 478.0 473.3
Currency in Circulation 111.3 118.5 119.1 140.7 173.7 167.4
Demand Deposits 158.8 178.9 171.1 251.6 304.3 305.9
Quasi Money 1245.5 1453.9 1794.8 2252.7 3065.3 3065.9
Time and saving deposits 682.4 764.1 876.9 991.8 1224.0 1250.1
Residents' foreign exchange deposits 517.6 631.4 862.2 1152.3 1746.2 1720.3
Securities Issued 0.0 0.0 0.0 0.0 0.0 0.0
Restricted Deposits 0.0 0.0 0.0 0.0 0.0 0.0
Other Items (Net) 378.9 473.7 632.9 703.1 931.5 996.2
Source: CBRT.
Annex 8: Central Bank Balance Sheet
Central Bank of Turkey Balance Sheet (TL Billion)
2016 2017 2018 2019 2020 2021-Feb
CBRT Assets 345.4 396.2 461.2 646.5 820.2 816.0
Foreign Assets 381.0 436.8 506.9 638.1 699.5 698.7
Domestic Assets 18.2 16.4 -0.7 58.1 188.6 165.8
Treasury Debt: Securities 13.9 14.5 13.7 19.4 89.5 88.6
Cash credits to Public Sector 13.8 14.4 13.5 19.3 89.4 88.5
Cash credits to Banking Sector 37.6 48.1 80.9 102.6 138.4 148.8
Credits to SDIF 0.0 0.0 0.0 0.0 0.0 0.0
Other Items -33.1 -46.1 -95.1 -63.7 -39.1 -71.5
FX Revaluation Account -53.8 -57.0 -45.0 -49.8 -67.9 -48.5
CBRT Liabilities 345.4 396.2 461.2 646.5 820.2 816.0
Total FX Liabilities 260.9 299.7 347.2 419.8 673.5 714.9
Foreign Liabilities 10.0 9.1 21.7 24.7 41.7 50.8
Domestic Liabilities 251.0 290.6 325.5 395.1 631.8 664.1
Central Bank Money 84.5 96.5 114.0 226.7 146.6 101.1
Reserve Money 168.0 174.1 192.2 203.8 382.3 349.9
Other Central Bank Money -83.5 -77.6 -78.2 22.9 -235.7 -248.8
Source: CBRT.
Annex 9: Fiscal Operations
General Government Budget
2014 2015 2016 2017 2018 2019
TL Billion, unless otherwise indicated
Revenues 691.2 799.2 904.3 1028.2 1238.5 1429.9
Tax Revenues 361.9 418.7 470.4 549.8 632.7 686.3
o/w Indirect 243.7 285.7 315.1 367.2 389.4 416.4
o/w Direct 106.0 118.9 138.1 164.3 221.3 245.0
Non-Tax Revenues 38.9 42.8 46.3 47.8 83.9 81.0
Factor Incomes 99.4 112.7 129.6 144.8 177.0 261.2
Social Funds 178.9 212.9 248.4 280.7 338.7 396.3
Privatization Revenues 12.1 12.1 9.6 5.0 6.2 5.1
Expenditures 701.9 801.5 940.5 1085.5 1327.1 1561.3
Current Expenditures 314.6 357.6 426.5 480.1 594.5 712.6
Investment Expenditures 66.9 81.1 91.4 115.1 141.6 115.2
Transfer Expenditures 320.4 362.8 422.6 490.3 591.1 733.4
o/w Current Transfers 295.8 339.4 399.9 466.4 564.2 698.1
o/w Capital Transfers 24.6 23.4 22.7 23.9 26.9 35.3
Overall Balance -10.6 -2.4 -36.2 -57.3 -88.6 -131.4
Interest Expenditures 51.7 54.9 52.7 60.3 79.3 108.3
Government Debt Stock 585.4 643.3 735.4 878.3 1134.0 1410.3
Primary Balance 41.1 52.5 16.6 3.0 -9.4 -23.1
Percent of GDP, unless otherwise indicated
Revenues 33.6 34.0 34.4 32.8 33.0 33.1
Tax Revenues 17.6 17.8 17.9 17.5 16.8 15.9
o/w Indirect 11.9 12.2 12.0 11.7 10.4 9.6
o/w Direct 5.2 5.1 5.3 5.2 5.9 5.7
Non-Tax Revenues 1.9 1.8 1.8 1.5 2.2 1.9
Factor Incomes 4.8 4.8 4.9 4.6 4.7 6.0
Social Funds 8.7 9.1 9.5 9.0 9.0 9.2
Privatization Revenues 0.6 0.5 0.4 0.2 0.2 0.1
Expenditures 34.2 34.1 35.8 34.6 35.3 36.1
Current Expenditures 15.3 15.2 16.2 15.3 15.8 16.5
Investment Expenditures 3.3 3.4 3.5 3.7 3.8 2.7
Transfer Expenditures 15.6 15.4 16.1 15.6 15.7 17.0
o/w Current Transfers 14.4 14.4 15.2 14.9 15.0 16.2
o/w Capital Transfers 1.2 1.0 0.9 0.8 0.7 0.8
Overall Balance -0.5 -0.1 -1.4 -1.8 -2.4 -3.0
Interest Expenditures 2.5 2.3 2.0 1.9 2.1 2.5
Government Debt Stock 28.5 27.4 28.0 28.0 30.2 32.6
Primary Balance 2.0 2.2 0.6 0.1 -0.3 -0.5
Source: Strategy and Budget Presidency, Treasury and Finance Ministry, WB Staff calculations.
Annex 10: Banking Sector Balance Sheet
Money and Banking Statistics of Financial Institutions
2015 2016 2017 2018 2019 2020
Assets Billion TL, unless otherwise indicated
Total assets 2338.3 2732.6 3263.0 3936.6 4659.4 6313.0
Net foreign assets -397.5 -433.2 -521.4 -543.7 -514.9 -692.2
Claims on nonresidents 117.3 182.2 214.9 378.7 430.2 496.2
Liabilities to nonresidents 514.8 615.4 736.3 922.4 945.0 1188.4
Claims on Central Bank 260.3 295.8 355.3 372.6 396.7 640.2
Currency 12.9 13.6 15.2 15.8 15.7 18.2
Reserve deposits and securities 247.3 282.2 339.7 356.4 379.8 613.0
Other claims 0.1 0.0 0.3 0.4 1.2 8.9
Net claims on central government 231.0 242.9 279.5 395.1 571.2 901.5
Claims on central government 287.8 307.1 353.8 470.3 666.1 1034.2
Liabilities to central government 56.8 64.2 74.3 75.3 94.9 132.7
Claims on other sectors 1533.7 1790.7 2168.0 2492.8 2821.7 3767.6
Claims on other financial corporations 40.8 48.8 61.8 69.9 84.5 111.8
Claims on state & local governments 17.6 23.4 34.4 36.9 35.6 40.2
Claims on public nonfinancial corporations 3.7 3.8 5.5 11.4 31.8 35.6
Claims on private sector 1471.6 1714.7 2066.3 2374.5 2669.8 3580.0
Liabilities Billion TL, unless otherwise indicated
Liabilities to Central Bank 112.9 106.8 99.2 119.7 105.1 228.3
Transfer deposits included in broad money 230.4 282.3 343.9 398.4 617.9 1072.1
Other deposits included in broad money 881.7 1028.7 1184.3 1442.5 1734.8 2114.9
Securities other than shares included in broad money 27.4 26.3 38.9 36.4 50.2 45.9
Deposits excluded from broad money 0.0 0.0 0.0 0.0 0.0 0.0
Securities other than shares excluded from broad money 1.2 1.5 2.3 1.6 11.3 10.9
Loans 12.3 17.4 30.4 53.5 56.7 73.2
Financial derivatives 1.6 2.7 2.7 4.1 4.2 5.5
Insurance technical reserves 0.0 0.0 0.0 0.0 0.0 0.0
Shares & other equity 269.0 308.3 366.2 429.4 507.0 611.7
Other items (Net) 91.1 122.2 213.5 231.3 187.4 454.5
Source: CBRT, BRSA, IFS.
Annex 11: Banking Sector Ratios
Selected Ratios for Banking Sector (end of period)
2015 2016 2017 2018 2019 2020
Liquidity Position
Liquidity Requirement Ratio 143.5 135.6 144.5 143.8 147.7 148.2
Loan-to-Deposit Ratio 123.4 123.6 126.6 122.6 109.6 108.3
Capital Adequacy
Core Capital Adequacy Ratio 13.3 13.2 14.1 13.8 14.2 14.5
Capital Adequacy Standard Ratio 15.6 15.6 16.8 17.3 18.4 18.7
Total Risk Weighted Assets (Net) / Total Risk Weighted Assets (Gross) 68.6 43.3 64.4 64.2 64.7 59.1
Regulatory Capital / Total Risk Weighted Assets 15.6 15.6 16.9 17.3 18.4 18.7
Profitability
Profit (Loss) Before Tax / Average Total Assets 1.5 1.9 2.0 1.8 1.4 1.4
Net Income / Average Total Assets 1.2 1.5 1.6 1.5 1.2 1.1
Net Income / Average Shareholder's Equity 11.3 14.3 15.9 14.8 11.5 11.4
Net Interest (Profit) Revenues (Expenses) / Average Total Assets 3.5 3.7 3.8 3.9 3.9 3.9
Asset Quality
Non-Performing Loans (Gross) / Total Cash Loans 3.1 3.2 3.0 3.9 5.4 4.1
Provision for Non-Performing Loans / Gross Non-Performing Loans 74.6 77.4 79.3 68.3 65.1 75.0
Credit Growth (FX-adjusted, y-o-y, in percent)* 10.7 10.2 18.2 1.0 8.5 37.5
Interest Rates (end-of-period)
Weighted average of Central Bank Cost of Funding 8.8 8.3 12.8 24.1 11.4 17.0
Weighted average Interest Rate for Deposits 10.7 9.6 12.4 22.2 10.6 15.1
Consumer Loans Rate 16.4 14.7 17.7 33.1 15.3 21.9
Commercial Loans Rate 15.7 14.3 17.1 28.3 13.5 19.0
Off Balance Sheet Transactions
Derivative Financial Instruments / Commitments 76.9 79.3 78.3 81.5 80.5 79.6
Source: CBRT, BRSA, IMF, WB Staff calculations.
*FX-adjusted credit growth is calculated using end-year average exchange rate value of US$/TRY since 2010 for adjustment.
World Bank
http://www.worldbank.org/tr
turkeywebfdbk@worldbank.org