50305 BRIEF Are Deposits a Stable Source of Funding for Microfinance Institutions? Poor people save. The conventional view is that low-income depositors transact more frequently than holders of larger accounts and are more prone to income disruptions from natural disasters, health issues, crime, and other factors. This perception makes financial institutions stepping into the under-served low-income space worry about whether they can use small deposits to fund their lending operations. But new research finds that under normal circumstances, aggregate balances for low-income accounts move gradually, and they are not prone to abrupt month-by month swings. This should make liquidity management easier because it gives the institutions enough time to adjust to changes in deposit supply over several months. Of course, financial institutions cannot take for Methodology granted that any of their deposits are stable, long-term sources of funds until they have Five financial institutions were studied: Allied carefully analyzed typical savings patterns in Bank (Pakistan), VTB (Georgia), BPR Kebomas their portfolio of deposit products. This analysis (Indonesia), Equity Bank (Kenya), and Banco informs their liquidity management and their Sol (Bolivia). They were chosen because they funding strategy. Deposit-taking MFIs should represented large markets of under-served use the same type of analysis on their deposit poor people, and all but one (BPR Indonesia) products, and refine their liquidity planning and maintained large deposit volumes from low- funding operations accordingly. income savers. Research study "Small balance" deposits were defined by analyzing account sizes by product and customer CGAP commissioned the Frankfurt School demographics, matched against the poverty line of Finance and Management to study the and per capita income in the country. stability of small deposits as a source of funds in five institutions.1 The study examined the Each institution's deposit base was analyzed in actual behavior of deposits raised from poor terms of long-term trend, seasonal patterns, individuals, and asked the following: annualized daily volatility, and average life of demand deposits, as well as peculiar patterns, 1. Are there recurring seasonal savings trend interruptions, and outlier values. Banks patterns? usually track these and other indicators for the 2. How volatile or predictable are the aggregate purpose of asset and liability management and balances of demand and term deposits? reporting. 3. How do external events (e.g., natural disasters, political turmoil, war) affect deposits? 1 The full report, The Stability of Small Balance Deposits, is available at www.cgap.org. In addition to data from a diverse group of five low- income deposit operations, the paper provides analytical tools that will help even small MFIs to start modeling deposit supply. June 2009 2 Small Balance Demand Deposits Provide performance of these deposits. The examples Relatively Stable Aggregate Balances from these five financial institutions show that neither savings nor term deposits are inherently Table 1 summarizes the main findings, along more or less volatile as products, but the specific with reference data from a bank in Germany. combination of these elements will determine The estimated volatility figures from the large in the end whether savings accounts or term German bank are in line with what conventional deposits are more volatile. wisdom would predict for deposit behavior in a country with fully developed financial markets. Deposit Balances Are Not Seasonal Current accounts are the most volatile: they are primarily used for transaction reasons, so The data show only weak seasonality in all five transactions are frequent. Savings deposits are institutions, which is good news for liquidity the bedrock of funding, with very low volatility management. The only marginally significant and low interest cost. Term deposits are more seasonal pattern was a recurring savings decline volatile than savings accounts because of their at the beginning of Ramadan at Allied Bank higher concentration, or "lumpiness," as they in Pakistan. However, money moved out of come and go in relatively large amounts. savings was often moved into current accounts or banker's checks, so there was little liquidity Equity Bank and Allied Bank conformed to this impact on the institution. expected pattern, but in VTB, BPR Kebomas, and BancoSol, ordinary savings accounts turned Deposits Are Not Very Sensitive out to be more volatile than term deposits. to External Macro Events This may have something to do with the fact that in the latter three institutions, aggregate The study found that natural disasters had no term deposits are several times larger than the effect on deposit balances in any of the institutions. total balance supplied by ordinary savings. This Deposits dropped during some sociopolitical suggests that bank staff pay particular attention crises in some countries, but certainly not all. to raising term deposits and cultivating the smaller circle of "elite" depositors, to manage the Even where there existed a potential match on volatility of term deposits. In many developing the timeline between a confidence-shattering countries there is anecdotal evidence that small stress event and a decline in deposits, there was traders use savings accounts instead of current no guarantee that there was indeed a causal accounts for short-term safekeeping of relatively link. There were always more undulations in the large amounts of cash that are deposited and deposit time series than there were stress events withdrawn several times a week. Such atypical that could have explained them. use of savings accounts may distort the volatility in institutions with a micro-enterprise focus, such In addition, the study notes that a positive macro as BPR Kebomas and BancoSol. In general, the event never leads to a sudden surge in deposits. stability or volatility of the deposits depends Mapping events to deposit patterns only makes on the size and frequency of transactions, sense for sudden bad news or disasters of a institutional reputation, and interest rates. broad-based nature that induce large numbers Also, the way the institution markets its savings of depositors to try to get their money out while products may help explain the unusual volatility they still can. 3 Table 1. Main Findings Allied Bank, VTB, BPR Kebomas, Equity Bank, BancoSol, For Comparison: Pakistan Georgia Indonesia Kenya Bolivia German Bank Type/charter Fully licensed Fully licensed Rural People's Fully licensed Fully licensed Fully licensed commercial commercial Credit Bank commercial commercial commercial bank bank bank (BPR) with bank bank limited license for savings and loan operations Supervised/ State Bank of National Bank Bank Indonesia Central Bank of Superintendent German regulated by Pakistan of Georgia Kenya of Banks and Financial Markets Financial Regulator & Entities & Bundesbank/ Central Bank European Central of Bolivia Bank Total assets, Dec-2006, $4.2 billion $275 million $641,000 $294 million $231 million Not disclosed US$ Customer deposits/ 81.7% 28.3% 59.0% 82.0% 66.4% 2002: 35% total assets GNI per capita (Atlas $690 $1,350 $1,280 $530 $1,010 $34,580 method) Consumer price inflation 7.9% 10.0% 13.2% 10.5% 4.3% 1.7% % p.a. 2006 % of population below 17.0% 6.5% 7.5% 22.8% 23.2% n/a $1/day % of population below 73.6% 25.3% 52.4% 58.3% 42.2% n/a $2/day Monthly product Jan-96 to Jan-01 to Jan-01 to Nov-01 to Dec-98 to Sep-97 to balances Jul-04 May-04 May-04 Dec-06 Dec-05 Dec-02 Weekly product balances Aug-04 to Jun-04 to Jun-04 to None None None Oct-05 May-05 May-05 Daily product balances Nov-05 to Jun-05 to Jun-05 to None None None Nov-06 May-06 May-06 Demand savings Some residual Insignificant Insignificant Insignificant Insignificant Insignificant accounts annual-cycle seasonality Current accounts on all three n/a n/a n/a n/a Insignificant Retail term deposits deposit Insignificant Insignificant Not tested Not tested Insignificant categories due to Islamic holidays and Zakat effect Demand savings 13.51% 27.6% (GEL 61.24% 14.43% 24.24% 4.39% accounts accts. only) Current accounts 40.89% n/a n/a 1.2012 n/a 0.1581 Retail term deposits 33.57% 6.73% 26.96% 34.50% 17.33% 11.88% Political or Yes, several Yes, several Yes, several Only minor Yes, several None, but special socioeconomic crisis political circumstances during data period? turbulence: during euro Dec-02 general introduction elections, Jan-02 constitutional referendum Dec-05 Large natural disaster Yes, Kashmir No Yes, Dec-04 Yes, drought in None Yes, River Elbe during data period? earthquake Tsunami 2005 flood Aug-02 Oct-05 Macro events coincide Yes, May- Yes, Nov-03 to Yes, May-01 No evidence Yes, Jul-02 No, however, with decline in deposit Oct-99 Kargil Mar-04 Rose to Jul-01 found in Equity liquidity crisis, surge of deposits supply? (Kashmir) Revolution President Bank data Feb-03 unrest in Dec-01 ahead conflict; 9/11 Wahid against Lozado of DM-Euro & Afghanistan impeachment govt, Apr-04 banknote invasion Oct-01 crisis; Jul-04 tax on financial conversion in to Sep-04 transactions Jan-02 presidential and gas elections referendum, Dec-05 presidential elections June 2009 Conclusion base to determine what proportion of it can safely be used for long-term loans. The analysis The study shows that some portion of small should be done by broad product categories All CGAP publications balance deposits in any institution can be (e.g., current accounts, demand savings, and are available on the considered a stable source of funds. time deposits) and should include at least three CGAP Web site at www.cgap.org. years of monthly and one year of daily data Banks and microfinance institutions should points to capture long-run trends, seasonal CGAP invest in analyzing their particular depositor effects and daily volatility patterns. 1818 H Street, NW MSN P3-300 Washington, DC 20433 USA Tel: 202-473-9594 Fax: 202-522-3744 Email: cgap@worldbank.org © CGAP, 2009 AUTHORS Julia Abakaeva and Jasmina Glisovic-Mezieres