Exercising Ownership Rights in State Owned Enterprise Groups: What China Can Learn from International Experience William Mako and Chunlin Zhang December 31, 2002 An enterprise group in China is defined by government regulation as "a combination of legally independent enterprises" made up of a parent company, subsidiary companies in which the parent has majority ownership, member companies in which the parent has minority ownership, and other member enterprises or institutions.1 At the end of 2000, there were 6027 registered enterprise groups2, in which 2655 are considered as "large enterprise groups", defined as those whose sales revenue and total assets exceed RMB500 million. In 2000, these large enterprise groups accounted for 11% of China's urban employment, 22% of exports, and 57% of assets in the industrial sector3 . The authorities expect enterprise groups to play an increasingly critical role in development of the Chinese economy in a post-WTO environment. The Communist Party Central Committee "4th Plenum" on SOE reform in 1999 called for "fostering strong and competitive large enterprises and groups" so that they can become "the pillars of national economy and major forces of China in participating in international competition". The Tenth Five-Year Plan approved by the National People's Congress (NPC) in 2001 also requires "development of a number of large companies and enterprise groups with well- known brand names, independent intellectual property rights, clearly-defined and strong core businesses". State ownership dominates large enterprise groups in China. Of the 2655 large enterprise groups, the state is majority owner of the parent of 1605. These tend to be the largest of the large. State-majority large enterprise groups account for 92% of the assets and 87% of sales for all large enterprise groups. While the state is moving to "withdraw" from non-strategic enterprises, significant state ownership of enterprise groups is likely to continue in the short to medium term. At the end of 2000, China still had over 190,000 state-owned enterprises. For the foreseeable future, deepening reforms in large state owned enterprise groups, for example, in the fronts of internal control and corporate governance, will depends critically on the way the state exercises its ownership rights, i.e., the reform of state assets management system. 1A legally independent enterprise in China can be a company registered under the Company Law, or a state owned enterprise registered under the All People Owned Enterprise Law (enacted in 1988), or other enterprises that are independent in terms of responsibility for accounting and debt re-payment, such as collectively owned enterprises, joint ventures, partnerships, and sole proprietorships. 2State Administration for Industry and Commerce: China Industrial and Commerce Administration Yearbook, 2001. 3NBS: China's Large Enterprises Groups in 2000, and China Statistical Yearbook 2001. 1 This note addresses the question of how the Chinese state could exercise its ownership rights in large state owned enterprises in a more effective way, by drawing lessons from relevant international experiences. Since large state owned enterprise groups share the same institutional framework of state assets management system with other SOEs, the policy recommendations of this note may have broader applicability. The rest of this note is organized as the follows. In section I, the institutional challenges of exercising state ownership rights in SOEs in China and past efforts of reform are highlighted. Section II presents relevant international experiences. Key policy recommendations appear in section III. 2 I. The Institutional Challenges China's SOE Sector: an Overview In 2000, China's 190,508 non-financial state owned enterprises (SOEs) produced roughly 1/3 of its GDP. Their balance sheets recorded over RMB16 trillion assets, with slightly less than RMB10 trillion liabilities. The total accounting value of state owner's equities was reported as RMB5.8 trillion by the Ministry of Finance. 49% of the SOEs were profitable, and reported a total profit of RMB 0.47 trillion in 2000, while 51% of them made a total loss of RMB 0.18 trillion, leaving a net profit of RMB 0.28 trillion to the state, which is 1.5 times higher than 1999 and 1.2 times higher than 1995. 52.7 million people were employed by these SOEs, accounting for 7.4% of total employment and 25% of urban employment. SOEs are classified as of large, medium or small size. In 1999, 72% of the state owner's equities was found in 9000 large SOEs, while 31000 medium and 177000 small ones shared the rest (12% and 16%, respectively). The Pre-Reform Regime Given the size and diversity of the SOE sector, the Chinese state has developed a system of division of labor in exercising ownership rights in SOEs. First, ownership rights are exercised by every level of government, from the national government to provincial, municipal and county governments. Second, ownership rights are shared by a wide range of government and Party agencies. This division of labor is reflected by the concept of "subordinate relation (li shu guang xi)". In the pre-reform regime, a SOE being "subordinated (li shu yu)" or "supervised (zhu guan)" by a particular level of government would typically mean that the government and party committee at this level exercised the following rights: · Appoint, monitor, motivate and replace managers. This was typically done by party committee and its personnel department at the level of government that was supposed to supervise the SOE in question4. In line with the political nature of SOEs, their managers were "administrated" as party cadres in the same way as other officials serving in the government bureaucracy, and were ranked in the same system. It was indeed quite common for a SOE manager to be appointed in a government position or the other way around. 4When a high rank manager was appointed to a SOE that is supervised by a lower level of government, this particular manager could be "administrated" by the higher level of party committee and personnel department. For example, a minister level manager could be appointed to a provincially supervised SOE. In such cases, this particular manager had to be administrated by the higher authority, the central party committee and personnel department, but other lower rank managers would follow the usual pattern. 3 · Make key business decisions. This was performed by economic departments of the level of government involved. For example, the planning commission would be responsible to formulate overall annual and Five-Year plans, which specified anything from budget funds allocation and investments to prices and quantities of key products of SOEs under its supervision. Planning department was also charged with the authorities to approve capital construction investment projects, or make recommendations to the higher-level government in cases of large investments. The economic commission in collaboration with line bureaus (ministries) was responsible for day-to-day management of SOE businesses. Labor and personnel departments held the power to manage human resources related issues, such as recruitment, wages and salaries, social benefits, etc.. · Provide finance and claim revenue. Financial department of the level of government was responsible to collect revenues from and allocate funds to SOEs. Enterprise revenue, in forms of taxes and profit remittance, accounted for roughly ½ of budget revenue in pre-reform China. Related to this function, it was also responsible to monitoring financial accounting and reporting of SOEs. State owned banks were responsible to collect deposits and provide working capital to SOEs. Bank loans were not an important source of finance for SOEs before the reform. Past Reform Efforts The SOE reform started in late 1970s has experimented numerous approaches and strategies, initiated by the central government as well as local governments at every level and managers and workers of SOEs. The relationship between the state and SOEs has been the central focus, and most reform measures were designed to change the way the state exercises its ownership rights in SOEs. Conceptually, these reform efforts can be viewed as having followed two strategies, which are often mixed with each other in specific policies and therefore less visible. The first strategy prevailed during 1978-93 and has remained influential since then. It has two pillars: empowering insiders to improve efficiency on one hand, and strengthening party-government control to contain insider control on the other hand. Managerial autonomy and earning sharing between the state and insiders, i.e., managers and employees, were introduced in an attempt to strengthen incentives and raise efficiency. This empowered insiders and inevitably led to significant degree of insider control with all the negative consequences such as assets stripping. To get insider control under control, the role of party and state organs in appointing, monitoring and supervising SOE managers was strengthened. The second strategy was officially adopted in the 1993 "3rd Plenum Decisions", which states that "building modern enterprise system is the direction of SOE reform" and calls for "institutional innovation" to replace the previous strategy of "delegating power and conceding profit". The 15th CCP National Congress in 1997 and the 4th Plenum of the CCP 15th Central Committee in 1999 reinforced the strategy by advocating the policies of "grasp the large and let-go the small", "adjust the layout of the state sector with a strategic view", and "withdraw from where the state should withdraw". This strategy is 4 essentially one of corporatization and ownership diversification, expressed in different official jargons in different periods though. Since neither corporatization nor diversification can be possible without reform of the state assets management system, the three tier model of state assets management reform deserves a closer look. The Three Tier Model Corporatization as a potential way of reforming SOEs started to be discussed and experimented in mid-1980s5. By 1993 when the 3rd Plenum of the 14th CCP Central Committee was held, there had been more than 3000 SOEs corporatized one way or another. One of the most important ideas behind this strategy of corporatization was to reform the institutional framework in which the state exercises its ownership rights by redefining its role in enterprises as a pure shareholder, whose rights and obligations are specified by law. To this end, a three-tiers model was developed and widely accepted by policy makers and advisors. The model was characterized by two separations: separation of the ownership function of the state from its other function, and separation of the ownership function from day-to-day operation of state assets. According to some common understanding of this model, the first tier consists of one state agency, under the State Council or even the NPC (the government or the local People's Congress in local levels), charged with the exclusive authority to act as the owner of state owned assets. Ownership function is thus separated from other functions of the state as all other party and government organs are supposed to transfer their ownership functions to this agency. However, this first tier agency is a government organ staffed with civil servants, rather than commercial units. The second tier institutions are designed to perform day-to-day management function of the state assets. These institutions are designed as commercial entities maximizing return on state assets, and most likely wholly owned by the first tier agency. They are supposed to exercise shareholder rights in corporatized SOEs who are found on the third tier, by voting in the shareholders assembly or sitting on their boards. In the Chinese Company Law, this kind of institutions are called "state authorized investment institutions" 5The World Bank played a role in advocating this strategy, according to some Chinese reformers such as Wu Jinglian, a leading advisor to the government on reform strategies and policies. In a report published in 1985, the Bank recommended that China may wish to delegate decision making power to a board of directors instead of workers and managers, as was happening at that time. And in order to isolate the board from political intervention as much as possible, a "socialist joint stock ownership" was suggested, which meant to "spread the ownership of each state enterprise among several different institutions", such as banks, pension funds, insurance companies, and other enterprises, in addition to central and local governments. In 1988, a Bank mission visited a number of major industrial cities in China and held extensive discussions with Chinese officials and academics in a workshop in November 1988. One of the key messages of this workshop and a subsequent unpublished report was that "the contracting responsibility system in China should gradually be phased out, and wherever appropriate it should be substituted by the share company system". See Wu Jinglian, Modern Company and Enterprise Reform, Tianjin People's Publishing House, 1994, p220-21; p234-38. The World Bank, China: Long-Term Development Issues and Options. The John Hopkins University Press, 1985, p166. The World Bank, 1989. China: Enterprise Management Reform, Issues and Options. Unpublished report. Report No. 7773-CHA. 5 Despite its strong influence in reform practices, this model has never been formally accepted by the central government as part of its SOE reform strategy. In particular, the idea of establishing the first tier agency has not been tried at the central government level. In terms of the second tier institutions, however, the central government did take some relevant actions. Many parent companies of SOE groups were "authorized (shou quan)" one way or another to act as the representative of state ownership in its subsidiaries. In mid-1990s, three parent companies of large industrial groups were selected for a pilot implementation of the idea of state holding company6 , which was later substituted by government-led reorganizations in the three relevant sectors before producing any useful results. A number of local governments went farther by setting up a state assets management committee (SOAMC) on the first tier. Shanghai and Shenzhen were among the most advanced. However, the SOAMC was only "virtual", as is usually called in China. First of all, the members of the committee are all current heads of party and government departments, instead of full time independent professionals. As a result, it has become more and more ceremonial. The gap has been filled by all kinds of party and government organs. In one case, eight party and government bureaus in charge of particular sectors, the Personnel Department of Party Committee, the Auditing Bureau, Supervision Bureau, Municipal Trade Union, and the office of SOAMC all have a role to play in exercising ownership rights in the second tier and third tier companies. The three-tier model has its own logic. Without a separation of ownership function from other functions of the state on the first tier, the second tier institutions can hardly differ from any other SOEs in terms of their relationship with the government. Not surprisingly, failure of the reform on the first tier has led to failure of commercialization of the second tier. In some places, the second tier institutions are known as "funnel (lou dou)", " setter (er chaun shou)" of the government, or "quasi-government". Their behavior is widely questioned. Some of these institutions see themselves performing multiple functions including a warehouse of non-performing assets and a guardian of social stability. Some invested in obviously non-profitable projects and enterprises in response to request from the government departments; some created "directly owned subsidiaries" and provided preferential treatment to them at the expense of other subsidiaries; some collected 30% of after tax profits of subsidiaries and could not cover their administrative costs. The fact that the second tier institutions have failed to become professional owners generated complaints from the third tier companies, which led to a return to the first strategy discussed earlier. Shenzhen government has implemented what is called "authorized operation" in some better-performing large groups, in which the government bypassed the second tier and signed performance contract directly with the management of each group. This is of course only feasible in wholly state owned companies, as was the contract responsibility system in the 1980s. Exercising Ownership Rights in SOEs: the Nature of the Challenge 6Involving China Petro-chemical Indutrisal General Corporation (Sino-pec), China Aviation Industrial General Corporation, and China Non-ferro Matel Industrial General Corporation. 6 The fact appears to be that China has spent more than two decades to reform the state assets management system and achieved only very limited success. None of the two strategies could turn SOEs into fully commercialized business entities that fits the notion of a modern market economy. Ownership rights are still exercised in a similar way as two decades ago in terms of institutional framework. The most significant change is probably the fact that some control rights have been transferred from the state to SOE management. Overall, the relationship between the state and SOEs is still such that it does not fit the requirement of a modern market economy. It is critical to ask the question of why this has been the case. In retrospective, one important reason seems to be that the reform in the past two decades have almost exclusively concentrated on the relationship between the state and SOEs, and did very little in terms of the relationship between the original owner, the people as a whole, and the state. This is important because the existing status of the relationship between the people and the state represents a soft budget constraint to the state. Legally speaking, the Chinese people as a whole is the ultimate owner of SOEs, and the state is supposed to manage them on behalf of the people. In this sense, the people-state relationship is comparable with one between the owners of a fund and the fund manager. However, the state faces a much more soft constraint than a fund manager. Mostly notably, there is no contract specifying the obligation of the state in terms of financial return to the assets owned by the people, as is found in the case of commercial assets management. This soft budget constraint has allowed the state to ignore the cost of state capital and pursue objectives other than maximization of the return on state capital. So long as the state is free to use its ownership function to achieve goals other than maximization of financial returns, two results follow naturally. First, SOEs are inevitably driven by the state to pursue multiple objectives as well, unless the state loses its control over certain SOEs. Second, it is probably the best arrangement for the ownership rights to be shared by various government and party agencies, which perform other functions of the state, because in this way, the ownership function is fully utilized to serve goals of other functions. The implication is clear: until the ownership function of the state is separated from its other functions and fully commercialized, i.e., maximization of the financial return on state capital is taken as the primary objective of the state as owner, SOEs are not going to become fully commercialized business entities that meet the requirement of a modern market economy integrated with the world market. The fundamental challenge to the next stage of SOE reform is how to make this happen. 7 II. International Experience Thus, China's current asset management system consists of three tiers: individual state- owned enterprises (SOEs) on the bottom; a "first tier" of central, provincial, or municipal entities (e.g., state asset management committees) that seek to exercise SOE governance on behalf of the State; and, in many cases, intermediate "second tier" entities (e.g., state asset management bureaus, the four asset management companies, or the parent companies of enterprise groups) that may also hold shares in individual SOEs and be in a position to exercise governance. In order to provide some perspective on China's three- tier framework and develop lessons for exercising the State's ownership in large enterprise groups, this section synthesizes case studies on reasonably well-run SOEs ­ in Sweden, Norway, Great Britain, Germany, France, Italy, Austria, New Zealand, Brazil, Singapore, and South Korea. This section discusses the following topics: · Manageable size of the state sector; · Appropriate goals for SOEs; · "First tier" state share ownership ­ i.e., principles, organization, and procedures; · Legal status and governance of SOEs; and · "Second tier" state share-holding institutions ­ i.e., special issues for asset management bureaus, parent companies of enterprise groups, and asset management companies (AMCs) created to resolve non-performing loans. A. Size of the State Sector Well-run state enterprise sectors in other countries typically represent a smaller portion of the national economy. The usual focus is on large enterprises that may be of strategic importance to the national economy. · According to one estimate, the value of Sweden's SOEs grew by more than 12 percent during 1998-2001, while the Stockholm Exchange's overall market capitalization showed just 6 percent growth. Sweden has just 59 SOEs, which account for about 7 percent of GDP and 5 percent of total employment. · Singapore's Temasek financial holding company owns shares in 20 large "government linked corporations" (GLCs), which represent about 12 percent of Singapore's GDP. Of these, twelve are listed on the Singapore Exchange (SGX) and account for about 20 percent of market capitalization. · New Zealand's 16 SOEs account for less than 12 percent of GDP. · South Korea's public enterprises accounted for 9.4 percent of GDP and 2.5 percent of employment in 1990. These included just 13 commercially-oriented, non-financial firms. As of 2001, South Korea's remaining SOEs had fewer subsidiaries (i.e., a narrower business focus), lower debt/equity, and higher returns on assets and equity than privately-owned business groups. By reducing multiple requirements on the state shareholder ­ e.g, reviewing business strategy and plans, appointing directors and top management, monitoring financial results ­ a smaller state sector is conducive to improved governance of remaining SOEs. 8 Some governments are content with majority-private minority-state ownership of strategic enterprises, so long as they retain an ability to block adverse developments. Austria. The government has sought to achieve market supervision and financing of former state-majority enterprises while retaining strategic influence through continued minority shareholdings by a state-owned fund, the Osterreichen Industrieverwaltungs Aktien Gemeinshaft (OIAG). OIAG's original mission, during the 1970s and 1980s, was to insulate SOEs from politics and enhance business flexibility; provide continuity in SOE supervision and strategy; end soft budget constraints; and enhance efficiency by centralizing the treasury function and other corporate services. This changed with a 1993 law requiring OIAG, in response to economic crises and mushrooming OIAG debt, to sell its majority holdings in Austrian SOEs and use the proceeds to pay off its debt. The law stipulated, however, that OIAG's shareholdings should not go below the level ­ 25 percent plus 1 share ­ needed under company law to block fundamental changes in a company's business or capital structure (see Table II-1). Since Austrian pension funds are not large shareholders and Austria's corporate laws do not give the state a "golden share" that could be used to veto a hostile takeover, the government has relied on OIAG's minority blocking shareholdings to ensure that strategic decisions on enterprises important to the national economy are taken within Austria.7 Table II-1. Austria: Shareholdings By State-Owned OIAG Enterprise (Sector) Year Shareholding VA Stahl (steel) 1993 75% 1995 43% 2000 35% OMV (oil refining/distribution) 1993 75% 1997 35% Austrian Airlines 1997 52% 2000 40% Telekom Austria 1997 100% 1998 75% 2000 47% Boehler Udenholm (steel) 1994 73% 1996 25% Source: OIAG, Butzbach South Korea. The government has reduced state shares in major SOEs, especially in 1988-89 and since 1997 (see Table II-2), but remains able to protect national strategic interests through various means. Korea Electric Power Company (KEPCO) and Korea 7Survey: Austria 99, Financial Times. 9 Telecom are both regulated businesses. In addition, while the Ministry of Finance and Economy (MOFE) shareholdings in KEPCO have fallen to 32 percent, the state-owned Korea Development Bank still holds about 22 percent ­ for a combined stake of 54 percent. Successive governments successively reduced MOFE shares in POSCO between 1987 and 1999. During this time, however, the government was able to protect national strategic interests through law. Under an enabling decree of the Securities and Exchange Act, the Minister of Finance and Economy has the authority to designate as a "public nature corporation" any corporation engaging "in an important industry for the national economy" so long as the state retains at least 15 percent of the corporation's shares. For any such "public nature corporation," the Act restricts proxy voting and allows the corporation's articles of incorporation to cap the voting rights of other individual shareholders at 3 percent. 10 Table II-2. Major SOEs in South Korea: Changes in State Shareholdings, 1988-2002 Enterprise/Date Gov't Gov't Other Foreign 1st tier 2nd tier Domestic KEPCO: Dec-88 100.00 - - Dec-89 77.83 22.17 - Dec-97 69.82 5.72 * 13.90 10.56 Dec-99 52.23 2.73 * 22.69 22.35 Dec-01 32.35 21.57 * 19.55 26.53 Korea Telecom: Dec-92 100.00 - - - Dec-98 71.20 - 23.80 5.00 Dec-99 58.99 - 22.30 18.71 Dec-01 40.15 - 22.65 37.20 May-02 - - 51.00 49.00 POSCO: Dec-68 56.20 43.80 ** - - Dec-87 32.30 67.70 ** - - Jun-88 20.00 42.70 ** 37.30 - Dec-92 20.00 30.10 ** 41.90 8.00 Dec-97 19.60 22.60 ** 32.70 25.10 Dec-99 - 17.86 ** 39.16 42.98 Dec-01 - 9.97 ** 28.02 62.01 Hanjung Heavy Engineering: Dec-97 - 100.00 *** - - Dec-00 - 76.00 *** 24.00 - Dec-01 - 28.34 *** 71.09 0.57 Source: Lim (2002). * Shares held by Korea Development Bank. ** Shares held by Korea Development Bank and other state banks, Korea Tungsten, and Pohang University of Science and Technology. *** Shares held by Korea Development Bank and KEPCO. 11 B. Goals for SOEs The best SOEs in the world focus on financial performance, especially on returns on capital. Sweden. The SOE shareholder ­ typically the Ministry of Industry ­ expects each SOE to do more than earn a profit based on standard financial accounting. Each SOE is expected to provide "economic value-added" (EVA), by earning more than its cost of debt and equity capital (see Box II-1 for summary). In Sweden, the emphasis on EVA has been endorsed by the labor union representatives of SOE workers on the assumption that there is only room for efficient companies ­ either private or state-owned ­ in a globally competitive market. The state shareholder has encouraged SOE boards to introduce value-based systems, such as EVA, to communicate the importance and effects of an efficient capital structure and to link employee and management compensation to EVA results. According to one estimate, the value of Swedish SOEs grew by more than 12 percent during 1998-2001, while the Stockholm Exchange's overall market capitalization showed just 6 percent growth (Detter, 2002) Box II-1. Economic Value-Added (EVA) Academic research shows that accounting measures (e.g., earnings per share) are only coincidentally related to stock prices. More significant is the cash, adjusted for time and risk, that investors can expect to get back over the life of the business. This raises the question of how to link discounted cash flow ­ which is the most analytically respectable approach to valuation ­ with actual financial management of the enterprise? Management should focus on maximizing economic value-added (EVA), which is operating profits less the cost of all the capital employed to produce those earnings. EVA will increase if operating profits can be made to grow without tying up any more capital, if new capital can be invested in projects that will earn more than the full cost of the capital, and if capital can be diverted or liquidated from business activities that do not provide adequate returns. EVA is the only performance measure that is entirely consistent with the standard capital budgeting rule: Accept all positive and reject all negative present value investments. (Earnings per share, on the other hand, will increase so long as new capital investments earn anything more than the after-tax cost of borrowing.) The rate of return on total capital is the return that should be used to assess corporate performance. The required rate of return for a particular enterprise should increase with the enterprise's operational risk and financial risk. Performance should be measured against total capital ­ both debt and equity. Net operating profits in excess of total capital times required rate of return is considered economic value added (EVA). Stewart, 1994. 12 Norway. Soon after Statoil's founding in 1972, the state shareholder established commercial targets for the company. The shareholder's decision in 1999 to proceed with an initial public offering (IPO) of 19 percent of Statoil's shares ­ in order to induce continuous market evaluation, facilitate benchmarking vis-à-vis competitors, and sharpen the distinction between enterprise governance and management ­ coincided with efforts to reduce costs of labor and capital. Labor reductions were agreed with the enterprise's unions and, as part of its 2001 IPO, Statoil's equity was reduced by 24 percent and capital was returned to the ministry of finance (MOF). This return of capital was financed through the issuance of new equity to public shareholders, transfers of cash and assets to the MOF, and additional debt. Singapore. Temasek expects those "government-linked corporations" (GLCs) in which it holds shares to (i) be world class and compete internationally, in order to attract talent; (ii) have a high-quality board; (iii) focus on core competencies; (iv) pay competitive wages; and (v) maximize financial performance in terms of EVA, return on assets (ROA), and return on equity (ROE). GLC operations may be benchmarked to international standards. To encourage each GLC to focus on core competencies, any diversification must be agreed beforehand by Temasek. Singapore GLCs compare favorably, in terms of returns on equity, with international private-sector peers (see Table II-3). Table II-3. Returns on Equity: Singapore GLCs vs. International Private-Sector Peers Sector/Company Temasek 2001 2000 1999 Ownership Telecoms: Singtel 68% 24.9% 30.3% 25.6% AT&T - 13.5% 4.5% 4.3% British Telecom - 15.5% 14.4% 22.6% Banking: DBS 13% 7.4% 13.2% 9.9% UOB - 7.3%% 13.5% 12.3% OCBC - 9.0% 10.3% 9.5% Airlines: SIA 57% 16.3% 10.6% n.a. British - 3.2% -0.7% n.a. Source: Tay and Ma South Korea. With its Public Enterprise Reform Act of 1983, South Korea began systematic efforts to improve performance at "government invested enterprises" (GIEs). Performance monitoring and incentives emphasized quantitative productivity measures (e.g., public profit divided by fixed operating capital); enterprise-specific technical indicators of operating efficiency; and qualitative indicators of performance in such areas as administration, research and development, and service. Initially, standard financial ratios ­ such as returns on equity ­ were not considered appropriate for public enterprises. In recent years, however, the government has increasingly included 13 economic value added (EVA) as a performance indicator. Among Korea's thirty largest enterprises in 2001, both state-owned and private, returns on equity averaged 5.2 percent at the state-owned versus 4.1 percent at the private. Financial and other business goals are sometimes supplemented by explicit statements of social goals. New Zealand is a good example. Under the SOE act of 1986, SOEs should meet three goals: (i) be as profitable and efficient as comparable businesses in the private sector; (ii) be a "good employer;" and (3) show social responsibility by paying attention to local communities' needs and interests. An annual Statement of Corporate Intent (SCI) is used to guide and monitor each SOE in New Zealand. In order to ensure that New Zealand SOEs achieve their objectives of profitability, efficiency, and social responsibility, the government established a framework that included state monitoring, control, and incentives. The two state shareholders ­ i.e., the ministry of finance and relevant line ministry ­ monitor and control the results and operations of the SOE and verify its compliance with an annual Statement of Corporate Intent (SCI) that covers the following: · Nature and scope of business activities to be pursued by the SOE; · Performance targets for the next three years; · Level of dividends to be paid; · Employees' safety, compensation, job security, and equal opportunity; · Adherence to long-term contracts with suppliers of goods and services; · Compliance with legal and regulatory obligations; and · Compliance with regulatory obligations Excessive SOE emphasis on social goals, however ­ such as the rescue of distressed companies or development of lagging regions ­ can destroy value for the state shareholder. The dangers from mixing business and social goals are clear in the case of the former state holding company in Italy. L'Instituto per la Ricostruzione Industriale (IRI) was set up in 1933 to take over heavily indebted companies in all sectors. After World War II, IRI played a key role in financing highways, telephone networks, steel, electricity, and engineering companies. In the 1960s and 1970s, IRI was assigned two non-commercial missions: (i) to rescue ailing companies, especially those hurt by oil shocks; and (ii) to dedicate 40 percent of its investment to developing Italy's poor southern region. IRI's management also became increasingly politicized during the 1970s, with management and board appointments, investments, and restructuring strategies fought over by political parties. By 1992, IRI had majority shares in 550 companies and minority shares in 280 companies. IRI's losses grew from $540 million in 1991 to $3.4 billion in 1992 and $8.3 billion in 1993. IRI's consolidated debt grew to $58.3 billion by 1992. C. "First Tier" Shareholding Institutional frameworks for well-run SOEs typically distinguish between the rights and responsibilities of the government to regulate economic activity and control the 14 disposition of state capital and those of the SOE's directors and management toward the SOE itself. · In Singapore, the Ministry of Finance ­ which is 100 percent owner ­ plays a minimal role in Temasek. The ministry appoints the chairman and members of Temasek's board. Every year, Temasek submits audited financial statements to the ministry for review. From time to time, the minister may ask for meetings with Temasek or its GLCs to discuss performance and plans. Otherwise, the ministry of finance gets involved only when an issue ­ e.g., acquisition, sale, or merger ­ affects Temasek's shareholding in a GLC. · In New Zealand, the relationship between the SOE's board and the state is regulated by the SOE Act, while the relationship between the SOE's board and its management follows commercial practice. SOE directors are legally bound to act in the SOE's interest. They report to the two shareholding ministers, who are themselves accountable to parliament. The board appoints the CEO and senior management of the SOE, determines strategy, takes decisions on large investments and dividend changes, ensures that the Statement of Corporate Intent complies with existing regulations, sets management compensation, and approves financial statements. · In the case of Statoil, the Norwegian government acts both as regulator and owner. By law, the state owns all oil resources and the Minister of Energy ensures that exploration and production conform with official resource policies. Legally, however, the state has no power over Statoil other than the shareholder rights it exercises at annual general meetings. By law, the directors and chief executive officer (CEO) of Statoil have a fiduciary duty to the company and all shareholders and are individually liable for any damage caused to the company. · In Sweden, governance of SOEs reflects the division of labor set out in the Swedish Companies Act, with the board responsible for providing oversight and guidance and the managing director in charge of day-to-day administration. Some matters, such as company closures and dividend policy, are considered part of normal administration and within the prerogatives of the government and ­ by extension ­ the SOE board. Other matters ­ such as changes in SOE ownership or capital ­ would be considered a disposition of state property and require consultation between the government and parliament. Thus there is typically a clear distinction between the government's regulatory rights and shareholding rights and a concentration of shareholder rights and responsibilities. The organization of "first tier" state shareholding, however, may vary. In Singapore, for example, the Ministry of Finance is the 100 percent owner in Temasek and relies on Temasek to manage state shares in GLCs. In New Zealand, state shares are divided between the Ministry of Finance and a sector ministry. In South Korea, the treasury bureau of the Ministry of Finance and Economy holds the state's shares in all GIEs. But the authority to exercise the government's shareholder rights ­ in consultation with the MOFE ­ may be delegated to a supervisory ministry, e.g., commerce, industry, and energy; construction and tourism. 15 Institutions for managing state equity may be modest and efficient ­ especially if the state shareholder implements effective reporting and controls, emphasizes strong boards for its SOEs, and follows best commercial practices. Sweden's 59 SOEs are estimated to represent about 7 percent of GDP and 5 percent of total employment. The 31 largest SOEs ­ including holdings in energy, telecoms, armaments, pulp and paper, pharmaceuticals, and transport ­ are administered by the Ministry of Industry, Employment, and Communications (MoI). State shares in these 31 SOEs are administered by the MoI's State-Owned Enterprises Division. This Division has a staff of 13, including 7 senior investment managers, 3 analysts, 2 assistants, and a director. Responsibility among senior investment managers is divided according to industry sectors in order to encourage focus and understanding of each holding. For example, the senior investment manager responsible for transport holdings maintains contacts with relevant bankers and consultants, tracks global competitors, and is MOI's main liaison with management of the transport SOEs, e.g., the airline SAS and the rail company SJ AB. Senior investment managers also represent the state shareholder as a non-executive director on SOE boards. The state shareholder completely revamped the boards and management of Swedish SOEs during 1998-2001. More than 85 percent of incumbent non-executive directors, 75 percent of board chairmen, and almost half of SOE CEOs were replaced during the period. The objective was to improve SOE performance by empowering the Non- Executive Board. By Swedish law, the Non-Executive Board is only institution responsible and accountable for development of a corporation ­ either private or state- owned. To revitalize the Non-Executive Boards of its SOEs, the state shareholder took a number of steps to upgrade board appointments and procedural rules for the Non- Executive Boards of SOEs (see Box II-10). Singapore. Temasek is a "second" rather than a "first tier" shareholder. It is worth noting, however, that Temasek's portfolio management infrastructure ­ especially relative to the market value of its portfolio ­ is quite modest. The Temasek financial holding company holds shares in over 20 major GLCs, 12 of which are listed on the Singapore Exchange (SGX). Temasek GLCs represent about 20 percent of SGX market capitalization. Major Temasek companies account for about 12 percent of Singapore's GDP. It is estimated that Temasek manages about $55 billion of market capitalization at an annual operating cost of less than $30 million ­ i.e., a 0.05% expense ratio. Temasek's Corporate Stewardship Division, which manages Temasek's GLC portfolio, has a staff of just 53. Temasek is able to get by with such a modest administrative infrastructure by focusing on GLC board appointments and ensuring that its GLCs have high-quality boards. New Zealand's Crown Companies Monitoring Advisory Unit (CCMAU) supports the two shareholding ministries in overseeing each of New Zealand's 16 SOEs. Set up in 1993, the CCMAU is owned by and administratively linked to the Treasury ­ for example, the Secretary of the Treasury appoints the CCMAU's executive director. But the CCMAU is operationally independent and functions like a private consulting business. It has 16 separate purchase agreements with each shareholding ministry to consult with the ministry of SOE issues. Many of the CCMAU's consultations are oriented toward balancing SOE profit maximization goals against broader social goals (see Box II-12). Box II-2. Outsourcing SOE Oversight: New Zealand's Crown Companies Monitoring Advisory Unit (CCMAU) The CCMAU prepares an annual outlook letter, which is then sent by the ministry to each SOE, detailing information required for business planning. The CCMAU reviews SOE Statements of Corporate Intent (SCIs) and provides comments to the shareholding ministries. The CCMAU also monitors SOE performance. To balance tensions between SOE profit maximization goals and state interests, the CCMAU advises shareholding ministries on SOE objectives; board performance and composition; definition of core businesses; capital structure and dividend policy; commercial assessment of business activity; performance measurement targets; actual performance relative to targets; and the impact of government policies on the SOE's value. More particularly, the CCMAU provides shareholding ministers with comments as to whether a board's proposed performance targets are consistent with relevant statutory requirements; the robustness of the board's proposed strategy and plans for achieving its proposed targets; whether the board's proposals on strategic issues (e.g., scope of business, balance sheet, dividend levels) serve the best interests of shareholders; whether the board is meeting its performance targets; and, if not, what action the shareholding ministries should take to hold the SOE board accountable. The CCMAU also participates actively in the appointment process for SOE directors, including commentary on skills needed for each SOE board. The CCMAU monitors board vacancies and conducts searches and develops short-lists of board candidates, for final approval by the shareholding ministers. Source: Butzbach. Austria. OIAG currently has about 50 staff. This group manages a portfolio of ten companies, some with subsidiaries. As of end-August 2002, the seven listed companies had a market capitalization of $3.6 billion. Careful development of strategic business plans, cash flow forecasts, and regular reporting are needed for the shareholder's representatives to exercise effective governance and to link management/employee performance and incentive compensation. Great Britain. British Steel (BSC) regularly submitted to the government detailed plans, both for the upcoming year and the medium-term. These plans were generally on a corporate-wide basis, although the operating budgets also showed details for each of the operating divisions of the corporation. Plans took into account government objectives for 17 the corporation as well as internal targets for performance improvements. Among other things, the plans included projections of BSC funding needs for the upcoming three years. Plans were subject of extended negotiations with government representatives, during which the plans were modified as appropriate. Monthly results for the corporation and its businesses were reported to the BSC board in the monthly management accounts. Copies were supplied to the Department of Trade and Industry (DTI) and the Treasury, and the results were subsequently discussed at the joint monitoring review meeting. This included detailed reporting on cash expenditures. South Korea. The 1983 GIE Basic Administration Act streamlined multi-layer controls by technocrats and gave GIE managers greater autonomy. While setting goals for GIEs and rigorously evaluating their performance, the government began to refrain from controlling the means through which these objectives were attained. In addition to streamlining procurement and audit procedures, the 1983 Basic Act entrusted GIEs with the authority to finalize their budget plans, subject only to common budget guidelines. Direct government representation on GIE boards facilitated this change. The Act established the GIE Management Evaluation Council (MEC), a ministerial-level council empowered to coordinate such matters as guidelines for managerial objectives, preparation of GIE budgets, and performance evaluations. Headed by the Prime Minister, the MEC had representatives from supervisory ministries and civilian experts. MEC oversaw enhancement of the GIE performance evaluation system (see Box II-3). 18 Box II-3. South Korea: Performance Monitoring of Government Invested Enterprises (GIEs) The system monitors three categories of performance indicators: · General indicators include quantitative measures of productivity and qualitative indicators of general management efficiency, such as efforts at responsible management. · Criteria for carrying out enterprise purposes include a number of enterprise-specific technical measures of operating efficiency. Most of these were evaluated on a performance vs. target basis, with the target based on a multi-year trend analysis. These criteria also allowed for benchmarking against peers. · Business administration criteria were mostly qualitative indicators such as responsiveness to changing business conditions, efficient operation of the board of directors, appropriateness of internal targets, customer satisfaction, and results from research and development. It was not uncommon for a GIE to have 30-40 performance indicators. A satisfactory grade for a quantitative indicator was based on past performance, controlled for historical deviations. All indictors were weighted, such that a GIE could receive up to 100 points. The evaluation score for each GIE was linked to a special annual bonus paid to all staff. In 1996, annual special bonuses ranged from 395 percent of monthly salary at KEPCO to 295 percent of monthly salary at Korea Coal Corp. Publication of the GIE Management Performance Evaluation Report provided additional incentives. Compared with the special annual bonus, the loss of face resulting from bad publicity may have been a greater concern for top managers at GIEs. Lim, 2002. D. Legal Status and Governance of SOEs A corporate form for state-owned enterprises, groups, and holding companies has become global best practice. The individual SOEs reviewed for this study are organized as corporations. For example, each New Zealand SOE is a limited liability company incorporated under the Companies Act of 1993. In the case of large state-owned conglomerates, such as CVRD and Petrobras in Brazil, both the parent company and its subsidiaries and affiliates have been organized as corporations and governed accordingly. The state-owned financial holding companies are organized as corporations as well. For example, Temasek was incorporated in June 1974 as a private limited company under the Singapore Companies Act. State shareholders need the flexibility to modify the corporate form of an SOE to meet changing enterprise goals. For example, · Statoil was founded in 1972 as a public limited company, subject to Norway's Public Limited Companies Act. When the state shareholder decided to sell a minority share 19 of Statoil to the public in 2001, it held an extraordinary general assembly to convert the company into a public listed company. · When IRI was assigned the mission of selling its assets to pay down its debt, in 1992, it was also incorporated as a joint stock company subject to the civil code. · When Austria's "second tier" shareholder was assigned to sell its majority shareholdings in SOEs to pay down its debt, in 1974, it was also converted from a limited public company (OIG) into a joint stock company (OIAG). · When two German railroads were merged and transformed from administrative divisions under the ministry of transport and operating under public administrative law in 1974, the resulting entity ­ Deutsche Bahn ­ was incorporated as a joint stock company. In 1999, to facilitate private investment into selected operations, Deutsche Bahn was converted into a holding company and its five operating divisions were each converted into a joint stock company with financial autonomy, its own management and corporate structure, and an ability to accept private investment. The form of board governance of SOEs tends to follow the pattern for non-state corporations. For example, Statoil, AssiDoman, OIAG, and Deutsche Bahn have the dual-board structure ­ a management board of executive directors and a supervisory board of non-executive directors ­ common for German, Austrian, and Scandinavian companies. In the case of industrial companies, workers may be entitled to a significant portion of supervisory board seats. Anglo-Saxon SOEs, e.g, in New Zealand and Great Britain, have had a single board consisting of executive and non-executive directors. A strong board can facilitate SOE restructuring, while a weak board can leave the SOE vulnerable to all sorts of harmful political interference. A comparison between Britain and Brazil is illuminating . British Steel had a strong board, which was given the autonomy to implement a drastic restructuring of the corporation during the 1980s (see Box II-4). By contrast, board governance of Brazilian SOEs was weak. While Brazilian SOEs had a two-tier board structure and each board had many formal duties, the SOE president and executive board were directly appointed by the President of Brazil or supervising minister, which created direct channels between management and the controlling shareholder that bypassed the board. In the case of Embraer, this set the stage for participation in a politically-motivated aircraft development project that almost destroyed Embraer (see Box II-5). 20 Box II-4. British Steel: A Strong and Accountable Board British Steel Corporation (BSC) went from a loss of GBP1.5 billion in 1979 to breakeven in 1985 and profit of GBP561 million in 1988, when the company was privatized. During 1977-1982, the board oversaw 30 major plant closures, capacity cuts from 21.5 to 15 million tons per year, and workforce reductions from 170,000 to 53,000. The state shareholder facilitated British Steel's transformation from a loss-maker to a viable enterprise by giving its board the autonomy to undertake managerial decisions (e.g., compensation policies to match performance) while extracting strict responsibility for performance, through annual and monthly reporting on planned and actual results. Halstead, 1996. Box II-5. Overriding a Weak Board: The Case of Embraer Founded in 1969, Empresa Brasileira de Aeronautica (Embrear) was closely tied to the President of Brazil, who appointed the company's president and three other members of the board of directors, with another two being appointed by the Ministry of Aeronautics. During 1969-1989, while continuously managed by a strong company president and his executive board, Embraer experienced several successes in aircraft design and manufacturing: the Bandierante commuter twin-turboprop, agricultural and military aircraft, and the hugely successful EMB-145 regional twin-jet, which was launched in 1989 and may soon be produced in China under a joint venture. By the early 1990s, Embraer was suffering the effects of global and domestic recession as well as the government's decision to discontinue several export finance and incentive schemes. Worse, company finances deteriorated significantly as a result of a project to develop with Argentina the CBA-123 turboprop passenger aircraft ­ motivated by desires to improve Brazil-Argentina relations at the onset of Mercosur. Faced with high development costs and unable to secure new capital infusions, Embraer borrowed at very short maturities and high interest rates. While technically sophisticated, the plane was too expensive ­ not least because the Argentine company did not have the capacities to cooperate with Embraer. In addition, promises by the two governments to buy the CBA- 123 did not materialize. In 1990-92, Embraer accumulated net losses of $776 million on $1,016 million revenues. The financial health of Embraer deteriorated to the point that, prior to privatization, the government had to inject $190 million of capital to make the company saleable. Castelar, 2002. 21 While a strong board seems a more straightforward and transparent way to creating strong SOEs, SOEs with weak board structures can sometimes exploit unique operating factors to gain a significant degree of independence. CVRD provides an interesting example. Like other Brazilian SOEs, CVRD's board of directors was appointed by the supervising ministry and Brazilian president's office. Indeed, CVRD's president was appointed directly by the President of Brazil. While CVRD's board of directors and executive board had a lot of formal powers, real power was concentrated in the hands of the company's president, who reported directly to the Minister of Mines and Energy. In practice, however, several factors enabled CVRD to avoid serious political interference: export performance; joint ventures with private companies; allocation of some reserves for public/social development programs; cultivation of a private sector culture ­ e.g., through financial transparency and investor outreach; and negotiation of a management performance contract (Castelar, 2002). A board composition that achieves an appropriate balance between independent perspectives and liaison with the state shareholder is an important aspect of SOE governance. South Korea and Sweden found it useful to have both government officials and non-governmental outsiders on SOE boards. In the case of British Steel, attendance of key government officials at board meetings provided necessary liaison. South Korea. The 1983 GIE Administration Basic Act stipulated that each GIE board was to consist of no more than 10 members, including the Chairman, the CEO, one representative each from the Economic Planning Board (now part of MOFE) and the supervisory ministry, and civilians with expertise in public enterprise management or consumer protection. The Chairman and CEO were appointed by the President. To cut down on "parachute appointments" of politically-connected individuals ­ often retired generals ­ which were demoralizing SOE managements, the Act also prohibited outsider appointments below the level of CEO. Other board members were appointed by the supervisory minister at the recommendation of the Chairman to serve on a non-standing "supervisory board" basis. Non-standing directors were reimbursed for expenses, but ­ to discourage "parachute types" from seeking director positions ­ received no fixed or incentive fees. Assessments of the 1983 reform conclude that segregation of politically- appointed directors in a non-executive supervisory board limited the risk of board members doing damage to the GIE. More positively, direct government representation on the board reduced transaction costs between GIEs and relevant ministries. At board meetings, government representatives could communicate directly with the CEO and other board members. Although half-way solutions, these measures to reduce and channel official influences on GIE management represented significant progress over previous practices (Lim, 2002). Sweden limited the size of SOE boards for the sake of efficiency. By 2000, the non- executive supervisory boards of companies with state ownership ranged from 3 to 14 and the average size had been reduced somewhat to 7.6 members. This included at least one senior investment manager from the supervising ministry. In the view of one such participant, it can often be advantageous for the state shareholder to be represented on an SOE board (Detter, 2002). 22 Great Britain. In the case of British Steel, the company had a single board with both executive and non-executive directors. The board's chairman and non-executive directors were appointed by the Minister for Industry for 3 year terms, which could be renewed. There were normally up to 7 non-executive directors. The CEO and executive directors were appointed to the board by the Minister for Industry on the recommendation of the chairman and the non-executive directors. Executive directors were appointed for operating divisions, finance, personnel, and research and development. Matters regularly considered at board meetings included the rolling corporate plan, annual budget, monthly trading report, CEO's report, major investments, divestments, and closures. Board meetings were attended by senior civil servants from the Department of Trade and Industry and the Treasury to liaise with the two main government departments. Non- executive directors staffed two board committees: audit and compensation. The compensation committee chairman had direct access to the minister of industry to discuss salaries and bonuses for the board chairman, CEO, and executive directors (Halstead, 1996). To encourage independent perspectives on the SOE board, various means may be tried to insulate the SOE board from political interference. E.g.: · The board may include non-executive directors ­ even a majority ­ from the private or non-profit sectors. For example, in New Zealand most SOE directors appointed by the state shareholders come from the private sector. In Norway, the current 7-person board of Statoil (which is 81 percent state-owned) includes a lawyer, a university representative, a trade union representative, and presidents of private businesses. According to Norway's Companies Act, employees may claim one-third of board seats. In Singapore, Temasek's 9-person board includes a private and public-sector directors. The current chairman is an ex-minister and current chairman of the banking group in which Temasek holds 13 percent ownership. Of the two deputy chairs, one is from the public sector and the other is from the private sector. Austria's OIAG has a 15-person supervisory board with five labor representatives from the Federal Labor Chamber and, according to the OIAG Act of 2000, ten members "selected for their achievements as renowned entrepreneurs from the business world, as members of the management of commercial companies or as individuals with many years experience in business." · Directors may have set terms (e.g., 3-5 years), with the presumption that early dismissal should only be for just cause. · Administration of the SOE may be subject to company law. · In the case of a 100 percent state-owned SOE, SOE directors will have a fiduciary duty to the state shareholder(s). It is probably impossible, to construct a separate fiduciary duty to the SOE itself that would enable directors to resist non-commercial interference from representatives of the state shareholder. By creating another ownership interest distinct from the state, however, the diversification of SOE ownership ­ e.g., through share sales to a strategic partner, institutional investor, or the public ­ can be a useful and effective means for promoting commercial goals over the particular interests of insider shareholders or broader political considerations. In an SOE with diversified ownership, rights of non-state shareholders can be protected 23 by company law, commercial code, or securities law and regulation. The concept of fiduciary duty is easier to advance in common law systems. In civil law systems, it may be necessary to codify key elements of fiduciary duty. Quebec's civil code is a useful example of such codification of fiduciary duties (see Box II-4). Box II-4. Codification of Fiduciary Duty: Key Elements from Civil Code of Quebec Quebec's Civil Code provides specific examples of what is meant by fiduciary duty. Excepts from cited articles from sections on legal persons, trusts, and the administration of the property of others follow. 322. A director shall act with prudence and diligence. He shall also act with honesty and loyalty in the best interest of the legal person. (see also 1309) 323. No director may mingle the property of the legal person with his own property nor may he use for his own profit... any information he obtains by reason of his duties.... (see also1313). 324. A director shall avoid placing himself in any situation where his personal interest would be in conflict with his obligations as a director. A director shall declare to the legal person any interest he has in an enterprise or association that may place him in a situation of conflict of interest....(see also 1310-1). 335. The board of directors manages the affairs of the legal person and exercises all powers necessary for that purpose.... 336. The decisions of the board of directors are taken by the vote of a majority of the directors. 337. Every director is, with the other directors, liable for decisions taken by the board of directors.... 1304. An administrator [i.e., director] is bound to invest sums of money under his administration in accordance with rules ... relating to presumed sound investments. 1315. ... no administrator may dispose gratuitously of the property entrusted to him. 1326. An administrator is bound to make an exact inventory ...of all the property entrusted to his administration...It also contains a statement of liabilities.... 1339. Investments in the following are presumed sound: ....common shares issued by a company that for three years has been meeting the timely disclosure requirements defined by the Securities Act to such extent as they are listed by a stock exchange recognized for that purpose by the Government on the recommendation of the [securities commission].... 1343. An administrator who acts in accordance with this section is presumed to act prudently. Civil Code of Quebec, (2000). Other useful measures to improve the ability of SOE boards to exercise effective governance include (i) careful attention by state and other shareholders to board appointments, (ii) appropriate division of labor among directors and use of committees, and (iii) development of board procedures. Singapore provides a good example. There, Temasek is paying a lot of attention to director appointments to GLC boards and 24 encouraging GLC boards to organize audit committees, board nomination committees, and executive compensation committees. Sweden is another good example. There, the state shareholder has overhauled the director appointments process and developed procedural rules for supervisory boards. Singapore. Temasek is able to get by with such a modest administrative infrastructure by focusing on GLC board appointments and ensuring that its GLCs have high-quality boards. Temasek nominates directors to the boards of GLCs. Such nominations are subject to the approval of a shareholders at the annual general meetings. The number of Temasek's representatives on each GLC is generally proportional to its shareholding. In turn, each GLC's board appoints the CEO. Where appropriate, Temasek's senior officers sit on GLC boards. Thus, Temasek monitors the strategic performance of its GLCs, but does not get involved in day-to-day operations. Temasek uses the nomination process and local standards for corporate governance and financial disclosure to promote high- quality GLC boards. Other key aspects follow: Depending on the nature of its operations and size, a Temasek company can have up to 12 directors. This size facilitates the formation of board committees. · Temasek is encouraging its GLCs to require, as suggested in the March 2001 Code of Corporate Governance approved by the government, that at least one-third of Corporate directors should be independent. In some cases, foreign directors have been appointed to boards of Temasek companies to provide a global perspective. · All listed companies, including those in which Temasek holds shares, are required to set up an audit committee. Temasek also encourages its non-listed companies to form audit committees. Under the Code of Corporate Governance ("Code"), an audit committee should be made up entirely of non-executive directors, at least two of whom should have accounting or financial management expertise or experience. · Consistent with the Code, Temasek is encouraging its GLCs to form Nomination Committees to nominate members to the Board of Directors. Temasek sees such Committees as a means to rejuvenate GLC boards. Nomination Committees assess Board needs, develop selection criteria, and recommend potential candidates to the Board for consideration. Temasek encourages GLC boards to pursue needed skillsets and ­ by reviewing the directorship of board members whose service has exceeded 6 years ­ to ensure that there is fresh thinking on GLC boards. · To ensure that remuneration packages are appropriate and competitive, Temasek is also encouraging its GLCs to form Executive Compensation Committees to set compensation packages, including the administration of stock option schemes. Sweden. In overhauling the composition of SOE boards, the state shareholder began by identifying a vision of where it wanted the SOE to be in 3-5 years, or even in 5-10 years. The three cornerstones for creating shareholder value ­ operations, capital structure, and business development strategy ­ served as a basis for identifying the competencies needed on each SOE board at different times in its development. In cases of listed companies, the state shareholder consulted with other major shareholders on board appointments through a Nomination Committee. A Nomination Committee of the 4-5 largest shareholders was constituted in advance of the Annual General Meeting (AGM) for each SOE. Agreed board nominations are published well in 25 advance of the AGM. Members of the Nominations Committee attend the AGM in order to justify their board nominations. In listed companies with a large state shareholding, at least one member of the Nominations Committee should represent the government. It is also the practice (??) for the Nominations Committee to propose directors fees and incentives to the AGM and to take responsibility for the selection of auditors. Finally, the MOI made efforts to change attitudes toward board membership from being a passive honor to an active and responsible position. In response to a Swedish Companies Act requirement that boards annually adopt rules of procedure, in May 2000 the government promulgated model rules. Compared with the established practices of private boards, the government aimed at stronger board involvement in SOE operations. Main points for rules of procedure for an SOE board include the following: · It should state the operational objectives, financial targets, and operational purpose of the company. It should refer to the obligation of the board and its members to administer the capital invested in the company by the government and Swedish people in order to create shareholder value, within limits set by the company's operational objectives in order to prevent investments outside the core business. · The board ­ through its chairman ­ should coordinate its views with representatives of the owner(s) on issues of strategic importance: changes in operations; major acquisitions, mergers, or divestments; changes in company risk profile or balance sheet (e.g., new debt or equity). · A clear division of responsibility between the board and the managing director should define which issues the board will resolve, and leave the managing director responsible for day-to-day management according to board guidelines and instructions. · The board chairman should be in regular contact with the SOE managing director to monitor direction and progress of the SOE. · On issues of strategic importance, the board should have access to the same information available to the managing director of the SOE. · SOE board rules should also cover the frequency of board meetings, notification, agenda, minutes, and board committees. E. Special Issues for "Second Tier" Shareholders Ownership and governance of "second tier" shareholders ­ e.g., shareholding funds, parent companies of enterprise groups, asset management companies ­ may follow the pattern that exists between "first tier" shareholders and standalone SOEs. For example, Temasek is subject to standard company law, is expected to earn a reasonable return on its investments, and is governed by a commercially-oriented board of directors. State- owned conglomerates in Brazil, such as CVRD and Petrobras, have relied on normal shareholding meetings, director appointments, and board procedures to exercise governance over large numbers of subsidiaries. 26 The special issues that arise for second tier shareholders have to do with the disposition of state assets and the creation of state liabilities. E.g.: 1. When a state-owned shareholding fund receives dividends from SOEs or proceeds from the sale of SOE shares, should these be returned to the state treasury or reinvested by the fund? 2. What should be the rules for debt or equity investments by a state-owned shareholding fund or state-owned parent company of an enterprise group? 3. How should asset management companies formed to resolve NPLs manage converted equity in distressed SOEs? 1. State shareholding funds Any proposal to allow a state-owned shareholding fund to hold and reinvest cash reserves in enterprises or to lend, borrow, or extend guarantees to provide debt financing for enterprises should be viewed with great caution. Temasek appears, so far, to have been a good steward of the state's cash reserves. Other problem cases, however ­ such as Austria's OIG and Italy's IRI ­ highlight potential dangers from allowing state- owned shareholding funds to provide equity and debt financing. Singapore. Temasek is not in the lending business. GLCs are expected to obtain debt financing on commercial terms from the market. While Temasek's financial results are not made public, there is nothing to indicate that Temasek is not continuing to exercise its relatively passive stewardship of the state's ownership interests in GLCs in a responsible and commercially prudent manner. Recent reports indicate, however, that the Singapore government may shift Temasek's emphasis from passive financial management of GLCs to a more active role. An economic review committee set up to examine how Singapore could become more competitive in the future recommended in May 2002 that Temasek should help GLCs to internationalize their core businesses and grow GLCs into globally- competitive businesses that are anchored in Singapore. Temasek indicates that it will sell businesses that are "no longer relevant or that have no international growth potential." Temasek is also prepared to dilute its stake by issuing shares or through M&A in companies that "have the potential to grow beyond the domestic market" into regional or global markets. However, the government will continue to own "critical resources" such as water, power, and gas utilities and the air and sea ports. Temasek is open to seeing GLCs that have already expanded regionally ­ such as PSA Corporation and DBS Holdings ­ partner with other companies or shareholders to regionalize or commercialize where this makes strategic or commercial sense. Temasek indicates it may also invest in new businesses with regional or international potential, especially in new growth sectors entailing high risk, large investments, or long gestation periods where Singapore's private sector is unable or unwilling to assume risks. While it "will exercise its shareholder rights to influence the strategic direction of companies," Temasek will remain hands-off in day-to-day operations.8 Austria. The original missions of OIG and (later) OIAG were to detach SOEs from politics and promote flexibility; guarantee continuity in SOE strategy and supervision; 8The Asian Wall Street Journal, July 4, 2002. 27 impose hard budget constraints; and centralize treasury and other corporate services. Starting in the 1970s, Austria entered a period of economic decline marked by oil crises and increased competition in basic materials from emerging markets. In the mid-1980s, the government responded with management changes both at OIAG and at OIAG-owned companies. The largest companies held by OIAG were split into independent, market- oriented firms. Parliament agreed on OIAG's re-capitalization but forbid any future injection of state capital. The late 1980s then saw an economic upturn. OIAG companies responded with an aggressive policy of debt-financed acquisitions. Economic conditions soured in the early 1990s: steel and aluminum prices were at thirty-year lows and OIAG debt grew to about $1.4 billion in 1992 and $5 billion in 1994. Both to deal with the public debt concerns (exacerbated by OIAG) and to revitalize OIAG portfolio companies, a 1993 law transformed OIAG's mandate. OIAG was directed to reduce its majority shareholdings to a blocking minority and to use share sale proceeds for repayment of debt. Since then, OIAG has raised about $60 billion from share sales that reduced its holdings to blocking minority positions (see Table II-1).9 These proceeds have mostly been used to repay some subsidiary debt prior to sale and to reduce OIAG's debt from Euro 37 billion in 1995 to Euro 4 billion by end-2001. The OIAG Act of 2000 gives the state a mandatory claim to 80 percent of the profits arising from OIAG sales of company shares. Upon repayment of subordinated state debt, the state would have a 100 percent claim on privatization proceeds. Italy. As noted earlier, the IRI holding company was expected to support the development of Italy's poor southern region and rescue distressed companies. Until 1987, these activities were partly supported through government transfers and endowments. IRI also borrowed from the European Investment Bank and through bond issues. Implicit government guarantees facilitated borrowing at the IRI level, at the sub- holding level, and at the operating company level. IRI extensively used this mechanism, especially after 1984 when government transfers began a rapid decline. The guarantee became explicit when IRI was transformed into a joint stock company 100 percent owned by the Treasury.10 This principle also held for sub-holding and operating company debt to the extent that there was an explicit guarantee from IRI.11 As a result, while IRI's losses increased from $540 million in 1991 to $3.4 billion in 1992 and $8.3 billion in 1993, IRI's consolidated debt surged to $58.3 billion by 1992. The government having exhausted its borrowing capacity and the lira having collapsed, IRI's reform became a priority. From 1992-93, IRI's main goal was to repay its debt, the government having agreed with the European Union that IRI would sell assets to reduce its debt/equity from 13:1 to 0.9:1 by 1996. This goals was actually achieved in 1999. During 1992-2000, IRI conducted 160 major asset sales ­ including of two banks, auto maker SEAT, Rome airport, toll roads, 31 percent of Alitalia, and a majority of aerospace company 9OIAG share sales also accounted for 45 percent of new publicly-traded equity issued during 1992-2000. 10Under Italy's Commercial Code, a sole shareholder is jointly and severally liable for the controlled company's obligations. 11In many cases, however, some sub-holdings had developed their own credit ratings and independent access to domestic and international credit markets. During periods of financial rationing, bank perceptions of credit risks are lower for operating companies ­ which are closer to underlying assets ­ than for holding or sub-holding companies. 28 Finmeccanica12 ­ and raised Lira 65 trillion ($53 billion?). IRI was liquidated in 2000 and remaining shares ­ including in public television RAI, 30 percent of Finmeccanica, and 54 percent of Alitalia ­ were transferred to the Treasury. While thus a failure in terms of regional development or corporate restructuring or financial management, IRI succeeded in liquidating itself and ­ like OIAG in Austria ­ stimulating development of Italy's capital market. Former IRI companies account for 45 percent of Italy's current stock market capitalization. 2. Enterprise groups Market-based investments, mergers, and acquisitions are more likely than administrative mandates to result in the establishment of successful enterprise groups. Brazil provides useful contrasts. Sidebras, a state-owned steel enterprises group formed in 1973-75, is a good example of a failed administrative approach. At best irrelevant and at worst counterproductive to development of Brazil's steel industry, Sidebras was liquidated by the government in 1990. By contrast, two other Brazilian enterprise groups ­ CVRD and Petrobras ­ evolved into successful conglomerates by pursuing market- based diversification over one or two decades. Brazil. In the late 1960s, a National Steel Plan proposed the formation of a steel sector holding company in order to centralize ownership, standardize management practices, rationalize capacity expansion, and allow firms to reduce costs by dividing markets to reduce transport costs and achieve greater economies of scale. While formally established in 1973, the Sidebras steel group did not receive significant capital and acquire control of all but one of the federal steel companies until 1975. Sidebras never achieved strong control of its subsidiaries, for several reasons: · It played no part in development of 1972-74 expansion plans. · It had its own capital, but was constrained in its ability to finance its subsidiaries. · Sidebras' subs mostly organized their own funding from the Brazilian National Development Bank (BNDES) or international banks. · Board appointments at subsidiaries were still controlled by the President of Brazil, the Minister of Industry and Commerce, regional elites, and BNDES. · Economic conditions and external financing constraints hurt operations. · Starting in 1979, investments, human resource policies, imports, and financing of large SOEs began to be controlled directly by a Special Secretariat for Control of SOEs (SEST), leading its subsidiaries to ignore Sidebras. In 1990, after putting Sidebras' subsidiaries ­ including Companhia Siderurgica Nactional (CSN) ­ up for sale, the government put Sidebras in liquidation. Although a good idea from an administrative point of view, Sidebras was never able to coordinate and plan the steel sector. In many instances, it only added another layer of bureaucracy, being unable to force its subsidiaries to adopt a unified strategy. Sidebras complicated management due to the duplication of functions. Sidebras created its own management structure, while the administrative and commercial departments of its 12Out of strategic interests, the Treasury retained a 30 percent shareholding in Finmeccanica. 29 subsidiaries continued to operate. In several cases, the same matter was dealt with both at the subsidiary and the parent company. The creation of Sidebras also reduced competition among the steel companies while strengthening lobbying, e.g, for financing. Lobbying was further encouraged because Sidebras management was often drafted from its subsidiaries, creating informal channels for lobbying. Before Sidebras, Brazil's steel SOEs generally competed and emulated best practices. With Sidebras' establishment and tighter price controls, rent seeking within the Sidebras group substituted for healthy competition. By contrast, CVRD provides a successful example of how to "grow" a globally- competitive enterprise group. CVRD was established in 1942 from British-owned iron mines and a railroad that were nationalized by the British government and given to Brazil to run as an SOE. Creating an ocean-shipping subsidiary in the 1960s, CVRD began to integrate vertically into transport, marketing, and iron processing. During the 1970s, it diversified horizontally and by 1985 its 34 subsidiaries and joint ventures were involved in engineering, geological research, forestry (re-forestation), cellulose, bauxite, aluminum, manganese, iron ore, and steel. In the early 1980s, CVRD bought out US Steel's interest in the Carajas mining complex and connected it by rail to Porto Madeira. In the early 1990s, CVRD bought large shareholdings in Brazilian steel SOEs undergoing privatization ­ Usiminas, CSN, CST, and Acominas. By 1997, the CVRD system consisted of 65 controlled or associated companies operating in (i) iron, gold, manganese, bauxite, potash, and kaolin mining; (ii) transportation systems consisting of two railways, ports, and 61 ships; (iii) production of iron ore pellets, aluminum, pulp, and paper; (iv) forestry; and (v) significant shareholdings in steel and alloy mills in Brazil, Argentina, France, and the US. In 1995, CVRD's assets and equity had book values of, respectively, $13.9 billion and $10.9 billion. Financing was a mix of state capital contributions; loans from BNDES and international financial institutions, operating cash flows, and trade finance. The government invested $594 million in CVRD and received $714 million in dividends before privatization, and received $5.3 billion in proceeds from two share sales in 1997 and 2002. Petrobras, established in 1953, was the other Brazilian SOE most active in diversifying and vertically integrating its activities through the creation of subsidiaries over a 9-year period. It created a petrochemical holding company in 1967, an oil derivatives distributor in 1971, an overseas exploration subsidiary in 1972, and a fertilizer holding company in 1976. With an end to its monopoly position in 1997 and increasing de-regulation, Petrobras reorganized itself in 1999-2000. Changes included (i) the creation of functional business segments to improve information and decision-making; (ii) incorporation of rate of return hurdles (similar to EVA) for these business segments; (iii) increased emphasis on integrated energy production that allows the company to participate in all parts of the energy value chain; and (iv) greater management emphasis on flexibility, transparency, results, and accountability. More specifically, Petrobras has created four business segments ­ upstream, downstream, international, and gas/power ­ and 40 business units; reduced the number of corporate units to 20; eliminated one level of management; and decentralized half of its head-office activities. 30 3. Asset Management Companies There are many examples of public asset management companies (AMCs) created to resolve non-performing loans (NPLs) and otherwise manage distressed assets, e.g., real estate. Under special circumstances, such AMCs may be able to play a useful role in financial stabilization and the resolution of some assets, but AMCs are ill-suited for operational corporate restructuring. Recent experience from the East Asia financial crisis (Mako, 2001) and earlier experience (Klingebiel) indicate that the AMC is more likely to succeed if assets taken over are small relative to GDP; if domestic financial markets are broad and deep; and if the asset taken over are relatively homogenous (e.g., mortgages, real estate, cash, and marketable securities). In addition, an AMC is more likely to be successful if tasked with relatively simple tasks, such as bundling and auctioning un-restructured NPLs. Cases of public AMCs having to manage significant volumes of converted corporate equity are few and far between.13 Poland's fifteen national investment funds (NIFs) have not been particularly successful in increasing the value of the 512 SOEs assigned to NIFs for management (see Box II-5). There is no reason to expect a public AMC to be able to drive the operational restructuring of distressed corporations. Such operational restructuring typically involves painful measures ­ e.g., sales of non-core assets, discontinuation of unprofitable products/services, plant closures, and workforce reductions ­ that a public AMC is ill-placed to force through. 13In Malaysia, Danaharta's assets were mostly loans, securities, and real estate. In Indonesia, IBRA's equity management arm has held corporate shares as collateral and sold some shareholdings, but has been unable to exert effective corporate control over these companies. In Korea, KAMCO has minority shareholdings in several Daewoo companies. Another unusual case that may warrant study is that of Germany's Treuhandanstalt. 31 Box II-5. Poland's National Investment Funds (NIFs) The NIF scheme involved 512 large SOEs, including those that were deemed in urgent need or restructuring or sale or which had not already been privatized by other means. Each NIF became the lead shareholder in 33 SOEs by receiving 33 percent of shares in 33 of 512 companies assigned to the NIF scheme, while the remaining 27 percent was held by the other 14 NIFs, 25 percent was held by the state treasury, and 15 percent was held by enterprise employees. Management of each NIF was outsourced to professional asset managers based on the results of a public tender. NIFs were able to reorganize the supervisory boards of portfolio companies and took a number of steps to turnaround SOEs in their portfolio: e.g., business plans, additional capitalization, SOE management changes, investor search, and asset sales. Between 1994 ­ the first full year of NIFs ­ and 1999, the gross profitability of NIF SOEs declined significantly, especially in comparison with other types of companies: 1994 1999 Change (points) Sold to domestic investor 13% 6% -7 Sold to foreign investor 7% 7% 0 Continuing state ownership 5% -2% -7 NIF 4% -4.5% -8.5 Source: PricewaterhouseCoopers Thus, a public AMC should seek to sell assets ­ e.g., unrestructured NPLs, restructured debt, collateral, converted corporate equity ­ as soon as possible. If a public AMC acquires significant holdings of corporate equity ­ e.g., through debt/equity conversions ­ it should seek to sell these shares as quickly as possible and, in the meanwhile, should insist on exercising its full rights as a shareholder and creditor. AMCs that have moved forward promptly on asset disposals ­ e.g., the U.S. Resolution Trust Corporation, Sweden's Securum, Korea's KAMCO, Malaysia's Danaharta ­ are considered to have been most successful. 32 III. Implications for China's State Assets Management System Recent global experience in the governance of relatively well-run SOEs suggests many lessons for China's state assets management system. These lessons may be grouped under the following five themes: · The need to focus on large SOEs; · The need to focus on the efficient use of capital; · The need to focus state ownership rights in a separate "first tier" government agency and provide sufficient transparency; · The need to focus authority and responsibility for SOE strategy and operations in each SOE's board and management; and · The need to provide an appropriate focus for "second tier" shareholders. A. Focus on Large SOEs The world's most successful SOE governance regimes have been responsible for relatively small numbers of SOEs: e.g., 16 in New Zealand; 59 in Sweden; and about 20 in Singapore. Italy's IRI which ­ with interests in about 500 companies ­ was probably the largest SOE manager in a Western economy, proved to be a financial disaster and was liquidated over an 8-year period. In addition to benefiting from well-developed legal and financial systems conducive to commercial operations by either state-owned or private businesses, narrower spans of control ­ e.g., in Sweden, Singapore, New Zealand, Austria ­ have probably facilitated good SOE governance. In hopes of achieving similar governance standards, it would be entirely reasonable for the Government to focus only on China's 9,000 large SOEs. Successful governance of this portfolio ­ with 60 times as many SOEs as in Sweden ­ would itself represent a huge challenge. At end-2000, China's 9,283 large SOEs accounted for 72 percent of the RMB 5.8 trillion book value of China's state industrial and commercial enterprises. By contrast, China's 181,225 small or medium SOEs accounted for just 28 percent of book value. Allowing for near-term ownership transformation of China's 181,000 small/medium SOEs and developing an adequate corporate governance regime for the 9,000 large SOEs would be consistent with the 4th Plenum Decision of the 15th Central Committee of the CCP in 1999 to "grasp" the large and "let go" the small and medium. This is a reasonable first cut at stratifying the issue of what to grasp and what to let go. It would, however, still be appropriate for the government to exit from non-viable large SOEs, e.g., through sale or liquidation. Contrariwise, some small/medium SOEs currently owned by state-majority parent companies of enterprise groups may remain state-owned. In any case, it is clear that implementation of the 4th Plenum Decision would lead ­ within five years, or so ­ to a major streamlining of the current infrastructure for state asset management. A government focus on "grasping the large" would still allow for equity structure diversification, as required by the 1999 4th Plenum Decision. Recognizing the benefits of 33 ownership diversification, the Government has increasingly allowed for sales of SOE shares to strategic investors or partners, institutional financial investors, and the public. The Government should be ready to reduce the State's share in many large SOEs to 51 percent ­ or even lower. Following the approach of Austria's OIAG, so long as the State shareholder(s) could forestall ­ e.g., through a blocking minority of shares ­ seriously harmful initiatives by the majority shareholder, it would be reasonable for the Government to accept minority ownership of some large SOEs. The Government and National People's Congress may wish to review existing and draft Company Law to see whether it does or could provide adequate support for a "blocking minority" approach. As small and medium SOEs are "let go," the configuration of China's state asset management system could be streamlined. Box III-1 provides a hypothetical example of a more streamlined configuration, which could have the following characteristics: · The "first tier" of a streamlined configuration would be State Ownership Agencies at the central level of government and at provincial and major sub-provincial levels. Fifty such agencies might suffice. The Ministry of Finance would be the state shareholder in some "second tier" shareholders ­ the four asset management companies (AMCs) established to resolve NPLs and the National Social Security Fund (NSSF). · In the "second tier" of a streamlined configuration, there would be no need for new asset management companies below the Ownership Agency. Indeed, with the "letting go" of small/medium SOEs, it might be possible to disestablish some existing second- tier asset managers. Most of the necessary second-tier asset management function could be performed by transforming the parent companies of enterprise groups. To begin to function effectively, these enterprise group parent companies would need to transform their approach to financial management of affiliated companies, internal controls, accounting systems, and human resources skills mix. Transformed parent companies of enterprise groups would be the main second tier agent for governing "third tier" affiliated SOEs. The NSSF and other institutional investors could play a supplemental governance role by appointing directors to the boards of SOEs ­ enterprise group parent companies, group-affiliated companies, and large unaffiliated companies. Board appointments would presumably be proportional to shareholdings. For risk management purposes, shareholdings by the NSSF or any pension fund should be limited both in terms of each investors capital and shareholdings in any single SOE. By pooling their votes, however, these investors might be able to place one or more directors on the boards of SOEs in which they hold shares. 34 Box III-1. Hypothetical Configuration for State Assets Management System After "Let Go" of Small and Medium SOEs Private Investors 1 Central Ownership Agency MOF and 1st Local Ownership Agencies (50) Tier 2nd Tier 0-75% 25-100% 100% 100% Enterprise Group 0-5% NSSF; AMCs Parent Companies Pension (4) (1600) fund(s) 3rd Tier 0-85% 25-100% 15-100% 0-75% 0-5% 100% Group Affiliated Individual Individual Companies 0- 0-5% Large Large SOEs 16,000* SOE s <500 (?) Assumed shareholdings are shown in percentages. In the case of an enterprise group parent company, for example, it is assumed that an Ownership Agency could hold anywhere from 25-100% of the shares; that private investors could hold up to 75 percent; and that the National Social Security Fund could hold up to 5 percent. On numbers of SOEs, at end-2000, the state was majority shareholder in the parent companies of approximately 1600 enterprise groups. [NB: delete 1600 from chart] Enterprise groups having then an average of ten affiliates, these state-majority enterprise groups had perhaps 16,000 affiliated companies. There being just 9,000 large enterprises, some of these 16,000 enterprise group affiliates must be small/medium SOEs. Some large SOEs operate as standalone companies ­ i.e., not affiliated with an enterprise group. 35 This scheme would more narrowly focus the attention of China's state asset managers on 1600 or so enterprise group parent companies plus some modest number of unaffiliated SOEs ­ perhaps for a total of 2000 or so SOEs. This might mean a requirement for managing 4000 board members.14 B. Efficient Use of Capital as the Central Objective The business strategy and operations of the world's best SOEs focus on the efficient use of capital ­ state equity, as well as debt financing and any private equity. An expectation from State shareholders that China's SOEs should produce positive economic value- added (EVA) should promote the development of globally-competitive companies and high-quality jobs ­ as appears to have been the case in Sweden and Singapore. Central to the efficient use of capital is the concept of opportunity cost of capital, defined as the rate of return attainable in an alternative use. A market-based rate of interest is the opportunity cost of debt capital. The opportunity cost of equity capital consists of two parts ­ the risk-free rate plus a premium for business and financial risk. An enterprise's overall cost of capital is just the weighted average of market interest on its debt capital and the expected return on its equity capital. If an enterprise's ratio of earnings before interest and taxes (EBIT) to capital exceeds its weighted average cost of capital, the enterprise is said to have positive EVA ­ i.e., it has created value. If this EBIT/capital ratio is less than the weighted average cost of capital, the enterprise has negative EVA ­ it has destroyed value. While it is natural for the Chinese state to pursue different political and social objectives, in exercising its ownership in large SOEs, the State shareholder should focus on maximizing returns on capital (which is best measured by EVA). In general, this focus would serve the interests of Chinese citizens ­ the ultimate owners of State capital. Moreover, a shift in focus to maximizing returns on capital paves the way for SOEs to be fully commercialized and in preparation for post-WTO competition. As illustrated in Section I, pursuit of multiple and non-commercial objectives by multiple representatives of the Government and Party does not support the commercialization of SOEs. More specifically, a focus on efficient use of capital by SOEs would discourage self- defeating business practices. For example, proposed capital investments, merger and acquisition transactions, and major changes in capital structure would be expected to clear designated "hurdle rates" and provide positive EVA. Such capital discipline would discourage administratively-mandated M&A transactions ­ e.g., forcing profitable SOEs to take over loss-making SOEs ­ that, in the past, have diminished state capital. A focus on efficient use of capital would also discourage practices such as administratively- mandated debt/equity conversions as the preferred response to an SOE's inability to service its debt. While a debt/equity conversion would reduce interest expense, it also represents a major opportunity cost in the use of state capital. Because an equity 14Based on the following assumptions: 2000 SOEs; average board size of 7; and average of 3-4 board appointments per qualified director. 36 investment is generally riskier than a credit investment, the state should earn more on its equity investments in SOEs. Such a disciplined approach to the deployment of state capital would encourage greater emphasis on operational restructuring, rather than superficial financial restructuring. Greater emphasis on operational restructuring would, in turn, enhance the global competitiveness of China's SOEs. While foremost emphasis on efficient use of state capital is recommended to enhance the competitiveness of China's SOEs (and, perhaps, to reverse past mis-deployments of state capital), it is entirely reasonable to expect SOEs to achieve other goals as well. Some of the other goals are not in conflict with the commercial nature of SOEs. For example, as in the case of New Zealand's SOEs, these other goals could include being a good employer, showing social responsibility to the local community, treating suppliers and customers fairly, and complying with environmental and other governmental regulations. These are the goals that even good private companies pursue. There can be, of course, non- commercial goals that the state may wish SOEs to fulfill, for example, providing poor farmers in remote areas with access to telephone by telecom company, which is commercially non-profitable. Even is such cases, however, a single-minded state owner is preferable, because it would generate more transparent measurement of the cost of such non-commercial activities. If, after recognizing the true cost, the state decides to proceed with its plan, it could still implement the plan by providing adequate compensation to the SOEs involved. Such compensation does not reduce the total wealth of the state at all, but it allows SOEs to operate on full commercial basis and therefore in a better position to compete with their private sector counterparts in the market. To implement a new state ownership function centered with efficiency of use of capital, two instruments can be important. The first is the concept of EVA. All large SOEs should be required to improve their accounting practice so that agencies that exercise ownership rights on behalf of the state can calculate EVA for each large SOE and evaluate their performance accordingly. This has been the experience of Sweden. The second is an annual statement of SOE approved their board that specifies goals to be achieved in the coming 1-3 years. This would help facilitating SOE performance monitoring and evaluation by the representative of the state shareholder. As in New Zealand, the annual statement for each SOE could specify the nature and scope of activities to be pursued; returns on state capital (e.g., return on equity, economic value-added) and other performance targets for the next three years; expected dividends; treatment of workers and suppliers; and compliance with regulatory obligations. C. Creation of Ownership Agency to Exercise State Ownership Rights Global experience illustrates the importance of insulating SOEs from misdirected bureaucratic and political influence so the SOEs can concentrate on maximizing shareholder value. A clear distinction between state ownership rights and other economic rights (e.g., regulation) is needed. Creation of a single government agency that specializes in exercising state ownership rights would help separate the ownership function of the state from its other functions. In particular, designation of such a 37 specialized Ownership Agency with the authority and the responsibility for exercising state ownership rights is needed to address the current fragmentation of authority and diffusion of responsibility for SOEs.15 Thus, efficient use of State capital could not be readily pursued without putting the ownership function into a separate Ownership Agency. Furthermore, exercising ownership with the single objective of capital efficiency requires vastly different expertise from other government functions. Thus, a new emphasis on the efficient use of capital would require a professional approach to the state ownership function. The state would have to exercise its ownership function in a way similar to professional investment agencies in the private sector. This provides further impetus for a separate and specialized Ownership Agency. Establishment of an Ownership Agency will require decisions on the following issues: (1) legal form; (2) scope of portfolio; (3) mandate; (4) appropriate exercise of ownership rights; (5) authority; and (6) financial management. 1. The Ownership Agency should be commercially-focused and professional. It may be easier to cultivate these qualities if the Ownership Agency is organized as a "public service unit" (PSU), which reports to the State Council and whose work is reported to the annual National People's Congress for its evaluation. The Ownership Agency should be staffed with full-time professionals whose compensation is market-based. Full-time staff are needed to ensure that the Ownership Agency is "real," rather than "virtual" as in some localities. Ownership Agency staff `s compensation should be market-based in order to attract staff with the necessary skill sets. Ownership Agency staff should have the business training and/or experience to be able to make integrative assessments of an SOE's marketing, production, financing, and other business plans and results. Such general management skills are more readily available in the non-state sector. To obtain a sufficient supply of necessary general business management skills, the Ownership Agency should have the ability and the readiness to match non-state salaries. 2. The Ownership Agency's portfolio should include only for-profit, non-financial SOEs. Establishing effective state ownership over China's large number of for-profit non- financial SOEs would be a huge challenge. Assigning non-profit organizations, state assets in government departments, and "public service units" to an Ownership Agency would distract the Ownership Agency from establishing effective governance over non-financial SOEs. While some 1st tier shareholders elsewhere oversee other types of organizations,16 this seems inappropriate for China's circumstances. As for financial SOEs, it would be a mistake to assign these to an Ownership Agency for non-financial SOEs. Assigning non-financial SOEs and financial SOEs to the same Ownership Agency would create or increase conflicts of interest ­ e.g., pressures for 15As noted in Section I, a variety of official entities (e.g., Ministry of Finance, State Development and Planning Commission, CCP Central Committee Enterprise Working Committee) perform multiple functions of social and political as well as an economic nature. 16In New Zealand, for example, the Crown Corporations Monitoring Unit oversees 18 non-SOE entities, including research institutes, airports, mass media, text book publication, and property valuation. 38 directed lending, avoidance of enterprise restructuring.17 It may be useful to develop some criteria for deciding whether to assign SOEs to a Local Ownership Agency (LOA) instead of the Central Ownership Agency (COA). SOEs that are very large, strategic, or involved in nationwide infrastructure networks would presumably be assigned to the COA, while most large SOEs and all small/medium SOEs would presumably be assigned to a LOA. 3. The Ownership Agency's mandate should include both management of state shares and, as appropriate, sale of state shares (equity). In performing its share management function, the Ownership Agency would ultimately be responsible for the value of State shares in SOEs. In organizing sales of state shares (equity), the Ownership Agency's goals would include maximization of sale proceeds. There is no inconsistency between the share management and share sale functions. Indeed, since maximizing the value of state shares would help maximize the proceeds from any share sale, it makes sense to link these two functions.18 Assuming that small/medium SOEs are assigned to LOAs and that the 4th Plenum Decision to "let go" is fully and expeditiously implemented, the LOAs would assume major responsibility for the sale of small/medium SOEs. 4. State shareholder interests should be exercised through normal means, i.e., annual shareholder meetings and SOE board appointments. "Shareholder interests" include seeing that the company's business plan, capital investment, financing, financial performance, risk management, key management appointments, staffing, compensation, and other business practices combine to produce positive EVA and a sufficiently robust balance sheet. To give company management sufficient space to operate, the shareholder's right to pursue these shareholder interests is normally exercised through annual shareholder meetings and board appointments. This should also be the pattern for China's SOEs. The Government's other non-shareholder interests (e.g., SOE fulfillment of contractual obligations or SOE compliance with health, safety, labor, competition, or environmental regulations) should be pursued through the appropriate commercial or administrative law channels. The "normal" exercise of shareholder rights will require the Ownership Agency to focus on making good appointments to SOE boards and on monitoring development/implementation of SOE business plans. Proportional to its shareholding, the Ownership Agency would appoint directors to 2nd tier enterprise group parent companies and to 3rd tier SOEs that are not affiliated with an enterprise group. Proportional to its shareholding, a 2nd tier enterprise group parent company would appoint directors to 3rd tier subsidiaries during annual shareholder meetings. A corporate form may be inappropriate for small/medium SOEs, but these should be "let go" in any case. 5. The Ownership Agency should have broad authority to perform its share management and share sale functions and should have sole authority to vote the State's shares at 17State-owned commercial banks and other financial institutions should be corporatized and assigned to a different shareholder. 18Share management and share sale functions are linked in other countries, e.g., Austria's OIAG, Singapore's Temasek, Sweden, and Norway. 39 SOE annual shareholder meetings and to make SOE director appointments (proportional to the State's shareholding). This would be consistent with a non- political, commercial, and professional approach to the efficient use of state capital. The Ownership Agency director would be responsible for liaising with the government and maintaining political support for the Ownership Agency's share management/sale mandate. The government would presumably provide financial performance targets and other guidelines (e.g., on appointments, conflicts, risk management) for the Ownership Agency. As long as it conforms with these guidelines, the Ownership Agency should have substantial autonomy to carry out its share management/sale mandate in pursuit of agreed financial targets.19 Consolidation of shareholder rights and responsibilities in an Ownership Agency is a major change. The Ownership Agency must have sufficient authority to perform its mandate in order to be effective. 6. While seeking to maximize the efficient use of state capital, the Ownership Agency should implement financial management procedures that provide transparency and minimize risk. The Ownership Agency's financial goals would presumably include both dividend payments from 2nd tier shareholders and 3rd tier SOEs as well as increases in the value of SOE equity. The Premier (or local equivalent) and the Ownership Agency would need to agree on "dividend policy" to guide SOE board decisions on whether SOE profits should be reinvested or paid as dividends to the 1st or 2nd tier shareholder. Measurement of SOE equity value should be as objective as possible ­ e.g., market value for listed companies, book value, EVA, or a valuation based on a multiple of dividend yield or cash flow. The Ownership Agency should have the authority to make additional equity investments in portfolio companies, but it should be prohibited from providing, guaranteeing, or directing debt financing for 2nd tier shareholders or 3rd tier SOEs. International experience (e.g., Italy's IRI and Austria's OIAG) show how dangerous it can be for a state shareholding fund or state shareholder to provide or facilitate SOE debt financing. Lastly, the Ownership Agency should not invest in or operate any other businesses. The Ownership Agency's sole mandate should be to manage or sell shares in SOEs assigned to it. In thinking about financial flows from the 2nd and 3rd tier to the Ownership Agency, linkages between the Ownership Agency to the Ministry of Finance (or bureau), and financial reporting for all three tiers, it would be useful to study New Zealand and other well-regarded cases. The Ownership Agency would focus on monitoring the performance of its SOEs and SOE boards and on board appointments and the development of SOE boards. South Korea approach to SOE performance monitoring may warrant consideration and adaptation to Chinese circumstances. The SOE governance framework outlined in this policy note will require a substantially higher degree of transparency in order to work ­ especially given the emphasis on more efficient use of state capital. Under the existing system, financial statements of non-listed 19In the case of particularly large share sales, the government may wish to retain authority to review and approve. 40 SOEs are treated as secret and available to only a few government departments. The mind-set behind this regime need to be changed. While listed companies are ones that are owned by a large number of public shareholders, SOEs are owned by all Chinese citizens. The Chinese citizens should be treated by SOEs in the same way as listed companies treat their public shareholders in terms of information disclosure and transparency. This is the case in many countries such as Sweden where SOEs are relatively well-managed. All large SOEs should be required to adopt the accounting and disclosure standards applied to listed companies in a specified period of time, say three years. Annual reports of large SOEs should be made available in the public domain. All "second tier" shareholders (e.g., AMCs and the NSSF) should also make their consolidated financial statement available to the public. The Ownership Agency or the NPC may wish to mobilize outside professionals such as investment bankers and accountants to publicly evaluate the performance of certain SOEs and second tier shareholders in terms of efficiency of their use of capital. Higher degree of transparency achieved through means such as this would greatly help the Ownership Agency in its monitoring the performance of SOEs. Approaches used by the Swedish government and Singapore's Temasek to monitor SOE board performance and fill board vacancies likewise warrant consideration. Lastly, the Ownership Agency should remain a lean organization ­ minimizing its staff size in order to avoid becoming another large bureaucracy. Focusing on working through SOE boards and on monitoring and high-quality board appointments should help. In addition, the Ownership Agency could emphasize a commercial approach to its operations and out-source SOE monitoring/advisory functions as much as possible. The Crown Corporations Monitoring Unit that assists New Zealand's Treasury provides one such model for consideration. D. Empowering SOE Boards of Directors20 Worldwide experience shows that a reasonably independent SOE board can facilitate value creation by the enterprise while a weak board may leave the enterprise vulnerable to value-destroying political interference. The composition of boards for China's remaining large SOEs should strike an appropriate balance between independent perspectives and liaison with the state shareholder. It would make sense for the Ownership Agency to appoint one or more directors and for any second tier financial institution shareholder(s) to have some proportional representation. To reinforce the distinction between administrative power and shareholder rights, civil servants should not be appointed as SOE directors. If the Ownership Agency 20This recommendation assumes ­ it goes almost without saying ­ that all large SOEs will be corporatized, that shareholders will have necessary flexibility to change corporate forms to meet changing business needs, and that the structure of SOE boards (e.g., German dual-board vs. Anglo-Saxon single board) will follow the pattern for private corporations. 41 is constituted as a PSU, it would be appropriate for Ownership Agency staff to serve as SOE directors. Such directors should be supplemented with the appointment of qualified individuals from business, other professions, or academia. Such outsider directors would provide an independent perspective. Demonstrated business management skills should be the key criteria for all appointments of SOE directors. The Ownership Agency should take extra care in appointing directors who have spent most of their working life in government administration. Additional "board empowerment" measures worth consideration include the following: · Setting a fixed term (e.g., 3-5 years) for SOE directors, with the presumption that early dismissal should be for a just cause; · Making administration of SOEs subject to the company law that applies to privately owned companies; · Encouraging further ownership diversification through sales of SOE equity to strategic investors, institutional financial investors, and the public; · Making SOE directors personally liable for damage to the SOE caused by their actions; · Providing oversight by the National People's Congress; and/or · Codifying fiduciary duties owed by directors to shareholders. Following the example from Quebec's Civil Code, this codification of fiduciary duties might cover a director's requirement to act with prudence; a definition of prudence; prohibitions on commingling of assets or use of inside information for personal gain; avoidance and declaration of any conflict of interest; reporting to shareholders; and director liability. The development of effective SOE boards will require careful attention to board appointments and directors' training as well as to board organization and procedures. Global best practices on formation and use of board committees ­ e.g., for audit and for compensation ­ should be considered. Annual adoption by SOE boards of board rules of procedure also warrants consideration. As in Sweden, such board rules of procedure could regulate the following: · Operational objectives, financial purpose, and operational purpose of the SOE; · The obligation of the board and board members to administer the SOE's capital so as to create shareholder value ­ within the limits set by the SOE's operational objectives in order to discourage ill-considered diversification; · The role of the board ­ through its chairman ­ in coordinating its views with shareholder views on issues of strategic importance: e.g., changes in SOE operations; major acquisitions, divestments, or mergers; or changes in risk profile or financing (e.g, new debt, equity, or contingent liabilities); · A clear division of labor between executive and non-executive directors, leaving the managing director responsible for day-to-day management according to board guidelines and instructions; · Regular contact between the board chairman and the SOE managing director to monitor SOE direction and progress; and · Board access to information. E. Focus for Second Tier Shareholders 42 Assuming implementation of the 4th Plenum decision to "let go" small and medium SOEs, it should be possible to organize second tier shareholding by empowering the parent companies of enterprise groups; the National Social Security Fund, pension funds, insurance companies, and any similar "fiduciary institution investors"; and the four asset management companies (AMCs). Comments on each of these groups follow. Enterprise Group Parent Companies The management and board of enterprise group parent companies will face continual decisions on whether to merge or acquire new businesses; on whether to support capital investments by affiliated companies; on whether to divest businesses and dispose of non- core assets. These transactions may generate cash reserves or absorb cash reserves or require debt and/or equity financing. Investment and financing proposals by parent company management and board decisions should focus on projections of economic value added (EVA). Merger and acquisition (M&A) transactions should be commercially motivated, not administratively mandated. If parent company cash reserves cannot be invested in a manner likely to generate positive EVA, capital should be returned to shareholders through cash dividends. Institutional Investors Institutional investors could include the National Social Security Fund (NSSF), other pension funds, insurance companies, and investment funds. While corporate governance and restructuring can be an intermediate goal, the ultimate goal of institutional investors should be to meet future obligations to their beneficiaries or investors. At present, it appears that equity holdings by institutional investors are modest (see Box III-2) and that institutional investors play no significant role in SOE governance. By providing a different perspective, directors appointed by such institutional investors could provide useful diversity on SOE boards. But prudent diversification by institutional investors ­ for example, limiting investment in a company's equity to 2 percent of institutional assets ­ may limit opportunities for institutional investors to claim director seats on SOE boards. Thus, it may be useful for China's nascent institutional investors to vote together in order to gain greater representation on SOE boards. 43 Box III-2. Current Status of China's Institutional Investors Licensed institutional investors include banks; pension funds, including the NSSF; insurance companies; investment funds; securities companies; and trust and investment companies. Their assets are estimated to amount to 7.6% of GDP. In addition, by one estimate, assets amounting to another 7.3% of GDP are managed by unlicensed "private funds." By comparison, institutional investor assets account for 50-100% of GDP in Germany and South Korea and 100% or more of GDP in the U.S., Australia, France, and Germany. National Social Security Fund assets amounted to RMB 74 billion at end-2001, with 98% in bank deposits and government bonds. In August 2001, NSSF did take an equity position in Sinopec, but this represents less than 2% of Sinopec's outstanding shares ­ not enough to entitle NSSF to appoint a director. According to provisional regulations issued by the Ministry of Finance and Ministry of Labor and Social Security in December 2001, the NSSF may invest up to 40% in securities funds or equities. Investment through one fund manager in the securities or securities investment fund issued by one enterprise may not exceed 5% of the securities portion of NSSF or 10% of total NSSF asset value. The NSSF Council plans to appoint other qualified investment management institutions to manage NSSF assets. Insurance companies. Assets of China's 52 insurance companies were RMB 459 billion at end-2001. Per 1999 regulations, insurance companies may invest 5-15% of their assets in equity investment funds. Just 5.5% of insurance company assets are in investment funds. By contrast, bonds and shares account for about two-thirds of insurance industry assets in Europe and North America. China's insurance market is growing rapidly, with new premium income rising by about 40% per annum. Investment funds. As of July 2002, China had 56 licensed investment funds with RMB 99 billion under management. These invest mainly in equities, apart from a 20% obligatory investment in government bonds. Securities companies were, until recently, to act as brokers/dealers and to refrain from operating discretionary management accounts for clients. Since November 2001, securities companies can obtain a separate license from CSRC to offer managed accounts. As of June 2002, uncommitted funds managed for clients amounted to RMB 428 billion. Kim and St. Giles (2002). Asset Management Companies Especially in cases where the AMC(s) has become a majority shareholder in the SOE as a result of a debt/equity conversion, the AMC may be called upon to function as an active "strategic investor" in turning around a distressed SOE, searching for strategic or financial buyers, monitoring the SOE's performance, and appointing SOE directors. Unfortunately, there are no examples of state AMCs successfully taking over, turning 44 around, and successfully governing large numbers of distressed companies. In cases where an AMC(s) holds only a minority shareholding in an SOE, it could not be expected to function as more than a passive "financial investor." Global experience shows, however, that public AMCs are not well-suited either for managing corporate equity or for turning around distressed enterprises. Thus, China's AMCs should seek to sell their assets ­ corporate equity as well as unrestructured NPLs ­ as soon as possible to investors better-positioned to create value. But in the interim, while the AMC still holds corporate equity, its rights as a shareholder in exercising corporate governance over the SOE should be commensurate with its shareholding. 45 Bibliography Olivier Butzbach, Case studies on OIAG and Deutsche Bahn, mimeos, 2002. Armando Castelar Pinheiro, "Corporate Governance Under Brazilian State Ownership: A Tale of Four Companies," mimeo, 2002. Dag Detter, "Valuable Companies Create Valuable Jobs: The Swedish Experience of Reforming the Management of State-Owned Enterprises, October 1998 ­ October 2001, mimeo, 2002. Ronald Halstead, "The Experience of British Steel," in Dominique Pannier, Corporate Governance of Public Enterprises in Transitional Economies, World Bank WTP323, 1996. Yongbeom Kim and Mark St. Giles, "Background Note: Developing Institutional Investors in China," mimeo, 2002. Wonhyuk Lim, "Institutional Framework for Exercising State Ownership Rights: Lessons from the Republic of Korea," Korea Development Institute, mimeo, 2002. Daniela Klingebiel, "The Use of Asset Management Companies in the Resolution of Banking Crises: Cross-Country Experiences," mimeo, 1999. William P. Mako, "Corporate Restructuring and Reform: Lessons from East Asia," senior policy forum, Korea Development Institute, November 2001. Russell Muir and Joseph P. Saba, Improving State Enterprise Performance, World Bank WTP306, 1995. Cases on Statoil, New Zealand Solid Energy, and IRI updated by Olivier Butzbach. G. Bennett Stewart, III, The Quest for Value, (Harper: 1991). Janet Tay and David Ma, "Corporate Governance of State Owned Equity: Lessons from Singapore," mimeo, 2002. Wenkui Zhang, "A Summary and Analysis of the State Asset Management System of Shanghai Municipality," Development Research Center, mimeo, 2002. 46 World Bank Note December 23, 2002 Executive Summary Reforming State Asset Management: Lessons From International Experience Issues: State asset management reform has been debated since at least 1993. During this time, authority and responsibility for exercising the State's rights as shareholder in state-owned enterprises (SOEs) has been widely dispersed. The 16th CPC Congress decision to concentrate the State's shareholder rights in SOE Ownership Agencies is a step in the right direction. But it raises many issues: How many SOEs can the State govern effectively? What should be the main goal for SOEs? What characteristics should an Ownership Agency have and how should it function? What role should 2nd tier shareholders play? This note draws on international experience to address these issues. Findings: (a) The world's most successful SOE governance regimes oversee relatively small numbers of SOEs. A narrower "span of control" facilitates good SOE governance. Full implementation of the 1999 decision to "grasp" the large and "let go" the small makes sense. Exerting effective governance over China's many large SOEs will itself be a huge challenge. Continued diversification of share ownership is highly desirable. The Government should be ready to reduce the State share in many large SOEs to 51% or less. Even with only a minority shareholding, methods for protecting vital strategic interests can be found or developed. (b) The world's best SOEs focus on efficient use of capital. The Government has many social and political objectives. But the Ownership Agency should focus on maximizing returns on state capital. (c) The new SOE Ownership Agencies should be commercially-focused and professional. An Ownership Agency's portfolio should include only for-profit non-financial SOEs. The Ownership Agency's mandate should include both management of state shares and sale of state shares. The State's shareholder interests should be exercised only through the "normal" means of annual shareholder meetings and SOE board appointments. The Ownership Agency should have broad authority to perform its mandate and sole authority to vote the State's shares and make SOE director appointments (proportional to the State's shareholding). Financial management policies and procedures should emphasize transparency and risk management (d) Worldwide experience shows that reasonably independent boards lead to stronger SOEs. The composition of boards for China's large SOEs should strike a balance between independent perspectives and liaison with the State shareholder. Additional measures to empower the board could include fixed terms for directors; application of company law to SOEs; further ownership diversification; personal liability for directors; and codification of fiduciary duties. International best practices in board organization and procedures and in director appointments and training deserve consideration. (e) Enterprise group parent companies can perform most of the necessary 2nd tier asset management functions, but must transform their approach to financial management. Institutional investors, such as the NSSF, can play a supplemental SOE governance role. The four AMCs may not be effective in turning around distressed enterprises. As long as they hold enterprise shares, however, the AMCs should enjoy full shareholder rights. Next steps: This note and background paper point toward major requirements for institutional development and capacity building among 1st and 2nd tier shareholders as well as training for SOE directors. Depending on client interest, a package of technical assistance (e.g., trust fund- financed advisors) could be developed.