10 Financial System Soundness and Reform

Wanda Tseng, and Markus Rodlauer
Published Date:
February 2003
  • ShareShare
Show Summary Details
Cem Karacadag

The past two decades have witnessed great strides in China’s financial system transformation. The authorities replaced the monobank system—typical of centrally planned economies—with a multilayered system and placed a central bank at the helm.1 In 1994, they created policy banks to take over policy lending duties from the state banks. A commercial bank law, enacted in 1995, laid the foundation for commercially oriented banking. The authorities intensified reform efforts in the wake of the Asian financial crisis and established asset management companies (AMCs), recapitalized banks, and restructured trust and investment corporations and small and medium-sized financial institutions. They strengthened prudential regulations through the introduction of a new loan classification and provisioning standards that approach international best practices. State-owned banks have taken steps to strengthen corporate governance, increase transparency, consolidate branch networks, and downsize staff. Also, stock, bond, and money markets are being developed. Finally, with its entry into the World Trade Organization (WTO), China has committed itself to allowing full market access to foreign banks by 2006.

Nevertheless, a substantial unfinished reform agenda remains. The extensive institutional reforms notwithstanding, the quality of financial intermediation needs to be improved by strengthening the incentives faced by banks, improving corporate governance, and upgrading financial market infrastructure. Virtually all banks are still state owned and enjoy the implicit guarantee of the government. The legal framework needs further development to provide effective recourse for creditors to pursue bad debtors among the state-owned enterprises (SOEs) and foreclose on collateral. The prudential framework needs to be improved to achieve international best practices in surveillance and enforcement. The stock and bond markets also need development to provide the competition and market discipline needed to steer banks toward greater efficiency and soundness. The tax system still overtaxes banks. Banks’ ability to extend loans on commercial terms remains constrained by the need to fund SOEs and government projects.2

The financial system thus has become saddled with nonperforming loans, which have emerged as one of the most important of China’s reform challenges.3 The large size of the Chinese financial system by emerging market standards (with total assets equivalent to nearly twice GDP in 2001) underscores the acute fiscal challenge and the potential drag on growth that continued financial system inefficiencies will pose over the next decade.

The interdependence of legal, regulatory, fiscal, state enterprise, and financial sector reforms has limited the speed with which reforms can be pursued in any one area. The scarcity of skilled bank and corporate managers presents enormous challenges to instilling profit-maximizing behavior in creditors and borrowers alike in any transition economy, even under the best of circumstances. The vast size of China’s economy and population, surpassing those of all of the transition economies in Central Asia and Europe combined, magnifies the challenge.

Looking ahead, the central policy challenge in the financial sector is to balance the intertemporal economic, financial, and social costs of reform. Experience in other countries suggests that the longer decisive action on financial sector reforms is delayed, the higher the economic and financial costs will be. China’s medium-term fiscal constraints, coupled with the country’s accession to the World Trade Organization, provide an outer bound on the time frame in which financial sector soundness needs to be achieved. The next 5–10 years are thus critical. Policymakers in China have reiterated their determination to deepen financial sector reforms to meet the challenges ahead, particularly by dealing decisively with the nonperforming loan problem.

Figure 10.1.Banking System Credit to the Private Sector and GDP per Capita in Developing Economies, 2000.

GDP per capita (in thousands of dollars)

Source: IMF, International Financial Statistics, various issues.

1Includes credits to SOEs.

Size and Structure of the Financial System

This section presents an overview of the size and structure of China’s financial sector.


China’s financial system is among the largest in the world. By the end of 2001 the combined assets of the country’s commercial and policy banks, credit cooperatives, and finance companies stood at nearly 200 percent of GDP, a ratio comparable to that in many G-10 countries and substantially higher than in most emerging market economies. Among its peers in terms of income per capita, China is a distinct outlier. Despite an income per capita in 2000 of only $839, bank credit to nongovernment entities was equivalent to 125 percent of GDP, a ratio that exceeded that of all other countries with incomes per capita up to $10,000 (Figure 10.1).4

The size of the banking sector reflects China’s high saving rate, depositor confidence in the system, and the lack of other investment vehicles. At 35 percent of GDP, China’s saving rate is second only to Singapore’s. Confidence stems from the public’s faith in the state guarantee of deposits held in the banking system, which in turn rests on state ownership of banks and the government’s creditworthiness. Despite their steady growth in recent years, the stock and bond markets are still small relative to banking assets, leaving deposits as the main financial instrument for savers.5

China’s large financial sector provides a basis for financial intermediation with the potential to play a key role in supporting long-term economic growth.6 However, the inefficiency of intermediation has thus far deprived creditworthy entrepreneurs and corporate borrowers of a critical source of investment financing, and therefore, it has yet to realize its full potential to contribute to economic growth (see Chapters 4 and 8).


The current structure of the financial system reflects the significant institutional changes made over the last decade to create a modern and commercially oriented system.7 The sector comprises state commercial and policy banks, joint-stock and city commercial banks, urban and rural credit cooperatives, and trust and investment corporations, whose origins, relative size, and systemic importance are summarized below.

State commercial banks (SCBs) remain the backbone of the financial system, accounting for two-thirds of system assets at the end of 2001 (Table 10.1). The four SCBs—the Agricultural Bank of China, the Bank of China, the China Construction Bank, and the Industrial and Commercial Bank of China—are among the world’s largest banks, with total assets in the range of $300 billion to $500 billion (Table 10.2). They are vertically integrated throughout their extensive branch networks (with a total of 120,000 branches at the end of 2000), with over 1.4 million employees. The SCBs have been the primary source of financing for large SOEs, which also account for the bulk of their nonperforming loans.

Table 10.1.Financial System Assets by Type of Institution

(In percent of total except as noted otherwise)1

Institution19981999200020012001 as

Percent of GDP
State commercial banks69.369.266.764.7127.1
Other commercial banks28.99.110.412.023.5
Urban credit cooperatives4.
Rural credit cooperatives8.
Finance companies1.
Specific deposit institutions7.
Source: People’s Bank of China.

Data are as of the end of the year.

Includes TICs and two of the three policy banks (China Development Bank and the Export-Import Bank of China).

Source: People’s Bank of China.

Data are as of the end of the year.

Includes TICs and two of the three policy banks (China Development Bank and the Export-Import Bank of China).

Table 10.2.Assets and World Ranking of State Commercial Banks

(Billions of Dollars)
Rank Among World’s

Largest Banks
Industrial & Commercial Bank of China483.015
Bank of China382.723
China Construction Bank305.930
Agricultural Bank of China262.637
Source: The Banker, July 2001.
Source: The Banker, July 2001.

The four SCBs were originally established in the mid-1980s, each to serve a different economic sector and to extend loans for policy objectives. In 1994 they were reestablished as commercial banks and officially absolved of their policy lending responsibilities. Since then they have expanded beyond their original sectors and have increasingly come into competition with one another (Wong and Wong, 2000). They also have been a prime focus of reform efforts in recent years.

Joint-stock commercial banks (JSCBs) and city commercial banks (CCBs) accounted for 10 percent of financial system assets at the end of 2001. Unlike the SCBs, the JSCBs have diverse ownership structures, with SOEs and local governments as their main shareholders.8 Also unlike the SCBs, the JSCBs finance smaller SOEs and nonstate firms, including small and medium-size enterprises, and maintain much smaller branch networks, typically confined to the region where they originated or to the fast-growing coastal provinces.9 CCBs have been created by the merger of urban credit cooperatives into commercial banks since the mid-1990s. CCBs also have diverse ownership structures: their equity is held by urban enterprises and local governments, and they lend to small and medium-size enterprises, collectives, and local residents in their municipalities.

China’s policy banks were created in 1994 to assume the policy lending roles previously performed by the SCBs. The three policy banks—the Agricultural Development Bank, the China Development Bank, and the Export-Import Bank of China accounted for around 10 percent of system assets at the end of 2001. Unlike the SCBs and other commercial banks, policy banks fund themselves through central bank loans, government deposits, and the issuance of government-guaranteed bonds held by commercial banks. Policy banks primarily extend long-term loans for infrastructure projects, including those financed by central government public investment bonds, and they finance development in the poorer western and central regions. The creation of the policy banks, however, has not put an end to noncommercial lending by the SCBs.

Rural and urban credit cooperatives (RCCs and UCCs) were established in the 1980s to diversify the financial system, provide small-scale banking services, and finance small, collectively and individually owned enterprises. RCCs, in particular, have become a key vehicle for channeling credit to farmers. The lending operations of RCCs, normally financed by deposits, have been funded increasingly by central bank loans, which accounted for 18 percent of their new loans in 2001. With assets equivalent to 14 percent of the financial system total in 2001, credit cooperatives have assumed systemic importance.

Trust and investment corporations (TICs) were also created in the 1980s to support the development of the nonstate sector and to provide financing outside the credit quotas imposed on commercial banks. TICs soon moved increasingly into the banking business, a trend the authorities have sought to curtail in recent years.

International trust and investment corporations (ITICs) were established by many provinces and cities in order to mobilize foreign funds (through the issuance of bonds) to finance local companies and infrastructure projects. Major financial weaknesses, including the bankruptcy of the Guangdong International Trust and Investment Corporation (GITIC) in 1998, involving a default on its debts to foreign creditors, have prompted the consolidation of TICs and ITICs since 1998. The subsector as a whole accounted for only 3 percent of financial system assets.10

The Regulatory Framework

This section describes the legal, regulatory, and prudential framework of the financial system.

The Legal Framework

The laws pertaining to monetary policy and banking were improved with the enactment in 1995 of laws governing the People’s Bank of China (PBC) and the commercial banks. Before then the framework consisted largely of a loose amalgamation of provisional regulations. The PBC law assigns the three key responsibilities of a central bank—monetary stability, banking supervision, and oversight of the payments system—to the PBC. The law also placed the PBC under the direct control of the State Council and prohibited intervention in its operations by local governments and other administrative bodies at the central level.

However, the PBC did not enjoy full independence: local governments interfered in the operations of its provincial branches. This prompted the reorganization of the PBC along regional lines in 1998 to reduce such interference. Moreover, the PBC does not have full power to conduct monetary policy or to close insolvent financial institutions. Major monetary policy and supervisory decisions must be approved by the State Council. The PBC is empowered to set supervisory standards, but the Ministry of Finance is responsible for rules on provisioning and loan write-offs.

Similarly, banks have not enjoyed the autonomy accorded by law to conduct their operations on a fully commercial basis. The commercial bank law provides a framework for transforming the SCBs into commercial entities, including by granting them operational autonomy and making them responsible for their financial performance, and the authorities are in the process of implementing the law (see Reform Strategy, below).

Important gaps in the legal framework also remain. Although several important laws have been passed in recent years (including laws on negotiable bills, insurance, auditing, securities, guarantees, contracts, and trust operations), other essential laws have yet to be adopted. A revised bankruptcy law that would strengthen the position of creditors is still to be passed. Moreover, even though several laws give creditors a senior claim on the proceeds from liquidation of bankrupt borrowers, a State Council regulation accords priority to fulfilling the social welfare obligations of workers in liquidated firms (Lardy, 2000).

The banking environment is also distorted by existing tax laws. The effective tax burden is excessive, undermining incentives to maximize profits. Banks are subject to a “business tax” of 6 percent and a corporate income tax of 33 percent.11 Business taxes are levied against gross income (including accrued interest on nonperforming loans), which results in the taxation even of institutions that lose money. At current levels of profitability, the effective tax burden approaches 70–80 percent.

Prudential Regulation

Prudential regulations have been gradually strengthened in recent years. Nevertheless, much more remains to be done to bring them up to standards practiced in more modern financial systems and to enforce them effectively.

Since 1998, important steps have been taken to tighten prudential regulation, especially with respect to loan classification. Until the end of 2001, the loan classification system was far behind those in industrial countries and advanced emerging market economies (Table 10.3). Effective January 1, 2002, the authorities have adopted the five-category system widely used in other emerging market economies, according to which the entire balance of the loan is classified as nonperforming, and interest accrual suspended, once interest or principal is more than 90 days past due.

Table 10.3.Selected Prudential Regulations
CategoryRegulations in Effect
Loan classificationAs of January 2002
  • Loans are classified in five categories: normal, special mention, substandard, doubtful, and loss, where the latter three categories are considered as nonperforming

  • A loan is classified as nonperforming once principal or interest is past due more than 90 days. Once a loan is classified, the entire balance of the loan is classified as nonperforming.

  • Unpaid interest is suspended after it is in arrears for more than three months.

  • However classification of a loan as loss is still subject to several preconditions related to the borrower (bankruptcy, decease, natural calamity, or deregistration), according to the Ministry of Finance regulation on provisioning, adopted in 2001.

Effective until December 2001
  • Loans were classified in four categories: normal, overdue, idle, and loss, where the latter three categories were considered as nonperforming.

  • A loan was classified as overdue after principal payment(s) were in arrears by more than 1 day, and idle after 12 months. Only the scheduled but unpaid portion of principal, not the entire balance, was classified

  • Unpaid interest was suspended after it was in arrears for more than six months

  • Overdue, idle, and loss loans could not exceed 8, 5, and 2 percent of loans, respectively.

Loan concentration
  • Lending to a single borrower may not exceed 10 percent of capital

  • A bank’s 10 largest exposures combined may not exceed 50 percent of its net capital.

Provisioning and write-offs
  • Until 2001 banks could maintain loan-loss reserves (stock) up to 1 percent of outstanding loans

  • Effective January 2002, banks are required to allocate general provisions of 1 percent of total loans and specific provisions in the following amounts: 2 percent of special-mention loans, 20–30 percent of substandard loans, 40–60 percent of doubtful loans, and 100 percent of loss loans. Banks are expected to fully comply with these requirements by the end of 2005

  • Classifying a loan as loss for provisioning purposes and loan write-offs require approval of the Ministry of Finance.

Capital adequacy
  • Definitions of capital and weighted risks are generally consistent with the Basel capital accord, with the important exception of the treatment of large SOEs

  • Risk weights assigned to large SOEs range between 50 and 70 percent, in contrast to the 100 percent risk weight stipulated in the Basel accord

  • Minimum ratios of total capital and core capital to risk-weighted assets are 8 percent and 4 percent, respectively.

  • Banks are required to maintain reserves equal to 6 percent of local currency deposits and 2 percent of foreign currency deposits

  • Loan-to-deposit ratio is to be kept under 75 percent for renminbi loans and deposits

  • Interbank debts are to be kept under 4 percent of deposits.

Foreign exchange risk
  • Foreign currency liquid asset-liability ratio must be no less than 60 percent

  • Loan-to-deposit ratio may not exceed 85 percent for foreign currency loans and deposits

  • Bank overseas borrowing may not exceed 60 percent of net capital

  • Open positions may not exceed 30 percent of paid-in foreign currency capital

  • Overnight open positions may not exceed 5 percent of paid-in foreign currency capital.

Sources: People’s Bank of China, State Administration for Foreign Exchange, and Fitch IBCA, Duff & Phelps (2001).
Sources: People’s Bank of China, State Administration for Foreign Exchange, and Fitch IBCA, Duff & Phelps (2001).

Provisioning rules have also been substantially improved. Until early 2001 the maximum allowable provision, on a cumulative basis, was 1 percent of outstanding loans. As a result, banks were not permitted to make provisions commensurate with the credit risks on their loan books. Effective January 2002, the authorities introduced a new provisioning regulation designed to complement the new loan classification system, and which is comparable to practices in more advanced emerging market economies. The regulation requires general provisions of 1 percent of total loans to cover potential losses on normal loans, and specific provisions of 2, 25, 50, and 100 percent of the loan amount for special-mention, substandard, doubtful, and loss loans, respectively. The specific provisions are not tax deductible. The PBC expects banks to fully comply with the new requirements by the end of 2005, but sanctions for noncompliance remain undetermined.

The minimum capital adequacy requirement is broadly consistent with the Basel capital accord, with the important exception of the treatment of large SOEs. Whereas the Basel accord stipulates a 100 percent risk weight for SOEs, the Chinese regulation assigns risk weights ranging from 50 to 70 percent for large SOEs. Given the heavy exposure of banks to large SOEs, their risk-weighted assets may be higher than reported. In addition, because of insufficient loan-loss provisioning, reported capital and capital-adequacy ratios do not reflect the true capital strength of banks.

The prudential regulation of foreign exchange risk management is an area of increasing importance, given the growth in cross-border transactions that is likely to result from China’s entry into the WTO and its growing integration with the world economy. The surge in foreign currency deposits in recent years (from $77 billion in 1997 to $135 billion in 2001), moreover, has already given rise to additional credit, liquidity, and exchange rate risks, which require careful management. Prudential regulation and supervision thus need to ensure that banks safeguard the liquidity and credit quality of their foreign currency assets to avoid their foreign currency liabilities becoming a contingent liability of the government and a potential claim on central bank foreign reserves. Foreign exchange risk management will assume even greater importance as the exchange rate regime becomes more flexible over time.

Regulations for managing foreign exchange risk need to be strengthened (Table 10.3). For example, the minimum liquidity ratio does not provide adequate protection against foreign currency liquidity risks, because short-term foreign currency loans are included among eligible liquid assets even though such loans may not be easy to liquidate in practice. In addition, not all existing prudential regulations appear to be enforced. For example, total foreign currency loans exceeded foreign currency deposits in 1998, indicating that some banks were noncompliant with the 85 percent cap on the loan-to-deposit ratio. Moreover, because bank capital is overstated as a result of underprovisioning, the 30 percent cap on open positions is not an effective limit on foreign exchange risk. Finally, most regulations on foreign exchange risk are administered by the State Administration for Foreign Exchange from an “exchange regulation” perspective, rather than by the PBC from a supervisory perspective.

Financial System Soundness and Efficiency

This section assesses the financial condition of the financial sector, with an emphasis on its contribution to economic growth and on the size of the nonperforming loan problem.

Systemic Soundness and Potential Losses

A notable achievement since 1998 has been the increasing transparency of the supervisory authorities and of commercial banks in disclosing financial information. Disclosure on nonperforming loans in particular has enabled more accurate assessments of potential systemic losses and set the stage for a more intense reform effort to meet the financial sector’s challenges (see Financial Sector Reform, below). The SCBs and policy banks have begun publishing annual reports and financial statements that are closer to international banking and accounting standards.

The financial system suffers from substantial weaknesses, largely reflecting the legacy of policy lending. Even after the transfer to AMCs of nonperforming loans equivalent to 14 percent of total loans in 1999–2000, the reported nonperforming loans of SCBs averaged over 30 percent of total loans at the end of 2001. However, reported non-performing loans may be subject to revisions,12 owing to the inherent difficulty in examining loan files in numerous branches, limited central oversight of branch operations, and differences in accounting practices among branches.13

JSCBs are younger than SCBs and thus carry less of the burden of past quasi-fiscal lending. Reported nonperforming loans of selected JSCBs indicate that their asset quality is stronger than that of the SCBs, but their nonperforming ratios are still high compared with other emerging market economies in Asia. For example, the Bank of Communications, the largest JSCB, reported a nonperforming loan ratio in 2000 of 18 percent, and several smaller banks reported ratios in the range of 9–39 percent in 1999 and 4–29 percent in 2000 (Table 10.4). The rapid growth of JSCBs, with credit expansion averaging 30 percent annually during 1998–2001 (“other commercial banks” in Figure 10.2), entails risks that a portion may become nonperforming in the future unless credit evaluation and monitoring improve significantly.

Table 10.4.Reported Nonperforming Loans at SCBs and Selected Other Financial Institutions

(In percent of total)1

State commercial banks2
Agricultural Bank of China35342
Bank of China392728
China Construction Bank42123
Industrial and Commercial Bank of China3430
All SCBs (weighted average)3030
Selected JSCBs and development banks
Bank of Communications2218
China Merchants Bank10108
China Everbright Bank392921
China Minsheng Bank94
CITIC Industrial Bank1715
Guangdong Development Bank23
Shanghai Pudong Development Bank911
Shenzhen Development Bank2523
Sources: Fitch IBCA, Duff & Phelps; Wong and Wong (2000); and institution annual reports.

Data are as of the end of the year.

Data for 2000 and 2001 are reported according to the new, five-category loan classification system, except where noted otherwise.

Reported according to the old, four-category system.

Includes Y 50 billion for 2001 and Y 17 billion for 2000 in debt-equity swaps from previously nonperforming exposures.

Sources: Fitch IBCA, Duff & Phelps; Wong and Wong (2000); and institution annual reports.

Data are as of the end of the year.

Data for 2000 and 2001 are reported according to the new, five-category loan classification system, except where noted otherwise.

Reported according to the old, four-category system.

Includes Y 50 billion for 2001 and Y 17 billion for 2000 in debt-equity swaps from previously nonperforming exposures.

Figure 10.2.Change in Credit Growth from Previous Year

(In percent)

Sources: People’s Bank of China; and IMF staff estimates.

Although information on credit cooperatives is very limited, they are probably the weakest subsector of the financial system. RCCs’ nonperforming loans reportedly reached 38 percent of total loans in 1996 and have since risen further (Lardy, 2000). Another sign of the RCCs’ troubled financial condition is that their consolidated balance sheet showed negative equity and a largely negative “other items net” in 1999.14 The health of UCCs is similarly poor, as evidenced by their liquidity problems, which prompted the central bank to merge UCCs into CCBs. Lack of information on the TICs and ITICs also rules out a quantitative assessment of their asset quality. However, the collapse of numerous TICs in recent years, starting with the high-profile case of the GITIC, suggests that this subsector has also been weak.

The potential losses of the financial system are thus substantial, and the bulk of those losses ultimately may have to be borne by the government, given the banks’ already weak capital positions, negligible loan-loss reserves,15 and poor profitability. Worldwide experience has shown that banks and supervisors often underestimate nonperforming loans and that asset quality can deteriorate quickly under adverse economic conditions. The analysis in Chapter 11 of the sensitivity of enterprise performance to modest interest rate and demand shocks confirms the potential for a further deterioration of borrowers’ creditworthiness under economic adversity.

Bank profitability has been very low. Among the SCBs, the Agricultural Bank of China registered losses in recent years (but then turned a small profit in 2000), and the remaining three have reported barely positive earnings; the SCBs as a group reported a return on assets of 0.2 percent in 2001. Even these levels of earnings are overstated given the lack of provisioning and, until recently, the accrual of interest on loans in arrears for six months. The low profitability of the SCBs also reflects their lack of business diversification, minimal sources of noninterest income, and still-high operating costs related to their large workforces and branch networks, despite the cutbacks of recent years.16 Available data also illustrate the tax burden imposed on financial institutions. For example, the SCBs paid taxes in excess of net income in 1999.

In contrast to the SCBs, the JSCBs appear to be more profitable, with a return on assets of 0.4 percent in 2001. This stronger profitability stems from their stronger asset quality (a larger share of loans to the nonstate sector), lower overhead costs, and fast-growth strategies. However, the JSCBs’ earnings are also inflated by shortfalls in provisions, and their rapid-growth strategies may backfire in the form of loan losses in the future.

Despite weak asset quality and low profitability, systemic liquidity is high. Reserves (cash and deposits with the central bank) account for 12 percent of system deposits, nearly double the required level of 6 percent. This ample liquidity reflects continued strong deposit growth—which averaged 16 percent annually during 1997–2001—driven by China’s high saving rate and the lack of investment alternatives.

Although high in the aggregate, liquidity is spread unevenly across the financial system, and the interbank market is neither developed nor integrated enough to bridge the liquid segments of the system with the illiquid segments. The SCBs are highly liquid, but pockets of illiquidity appear to exist among the credit cooperatives. At the end of 2000, the SCBs held reserves in excess of 8 percent of their total assets, and another 19 percent of their balance sheet was invested in government securities, in part reflecting the SCBs’ growing risk aversion and shift toward assets with lower credit risk. Signs of illiquidity among the rural credit cooperatives are visible in the PBC’s lending patterns to them: its loans to cooperatives have doubled each year since 1998, albeit from a low base (Figure 10.3).

Figure 10.3.Lending by the People’s Bank of China to the Financial System

(In percent of deposits)

Source: People’s Bank of China.


State ownership of the financial system and of major borrowers reduces incentives for good corporate governance.17 With the sole exception of China Minsheng Bank, a medium-sized JSCB, the financial system is entirely state owned. This is in sharp contrast to the nonfinancial part of the economy, where nonstate enterprises account for more than half of gross annual output. The asymmetry in ownership structures between the financial and the nonfinancial sectors limits the economy’s growth potential because of the financial system’s inability to adequately select and support nonstate firms. Banks have little incentive to exert market discipline over borrowers, particularly the large SOEs, given the government’s implicit guarantee. That guarantee, in turn, creates moral hazard on the part of both lenders and borrowers and slows the development of a “risk culture” in which lenders are able to assess and price credit risks, and of a “credit culture” in which borrowers assume responsibility for repaying their debts.

Mirroring the state ownership of banks, the bulk of bank lending has gone to SOEs rather than to the faster-growing nonstate sector. Although a precise measure of the composition of lending by borrower is unavailable, the stock of “direct” loans to the nonstate sector was only 5–6 percent of total lending in 2001, and under 50 percent even when “indirect” loans are included, according to a PBC survey (PBC, 2000).18 A 1999 survey conducted by the International Finance Corporation of 600 private Chinese enterprises also revealed that private firms primarily relied on self-financing and that 80 percent cited lack of access to outside financing as a major constraint (Gregory and Tenev, 2001). In the same vein, the contribution of financial intermediation to growth appears to have been marginal: Chapter 4 finds, based on provincial lending and growth patterns, that bank lending to the state sector has made no discernable contribution to economic growth, and that lending to the private sector has had only a small impact.

Inadequate competition in the banking industry has contributed to the inefficiency of financial intermediation, especially on the part of the SCBs (Wong and Wong, 2000). In particular, a number of administrative rules discriminate against JSCBs’ ability to compete with the SCBs. The central bank imposes quotas on the number of new branches that each JSCB may establish, and enterprises usually have to open deposit accounts with banks from which they borrow, a rule that works in favor of the SCBs. Moreover, government administrative and nonprofit units are required to open their accounts for revenue collection and extra-budgetary funds with SCBs, and central bank lending appears to favor SCBs, which receive a much larger share of central bank financing than do the JSCBs in relation to their own resources. (However, this could be attributable to past policy loans of the SCBs funded by the PBC; see Figure 10.3). Although these distortions are detrimental to competition, JSCBs have still experienced rapid growth in recent years.

The absence of reliable financial information on enterprises is an impediment to credit risk assessment. Corporate fixed assets, inventories, receivables, and earnings are often overvalued, whereas leverage is understated (see Chapter 11). Having developed in a difficult environment that until recently favored state enterprises, private firms tend to be opaque (Gregory and Tenev, 2001). Channeling a larger share of credit to creditworthy private enterprises thus hinges on fostering greater adherence to formal accounting rules.

The rapid pace of credit growth throughout the 1990s has been a major contributor to the weaknesses of the financial system today. Credit growth was particularly fast from 1991 through 1997, averaging 26 percent a year. Bank lending financed massive investments in real estate and property, excess capacity building, and inventory accumulation (Lardy, 2000). Although the rate of expansion in overall credit has decelerated by half since 1998, the surge of new credits in certain financial subsectors, particularly the JSCBs, risks new problem loans emerging in the future.

Financial Sector Reform

This section reviews the progress made in financial sector reforms to date and discusses elements of a comprehensive reform program for the future.

Reform Strategy

The Asian financial crisis of 1997 pushed financial sector reform to the top of the government’s reform agenda. The reforms aimed to gradually transform the SCBs into viable and competitive universal banks; to contain losses in and consolidate the credit cooperatives and the TICs and ITICs, which remain the weakest links in the financial system; and to increase competition by allowing commercial banks other than the SCBs to assume an increasingly important role in domestic financial intermediation. The reforms have involved greater transparency with regard to the condition of the financial system; stricter regulatory standards; reduction in government interference in the system; recapitalization, balance sheet clean-up, and operational restructuring of SCBs; the creation of AMCs; the merger and closure of ailing credit cooperatives, TICs, and ITICs; and opening to foreign competition pursuant to China’s WTO commitments.

Recognition and public acknowledgment of the extent of the financial sector’s weaknesses have been a catalyst for deeper reforms. The government has announced more-credible statistics on banking soundness, particularly with respect to asset quality. Official recognition of bad loans has been complemented by a willingness on the SCBs’ part to disclose more comprehensive and accurate financial information, and there has been some movement toward international accounting standards with the aid of external audits of main branches.

Progress on transparency has been brought about in part by improvements in prudential regulation, particularly with respect to loan classification and provisioning. The recent adoption of the five-category risk-based loan classification system and corresponding provisioning requirements represents a critical step in improving the accuracy and credibility of bank financial statements. They are also prerequisites to building risk and credit cultures in the financial system.

The renewed effort to reduce government interference in the financial system began with the National Financial Sector Work Conference in November 1997, where a consensus was reached on allowing the financial sector to operate commercially. This was followed by the restructuring of the central bank in 1998. PBC branches in each of China’s provinces were replaced by nine regional branches, which enjoy greater independence on credit policy and supervisory decisions. That same year the government eliminated credit quotas on the SCBs, conferring greater responsibility on them for their lending decisions. In 1999 the PBC, provincial governors, and Communist Party secretaries held another Financial Sector Work Conference, which explicitly forbade government interference in commercial lending.

The rehabilitation of balance sheets has involved recapitalization and the removal of nonperforming loans. The government injected Y 270 billion (equivalent to 3½ percent of GDP) in equity into the SCBs in 1998. During 1999–2000 the four SCBs and one policy bank (the China Development Bank) transferred, at book value, Y 1.4 trillion in nonperforming loans (14 percent of their total loans and the equivalent of 17 percent of 1999 GDP) to four newly established AMCs.19 The AMCs provide a basis for asset resolution, and notable progress has been made in asset valuation and disposal, aided by the use of foreign expertise. Thirteen percent of AMC assets had been sold by the end of 2001, including some sales to foreign investors, with an average cumulative cash recovery rate of 21 percent (Box 10.1).

SCB reforms have focused on cutting costs and strengthening bank governance, internal controls, and credit risk management. The four SCBs cut staff by nearly 130,000, and branches and savings outlets by some 40,000, between 1997 and 2000. Each SCB is establishing boards of supervisors and boards of directors to monitor and oversee their operations. Credit departments are developing internal risk rating systems, and credit analysis and approval functions have been separated, with approval authority centralized at headquarters and subjected to credit committee review. Moreover, the composition of new loans has shifted in favor of lower-risk housing and consumer loans, and among corporate loans to more creditworthy clients in the coastal regions. Nevertheless, much remains to be done to ensure sustained improvements in risk management and internal controls. In particular, skills in pricing and managing credit risks are scarce, and training of staff in risk management skills will take time, given the size of the SCBs. More-sophisticated information systems also need to be put in place. The PBC has set a goal of reducing the SCBs’ combined nonperforming loans ratio to 15 percent by 2005 (from 30 percent at the end of 2001), implying a 4-percentage-point reduction per year in the nonperforming loans ratio of each SCB.

In the long term, the authorities plan to list the SCBs on the stock exchanges in order to diversify their ownership, in a manner akin to JSCBs. Three JSCBs have already listed on the domestic bourses, most recently China Minsheng Bank in October 2000, and two others have announced plans to be listed during 2002, with more to follow.20

Closures and mergers have been the principal means used to consolidate the credit cooperatives and the TICs and ITICs. Insolvent rural credit funds were liquidated, and solvent ones were merged with RCCs. The number of RCCs has been cut by several thousand in recent years. Restructuring plans for the RCCs involved the consolidation by merger at the city and the provincial level and a strengthening of supervision. However, the RCCs’ nonperforming loans problem remains to be addressed. A key factor influencing restructuring has been the desire to avoid disrupting the flow of credit to farmers at a time when raising rural incomes has become a prime policy objective. The reform of UCCs is proceeding along similar lines; closures are also part of the process.21 As with the UCCs, the consolidation of the TICs and ITICs involves a mix of mergers and closures, through which their number will be cut to 60 from 239.22

Box 10.1.Operations of the Asset Management Companies

In 1999 the government established four AMCs—one for each SCB—each of which assumed problem loans from its SCB counterpart. (Huarong AMC also assumed problem loans from the China Development Bank.) Nonperforming loans with a face value of Y 1.4 trillion, the equivalent of 14 percent of total loans and 17 percent of 1999 GDP, were transferred at book value. In exchange, the AMCs recorded a debt to the SCBs of about Y 820 billion, with a 10–year maturity and an annual interest rate of 2.25 percent, and were released from corresponding PBC refinance credits (about Y 570 million), which are now PBC claims against the AMCs. Nearly one-fourth of the loans transferred—an estimated Y 300–350 billion—is being swapped for equity in SOEs, which will be held by the AMCs in the form of an ownership stake to facilitate the restructuring of selected SOEs. The AMCs will assume ownership stakes in some 580 SOEs. (See Chapter 9 for a more detailed description of the debt-equity swap program.)

The AMCs were created as independent financial institutions with the primary goal of maximizing recoveries from nonperforming loans through various asset resolution techniques. In light of the diversity and large number of debtors under their management, the AMCs have increasingly used foreign expertise to manage and dispose of the loans, including through auctions to foreign investors. Thus far the AMCs have sold only 13 percent of total assets assumed. Cash recovery rates have ranged between 7 and 31 percent among the four AMCs, with an average cumulative cash recovery rate of 21 percent as of the end of 2001 (Table 10.5). These recovery rates are likely to decline as the AMCs dispose of assets of increasingly poor quality.

Despite the progress made in asset disposal, the AMCs face important hurdles and deficiencies in governance, the removal of which will be important in maximizing the value of assets they hold:

  • Shortage of staff and skilled expertise. AMCs are understaffed and lack the skilled expertise to assess the value of their nonperforming loans and to manage the companies in which they have ownership stakes. For example, at one AMC a workforce of 1,300 (including clerical staff) manages the debts of 70,000 borrowers.

  • Legal impediments. AMCs lack the power, even when they are majority shareholders, to replace SOE management and restructure their companies. They are also unable to sell their shares for a period of three years, or at a discount at any time. Moreover, legal uncertainties remain over the transferability of ownership, the tax treatment of potential capital gains, and foreigners’ ability to repatriate their investments in AMC assets.

  • Governance. The financial operations and policies of the AMCs diverge widely: some have not serviced their debts to the SCBs and the PBC, while others have made payments in full. The lack of transparency and consistency in AMCs’ financial operations may undermine incentives to maximize recoveries and make it difficult to gauge their performance and hold management accountable.

Table 10.5.Operations and Recoveries of Asset Management Companies, December 31, 2001(In billions of yuan except where noted otherwise)



Great Wall

Book value of assets assumed373.0407.7267.4345.81,393.9
Recoveries at face value61.231.123.461.0176.7
Cumulative recoveries27.315.811.012.967.0
Assets disposed of (in percent)
Cash recovery rate (in percent)28.330.924.47.220.9
Total recovery rate (in percent)44.650.847.021.137.9
Sources: AMC data, news reports, and IMF staff estimates.
Sources: AMC data, news reports, and IMF staff estimates.

China’s accession to the WTO will open the financial system to greater foreign competition, which will serve as an important source of market discipline. Foreign banks will be allowed to conduct domestic currency business with Chinese firms in 2003, and with retail customers in 2006. Although foreign banks’ market share in these sectors is likely to remain modest, their entry should still have a positive impact on domestic banks by intensifying competition and introducing knowledge and new technology.

The Future Reform Agenda

The achievements to date notwithstanding, there is a broad consensus on the need to deepen and accelerate reforms. Building on recent and planned reforms, the future reform agenda should include the following items.

Formulation and Public Announcement of a Comprehensive Reform Strategy

The plan could include, among other things, a statement of the authorities’ reform objectives and the legal, regulatory, prudential, and institutional measures planned over the next 3–5 years to achieve those objectives. The strategy should also include a long-term vision for the financial sector, which could envisage creating a financial system that is more diversified and specialized—reflecting China’s geographic and socioeconomic diversity—than the current one, which is dominated by vertically integrated SCBs.

Transfer of Ownership to the Private Sector

Transferring management control and ownership to the private sector could proceed along the following lines: comprehensive diagnosis of the asset quality and solvency of financial institutions; identification of viable and nonviable parts of financial institutions and the separation and spin-off of good assets under wholly owned subsidiaries of the parent; and the listing or sale to strategic investors of solvent institutions and subsidiaries.

Stricter Accounting, Internal Reporting, and Disclosure Standards

Accounting rules and practices should be strengthened by moving toward international standards and by requiring more independent audits of banks and nonbank firms. These steps should be accompanied by more frequent publication of detailed financial information to enable the exercise of market discipline.

Gradual Liberalization of Interest Rates

Deposit and lending rates remain largely fixed,23 inhibiting banks’ pricing of credit risks. More market-based interest rates would allow banks to allocate resources more efficiently. Interest liberalization could proceed in stages, starting with a gradual widening of the existing band around lending rates.

Reduction of the Tax Burden

Consideration could be given to phasing out the business tax altogether, while holding banks more accountable for their financial performance.

Accelerated Operational Reforms

Weak banks not in compliance with prudential requirements should be subject to time-bound operational restructuring agreements, including specific financial performance and prudential compliance targets against which management is held accountable. When banks’ business plans and performance fail to meet the specified criteria and targets, management should be improved and possibly replaced. Operational reforms should emphasize the strengthening of risk management practices relating to credit, market, and liquidity risks. Risk management skills will also have to be strengthened in parallel with the growing market and foreign currency risks that more-flexible interest and exchange rates will engender.

Strict Regulatory and Prudential Enforcement

With prudential regulations having been considerably strengthened in recent years, strict enforcement of rules is critical to securing their intended benefits, including the exit of nonviable institutions and the development of risk and credit cultures in the economy.

Improvement of the Framework for Asset Resolution

Although the AMCs provide a basis for efficient asset resolution, important impediments remain. Legal uncertainties about loan collateral, foreclosure, and creditor rights deter potential investors in distressed assets, and AMCs possess little power to restructure and improve the performance of ailing enterprises in which they hold substantial shares. The diversity and magnitude of distressed assets also highlight the need for AMCs to acquire skills, train staff, and maximize the use of foreign expertise in asset valuation and recovery. More generally, a good framework for resolving distressed corporate debts is critical to the development of a sound corporate sector as well as a sound banking sector.


    Fitch IBCA Duff & Phelps2001“Chinese Bank Prudential Regulations”September (London: Fitch IBCA, Duff & Phelps).

    GregoryNeil and StoyanTenev2001“The Financing of Private Enterprise in China,”Finance & DevelopmentVol. 38No. 1 (March) pp. 1417.

    LardyNicholas2000“When Will China’s Financial System Meet China’s Needs?”paper presented at the Conference on Policy Reform in China, Stanford, CaliforniaNovember 18–20, 1999, revised February 2000.

    Organization for Economic Cooperation and Development2002“Challenges to China’s Banking Industry” in China in the World EconomyMarch (Paris: Organization for Economic Cooperation and Development).

    People’s Bank of China2000Annual Report (Beijing: People’s Bank of China).

    WongRichard Y.C. and Sonia M.L.Wong2000“Competition in China’s Banking Industry”October (Hong Kong SAR: Hong Kong Centre for Economic Research).

Under the monobank system, the People’s Bank of China combined the roles of central and commercial banks and was subject to strict cash and credit plans set in accordance with the production plans drawn up by the State Planning Commission. Banks were part of the administrative hierarchy, whose primary goal was to ensure that national production plans were achieved (Wong and Wong, 2000).

As discussed in Chapter 11, SOEs continue to employ over half the industrial workforce and still provide essential social welfare services to workers, including health, education, and housing.

See the section on “Financial System Soundness” below and Chapter 8.

If nonbank financial institutions’ credits to the private sector are included, the financial sectors of the Republic of Korea and Malaysia (with incomes per capita of $9,900 and $3,880, respectively) surpass that of China on this measure.

At the end of 2000, total bonds outstanding, including those issued by the government, banks, and enterprises, equaled 24.5 percent of GDP. Stock market capitalization, including the Shanghai and Shenzhen stock exchanges, stood at 45 percent of GDP at the end of 2001, but most shares are owned by the state, and only about one-third are tradable.

As noted in Chapter 4, the literature on the relationship between finance and economic growth has amply demonstrated the strong correlation between the depth of financial intermediation and growth.

This section draws on OECD (2002).

China Minsheng Bank, with total assets of Y 48.6 billion ($6 billion) in 2000, is the one exception: the majority of its shares are held by nonstate entities. It is the only privately owned Chinese bank.

The Bank of Communications and CITIC Industrial Bank, China’s fifth-and sixth-largest banks, with assets of Y 628 billion ($76 billion) and Y 235 billion ($28 billion), respectively, in 2000, are the two largest JSCBs.

The China International Trust and Investment Corporation (CITIC) dominates the subsector, accounting for over half the assets of all TICs, although the bulk of its consolidated assets derives from its fully owned banking subsidiary, CITIC Industrial Bank.

The business tax rate is scheduled to decline to 5 percent in 2003 but remains excessive even at that level.

For example, the PBC’s own estimates of nonperforming loans have been subject to upward revisions in recent years. In 1998 the PBC estimated the SCBs’ nonperforming loans to be 25 percent of their total loans. After the transfer of many nonperforming loans to AMCs in 1999–2000, the PBC once again released an estimate of 25 percent for the SCBs, based on the old classification system, implying a substantial increase in the estimated ratio from 1998. Following the introduction of a new loan classification system effective January 2002, the four SCBs revised their nonperforming loan ratios upward, to a combined average of 30 percent of total loans at the end of 2001.

Information on asset quality for the remainder of the financial system outside the SCBs is more sketchy and subject to even greater margins of error (again mostly on the downside). Nonetheless, this chapter attempts an overall systemic assessment based on a range of qualitative factors.

See Table 8 (p. 216) in the Almanac of China’s Finance and Banking (English edition), 2000.

Provisions averaged only 1.3 percent of loans in 1999.

Despite their financial costs, the four SCBs’ extensive branch networks—even at potentially smaller levels—carry considerable franchise value, which should support their ability to compete with foreign banks as well as their future restructuring and privatization.

Evidence from the JSCBs suggests that the more concentrated public sector ownership is (that is, the more a bank is owned by a single local government rather than by multiple state enterprises in different provinces), the more likely is the bank to experience interference in its operations (Wong and Wong, 2000).

Indirect loans include those made to SOEs, which in turn relend some of those funds to nonstate firms with which they conduct business (OECD, 2002).

The transfer removed the following from the banks’ books: those loans with no prospect of recovery as of September 1999; nonperforming loans made before the end of 1995 and loans that as of the end of 1998 were classified as “idle” (that is, over a year past due); loans extended before 1995, which will be converted into equity; and interest receivable on the transferred loans (provided it was recorded on the balance sheet).

The Bank of China also recently listed its Hong Kong SAR subsidiaries on the Hong Kong Stock Exchange.

In 2000 the number of UCCs was reduced by 1,296, to about 1,650, through conversion into city commercial banks (90 former UCCs), outright closure (56), acquisition by SCBs (114), and conversion into RCCs (1,036).

Since the GITIC closure, the authorities have emphasized negotiated settlements with foreign creditors in the 20 or so remaining TICs. The PBC has also held firm on its “no bailout” policy, leaving local governments to take charge of the restructuring and assume responsibility for the foreign debts of their TICs. Also noteworthy is the PBC’s issuance of a new regulation aimed at tightening supervision and control over the TICs by placing them firmly under PBC supervision; barring them from raising foreign debt, issuing bonds, and taking deposits; and confining the scope of their business to provision of trust services for investment funds.

Since October 1999, banks are allowed to set renminbi lending rates within a range from 30 percent above to 10 percent below the official lending rate for small and medium-size enterprises, and 10 percent above or below that rate for SOEs in urban areas; since January 1998 the corresponding range for small and medium-size enterprises in rural areas is from 50 percent above to 10 percent below the official rate.

    Other Resources Citing This Publication