See, among others, Bell, Khor, and Kochhar (1993) and Perkins (1988).

Decision of the Third Plenum of the Fourteenth Central Committee on Issues Concerning the Establishment of Socialist Market Structure.

See Sachs and Woo (1994). The authors argue that the divergent reform experiences of China, on the one hand, and Eastern Europe and some countries of the former Soviet Union, on the other, originate in China’s beginning its reform as an agricultural economy facing normal problems of economic development, as opposed to Eastern Europe and the countries of the former Soviet Union, which started as urban and industrialized economies confronted with problems of structural adjustment.

For a more detailed overview of the approach taken in Southeast Asian countries, see World Bank (1993).

For publications that cover China’s reform process in all sectors, see, for instance, Perkins (1988), Blejer and others (1991), Bell, Khor, and Kochhar (1993), Tseng and others (1994), and Fan and Nolan (1994).

Decentralization gave local authorities the leeway to experiment. However, it has also given them the possibility to “exploit” some of the old structures and, in fact, undermine central decision making. For instance, decentralization put the local authorities in a better position to influence branches of the PBC and specialized banks in promoting local projects. On several occasions, these policies undermined the national monetary policy objectives. This undermining of policy objectives in several policy areas has certainly contributed to the cyclical swings that have typified the Chinese economy since the start of the reforms.

It has become common practice to divide the reforms into phases. The first phase of economic reforms (1979—84) was aimed at allowing market forces to play in agriculture and nonurban industry and to open the economy to the rest of the world. The second phase (1984–88) included the establishment of a two-tier price system, the introduction of enterprise taxation, and the reform of the wage system. The introduction of a two-tier banking sector and the transformation of the PBC into a central bank also date from 1984, while swap centers for trading retained foreign exchange earnings were introduced in 1986. The period 1988—91 was characterized by retrenchment measures as a result of the acute inflationary pressures experienced in 1988. The rectification program reintroduced some price controls, and the pace of financial reforms slowed appreciably. Credit quotas were applied more tightly, and the freedom for banks to set their interest rates was scaled down. However, these were also the years in which secondary government securities markets emerged and the NBFI sector started growing. The fourth phase started in 1992 with the declaration ending the rectification program. The macroeconomic cycles (see also Section VII) do not coincide perfectly with these reform phases because of lagging effects, but both types of phases are interrelated.

The PBC Council was composed of the Governor and five Deputy Governors of the PCB, one Vice-Minister each from the Ministry of Finance, State Planning Commission, and System Reform Commission, and the presidents of the four state-owned specialized banks and the People’s Insurance Company of China.

Trust deposits are invested or lent at the discretion and risk of the TIC and are thus similar to bank deposits; entrusted deposits are lent or invested at the specific instruction and risk of the depositor.

The rectification program was a stabilization program introduced in the wake of a bout of inflationary pressures (by mid-1988, the 12-month inflation rate reached 30 percent). Structural reforms were given lower priority than stabilization, and administrative measures were used to supplement nascent indirect instruments of macroeconomic control. The plans for a new round of price reforms were deferred, and some earlier reforms were reversed (e.g., some price controls were reintroduced). For more details, see the “Monetary Developments” subsection of Section VII.

The paper tries consistently to use the term “state-owned specialized banks” for the period before 1992–93 and “state commercial banks” after 1992–93. In a few cases, it was difficult to draw this strict line.

For instance, the Law of the Autonomy of the Banco de España and the Statute de la Banque de France include similar provisions in this respect: “The Bank is an institution under public law with its own legal personality and full public and private legal capacity. It shall pursue its activities and fulfill its objectives with autonomy from the administration, carrying out its functions as specified in this law and other legislation” (Article 1). Also, Article 12 of the Federal Bank Act in Germany states that the Bundesbank “shall be independent of instructions by the federal Government in the exercise of the authority granted to it by this Act.” Similar arrangements had been enshrined in the Statute of the European System of Central Banks and the European Central Bank adopted under the Maastricht Treaty. This trend is also present in Latin America: in Chile, the Central Bank is an autonomous institution, with its own resources; in Mexico, the central bank was granted constitutional status and provided with full autonomy in its functions and administration; in Argentina, the Central Bank is granted the highest degree of autonomy under Argentine law. However, this is not the case in some Asian countries, such as Thailand, where the Minister of Finance has charge and control of the execution of the Central Bank Act. Moreover, the general supervision of the affairs of the Bank of Thailand is vested in the Minister of Finance.

The Bank of Japan’s objectives, as detailed in Article 1.1 of its central bank law, for instance, include not only “the regulation of the currency” but also “the control and facilitation of credit and finance, and the maintenance and fostering of the credit system, pursuant to the national policy, in order that the general economic activities of the nation might adequately be enhanced.” In the case of Korea, the primary purposes of the central bank are, according to Article 3 of its central banking law, “to maintain the stability of the value of the money in the interest of national economic progress, and further economic growth and efficient utilization of national resources by the sound operations and functional improvement of the nation’s banking and credit system.” In Malaysia, besides the promotion of monetary stability and sound financial structure, the principal objective of the central bank is to influence the credit situation to the advantage of the federation.

Among the members of the European Union, this is the case in Austria and Portugal.

In Japan, the central bank law is silent on the role of the central bank in exchange rate policy. In practice, the Minister of Finance, in close contact with the Bank of Japan, decides on exchange rate policy and instructs the Bank to intervene to prevent extreme fluctuations. In Korea, Article 43 of the central bank law explicitly indicates that the Government decides on the exchange regime, as the Bank of Korea reports annually on the “Government’s exchange rate policies.” The same situation applies in Malaysia, where Article 19 of the relevant law states that “the parity of the ringgit shall be determined by the Minister on the recommendation of the Bank.”

For instance, among the central banks of the European Union, only the central bank of Sweden has direct responsibility for defining the exchange regime. In some countries (Spain and Portugal), the central bank is consulted; in other countries (Finland), the central bank proposes to the government, or, as in Austria, it cooperates with the government. In all the other European Union countries, the exchange rate regime is the sole responsibility of the government.

Recent European experience provides examples of greater autonomy for central banks. In Spain, coordination is assured by the participation without vote of the Director-General of the Treasury and Financial Policy and the Economy and Finance Minister in central bank board sessions; a similar provision appears in the French law. In other cases, the law regulates directly the solution of potential disagreements. The Bank of Canada Act provides that the Minister of Finance and the Governor shall consult regularly. In case of differences of opinion on monetary policy, the Minister may, after consultation with the Governor and with the approval of the Governor in Council, give to the Governor a written directive, with which the Bank of Canada shall comply. In a similar vein, the Reserve Bank of New Zealand Act requires the Reserve Bank to agree with the Minister of Finance on target price increases. In some Asian countries, central bank autonomy remains more limited. In Japan for instance, although the central bank board has the sole authority over interest rate policies, the Minister of Finance can intervene if deemed necessary. In Korea, the Government may also formulate monetary and credit policies but shall consult with the central bank board.

In Canada and Germany, banking supervision is vested in a separate agency, although the Bundesbank in Germany assists the Supervisory Office in many ways. In the United States and Japan, the central banks share formal responsibilities. Meanwhile, in France, the United Kingdom, New Zealand, and Korea, for instance, the central bank has prime responsibility. The Statute of the European System of Central Banks and the European Central Bank foresees only a limited role for the central bank in the area of prudential control, as the national authorities are to remain the main players in this area, and a certain coordinating role is envisaged for the European Central Bank.

The 1994 decision to phase out PBC direct credit to the Government is now enshrined in Article 28 of the PBC law. This provision is similar to those in the laws of the central banks that enjoy the highest degree of autonomy. In the case of China, it is also important that this prohibition against lending to the Government is extended to all levels of government. This disposition, together with the dispositions enhancing the authority of the PBC over its branches, will facilitate the phasing out of (autonomous) lending by PBC branches to local governments.

Soon after World War II, a central banking system was set up in Germany consisting of legally independent regional central banks (the “Land Central Banks”), established by the State Government of each Land, and a joint, “umbrella-type” subsidiary of the Land Central Banks (the “Bank deutsche Lander”). The latter issued notes and performed other central bank functions on behalf of the regional institutions. The present Deutsche Bundesbank was established in 1957 through the amalgamation of the Land Central Banks and the Bank deutsche Lander. The Land Central Banks lost their independent legal status and became “main offices” of the Deutsche Bundesbank. However, the establishment of a centralized central bank did not preclude a permanent regional participation, as the Central Bank Council includes all the presidents of the Land Central Banks.

The United States offers a model of a rotating provincial representation. The presidents of the Federal Reserve Banks—except for the President of the Federal Reserve Bank of New York, who is a permanent member—serve one-year terms on a rotating basis as members of the Federal Open Market Committee. In Germany, therefore, regional participation takes place at the level of the Central Bank Council, whereas in the United States it takes place at the level of the Federal Open Market Committee.

Until recently, part of the country’s foreign exchange reserves were held by the Bank of China. Management of the other part was in the hands of the State Administration for Exchange Control (SAEC), which, until 1994, reported directly to the State Council. Since 1994, the SAEC has operated under the leadership of the PBC (see Section VIII).

As mentioned in Section III, during most of the reform since 1984, the four state-owned specialized banks and the PBC were politically responsible in the same way to the State Council, even though the PBC was given some supervisory powers over the specialized banks.

Government bonds are the underlying instruments of these repurchase agreements. Turnover was ¥ 1,832.7 million in December 1994, which is approximately one fourth of the turnover in the secondary government securities market in the same month (World Bank (1995), p. 19).

Among these 44, 6 centers (Shanghai, Wuhan, Beijing/Tianjing, Shenyang, Xian, and Chongqing) had a regional significance.

Thus, from a nationwide point of view, the interbank market could not play a redistributive role. Redistribution of funds between banks (or regions) with excess liquidity and those with shortages actually was done through the PBC on the basis of its lending to the banks and its reserve requirements (see Section VII).

The difference in growth and volume between borrowing and lending in the interbank market (see also Table 2) is due to classification problems experienced by the Chinese authorities. Part of the missing funds reflect lending from state bank branches to affiliated NBFIs not reported in the PBC’s statistics. In addition, some enterprises have direct access to the interbank market, a flow that is also not captured in the available statistics.

The procedures for issuance and redemption of treasury bonds to individuals are quite complex. Each year, during the budget-planning process, the Ministry of Finance estimates the volume of securities that can be issued based on an analysis of household income and savings and, for enterprises, their financial position. Local governments bear responsibility for organizing and promoting the bond sales in their jurisdictions. The PBC’s network is in charge of delivery and distribution of the bonds to the branches of the specialized banks that participate in the sales at the retail level. Actual sales of treasury bonds begin in early April of each year and can take as long as six months to complete. However, all bonds start earning interest on July 1.

The treasury bonds issued to enterprises only take the form of a receipt of purchase.

See also McKinnon (1993).

Procedures are also complicated by the wide range of denominations for household-held bonds, which run from a minimum of ¥ 5 to a maximum of ¥ 100. This places heavy administrative burdens on entities involved in issuing, redeeming, and trading the bonds. In addition, all household-held bonds are issued in the same period of the year and, like all other treasury bonds, mature on July 1.

The indexation mechanism ensured that the real interest rate on these bonds remained close to zero. The mechanism also applied to bank deposits with maturities of over three years. It was reintroduced in 1993, when the inflation rate again exceeded the interest rate on these financial instruments.

The April 1991 underwriting was jointly organized by the Ministry of Finance, the State Commission for Restructuring the Economic System, the Stock Exchange Executive Council, and the PBC. The syndicate was composed of 58 financial institutions selected from among numerous applicants on the basis of experience in handling securities, the institution’s number of counters and trading desks throughout the country, and its overall size.

The main reasons for this disappointing performance seemed to be that (i) the issue came too soon after the initial six-month issue, (ii) the maturity term offered was longer than suited investors’ preferences, and (iii) the size of the offering made it difficult to place.

The city of Shanghai permitted SOEs to issue shares from 1984 onward.

Rates were as high as 20–40 percent in 1985–86 and even ranged between 50 percent and 100 percent in 1988 (Goldstein, Folkerts-Landau, and others (1994), p. 93).

This market was fed mainly by people who were forced to sell their securities. These securities were bought at steep discounts by speculators who redeemed them at maturity for substantial profits (Goldstein, Folkerts-Landau, and others (1994), p. 94).

The establishment of these stock exchanges had been preceded by a few local experiments: in Shanghai, for example, the first official secondary market for equity was opened in 1986, starting with one trading center (organized by the Shanghai branch of the ICBC). By the end of 1988, there were 33 trading centers in the city, trading seven enterprise shares and eight bonds (Goldstein, Folkerts-Landau, and others (1994), pp. 94–95). Another experiment, not officially recognized by the central authorities, took place in Shenzhen. An over-the-counter (OTC) market emerged in 1988, mainly trading shares of the Shenzhen Development Bank. By 1990, several other shares and bonds were traded in this market, leading to a significant increase in volume that the OTC market ultimately could not handle, owing among other reasons to inefficiencies in trading practices and to a lack of appropriate regulations. The latter, in turn, led to tax evasion, insider trading, and the emergence of a curb market. In December 1990, the municipal government opened a centralized trading center that began trading in equities.

In the first year, about 200,000 debit cards were issued, with an annual turnover of ¥ 3.3 billion (World Bank (1990)).

Financial institutions have no direct access to the system. Their interbank transactions are serviced by and through PBC branches. In general, each branch of each financial institution has two accounts at the PBC branch at the same administrative level: a required reserve account and an excess reserve account. The latter is used for payments and settlement purposes.

The PBC uses the term “credit quota” instead of credit ceilings.

The specialized and universal banks are responsible for over 80 percent of the total credit in the economy (see Chart 4).

Approval of NBFIs to borrow from the PBC is based on both balance sheet and profitability considerations. Approved NBFIs that wish to borrow from the PBC are required to open settlement accounts with the PBC. NBFIs can only borrow from the PBC twice a year.

Annual quotas are based on the market share of financial institutions, as well as on the extent to which they engage in policy lending. The PBC can adjust its lending quota for a specific bank or adjust the allocation of its lending among banks, as necessary, during the year, taking into account the growth of deposits relative to credit ceilings of specialized banks, as well as the overall liquidity situation of the financial sector and the changing needs of priority sectors of the economy.

Since then, reserve requirements have been refined to cover demand and savings deposits of specialized banks and to exclude fiscal deposits. Interbank deposits are also excluded from reserve requirements. Trust deposits of other financial institutions with specialized banks are subject to reserve requirements.

Required reserves of RCCs are deposited with the Agricultural Bank of China (ABC). These deposits with the ABC are, in turn, subject to reserve requirements. UCCs deposit their required reserves directly with the PBC. The reserve ratio for UCCs, however, differs somewhat from other banking institutions; it is in the range of 10—20 percent of deposits, at the discretion of individual cooperatives.

Interest rates are administered by the PBC, but changes in the rates need prior approval by the State Council.

That is, if the administered rate is 10 percent, banks can charge between 10 percent and 11 percent.

Only UCCs and RCCs were allowed some flexibility in setting deposit rates within prespecified margins.

In practice, in northeastern and northwestern China, where there are more loss-making SOEs, banks charge the official rates; in the coastal areas, banks usually lend at rates close to the ceiling of the range or even at rates above the permitted floating range.

Since then, the indexation mechanism has gone into effect every time that the inflation rate exceeds the three-year deposit rate. The mechanism broadly ensures that the real rate is zero over the savings period. A similar scheme applies to government bonds with initial maturities of three years or more.

Open-market-type operations refer to the use of primary issues of securities through auctions of central bank or government deposits at the initiative of the central bank to influence monetary conditions in the financial markets.

UCCs, RCCs, and small commercial banks were requested to deposit part of their free reserves in special deposits at the PBC. In 1994, a total of ¥ 14 billion was collected in one-year time deposits, yielding an interest rate of 12.6 percent. In this connection, the PBC’s provincial branches were assigned quotas to be collected within their jurisdictions.

See Khor (1991) and Bell, Khor, and Kochhar (1993).

PBC branches had to follow strict rules and guidelines in implementing the credit plan, and Huang (1994) reports that strict discipline and punishments were imposed for failures to achieve the binding plans.

In addition, however, as can be seen from Chart 9, the austerity measures led to the quick reconstitution of total reserves (to 25–30 percent of total deposits by mid-1990, similar to the level attained in 1985).

Excess reserves are defined here as the sum of the banks’ required excess reserves (5—7 percent of their deposits to be held at the PBC) and any free reserves that they have over and above these required amounts (basically “free” deposits at the PBC and vault cash).

This analysis disregards the behavior of “other items, (net)” on the PBC’s balance sheet. The erratic fluctuation of this aggregated component means that any attempt to analyze the sources of banks’ reserves can only be approximate.

See also Khor (1993) for an overview of the swap centers and their contribution to the economy.

In an open outcry arrangement, buyers and sellers enter their bids and offers on computer terminals that are networked. Members can enter only what they have customer orders for, but they are not compelled to do so, and they can choose at what time during the session they make the entries. Entries can also be changed at any time. All participants (and observers) can see the entries made by the other participants on their screens, as well as on a large wall screen. Each bid or offer is identified by dealer. Clients may watch the proceedings from behind a barrier and pass instructions to their brokers. Brokers also stay in touch with their clients through telephones. In the Chinese version of this arrangement, there were 60 dealing positions, but only a fraction of the capacity was used on a typical day.

Lower-priced selling orders had priority over higher-priced selling orders; higher-priced buying orders had priority over lower-priced buying orders. For orders at the same price, priority was based on time of data entry; early entry had priority over late entry. The computer system identified the lowest-selling and highest-buying orders that could be matched and highlighted them on the screens for 30 seconds. If neither party wished to change its orders, or if no orders at a better price were submitted within the 30 seconds, the deal was struck. A digital clock on the screens showed the time remaining of the 30-second period. Deals done were also shown cumulatively on the screens, together with information about prices. Deals were struck throughout the trading day in this manner, as dealers adjusted their prices on the basis of their objectives and the observed transactions.

From the market participants’ point of view, they have only one counterpart in the market, the CFETS. The normal evaluation of counterpart risks and credit limits has no practical role in the current system, which is totally impersonal, providing no information to participants about the identity of other market participants.

During the sessions, therefore, it is difficult or impossible for banks to monitor and control the operations of their traders.

If larger changes in the daily rate were allowed, arbitrage opportunities might arise between the CFETS and retail markets.

In general, and compared with many other countries that embarked on broad-based financial reform processes starting from high degrees of financial repression, China’s interest rate policy record has been encouraging. During most of the period since 1978, interest rates have been positive in real terms. Only during some episodes of high inflation (1988–89 and 1993–94) have real interest rates turned negative (although indexation schemes reduced the adverse impact of inflation on savings). Even in those periods, the discrepancy, as measured by international standards, was not dramatic. During periods of low inflation, administered interest rates must have been close to market equilibrium level. The favorable assessment of China’s overall record on interest rate policy is also supported by the country’s continuously high savings ratio. Distortions usually introduced through interest rates that are administratively set below their competitive free market equilibrium level have been less significant than in several other countries.

However, the proportion of market interest rates remains small, as evidenced by the large share of credit extended by the state banks (some 80 percent).

It is estimated that the SOE sector still accounts for about 45 percent of industrial production and one third of GNP.

An estimated one third of SOEs are incurring operating losses—accounting for about 6 percent of GNP—that are covered by direct budgetary subsidies and bank lending.

The SOEs are estimated to absorb over 80 percent of total bank credit to the economy.

Policy lending banks are intended to receive capital from the Government, budgetary appropriations, and bond issues. So far, the bulk of the resources have been raised from bond issues, which were allocated to the four state commercial banks according to their share in total deposits, and from PBC lending (in the case of agricultural procurement loans).

According to the authorities, the state commercial banks currently satisfy the “core” capital ratio of 4 percent but fall short of the “second-tier” capital ratio requirement of 4 percent.

Provisions for loan losses are to increase from their 1995 annual average level of 0.6 percent of outstanding loans by 0.1 percent a year until reaching 1 percent of the loan portfolio in 1998.

In Malaysia and Thailand, measures were taken to ensure transparency in the market and thus fair competition. Banks in the two countries are required to post what is referred to as a “minimum reference rate” in Thailand and a “base lending rate” in Malaysia. These reference rates are based, for each bank, on its average cost of funds and indicate the reference lending rate of that bank for prime-quality customers. In Thailand, the minimum reference rate includes four components: the average cost of funds, the operating cost, a tax levied on financial transactions, and a normal profit margin set at 2 percent. See Mehran, Laurens, and Quintyn (1996).

At the end of September 1994, the PBC’s claims on the banking system amounted to ¥952.8 billion and exceeded the banks’ required and other deposits at the PBC, which stood at ¥590 billion.

In general, country experiences with refinance instruments indicate that the choice between instruments to provide liquidity to the system as a whole, either through a discretionary and competitive mechanism—at the initiative of the central bank—or through instruments that work on a bilateral basis and at the discretion of the commercial banks—such as a discount window—is to a large extent dictated by the degree of development and liquidity of the money markets, especially the interbank market.

The development of a framework to manage the supply and demand for reserve money at the level of the central bank’s balance sheet, that is, a reserve money program, is discussed in the next subsection.

The stipulation in the commercial bank law that a head office has authority over its branches should form the basis for arrangements whereby branches can conduct interbank operations by delegation from headquarters, thus ensuring that the bank’s creditworthiness backs each and every transaction at the branch level.

The characteristic of stimulating small institutions alongside large, state-owned ones is also found in other sectors of China’s economy. See, for example, the development of the small and collectively owned township and village enterprises versus that of the large state-owned enterprises.