March 2, 1999

Wing Thye Woo

University of California, Davis

The economic success of China in the last two decades is well-known, and so are its causes. The post-1978 deregulation of China’s economy, its integration into the world trade and financial systems, and the harmonisation of its economic institutions to those of a normal market economy have allowed market-directed economic development to draw China’s vast surplus agricultural labor into the industrial and service sectors, make China the biggest recipient of foreign direct investment (FDI) in the world, and render China the world’s most competitive exporter of labor-intensive manufactures. The outcome of these economic changes is that China has quadrupled its per capita income over this twenty-year period.

These are impressive achievements, but there are still many difficult structural reforms (e.g. ownership reform, legal reform) that need to be completed before China can join the ranks of its richest neighbors, not to mention the G-7. Besides wrestling with the formulation and implementation of the complicated structural reforms, China is now facing extreme problems in short-run macroeconomic management. Economic growth has slowed significantly in the last two years and is generally predicted to slow further in 1999, the urban unemployment rate has jumped to double-digit in some parts of the country, and income in the countryside is stagnating. In response, the government is implementing a wide-ranging reflation program in earnest to restore economic growth – but will the recovery be a self-sustaining one?

At the present, the government is also undertaking drastic measures to maintain the viability of the existing peg of Renminbi (RMB) 8.3 per US$. Beside implementing the conventional array of policies to improve the balance of payments like ordering the state banks to make more credit available to exporters, the government has also used administrative measures like ordering the military to turn over the management of its commercial and industrial enterprises to the State Economic and Trade Commission in order to curb unauthorised capital outflows and smuggling. Will the defense of the existing exchange rate work, and even if it works, is the defense desirable in the first place? And, more generally, will China be able to escape the Asian financial crisis?

We are cautiously optimistic in our assessment of China’s ability to address its short-run macroeconomic challenges. Our assessment of China’s economy is organised as follows. We first outline the current macroeconomic situation and the thrust of present macroeconomic policies, and then attempt answer some questions about the effectiveness of the reflation package, the possibility of China being part of the Asian financial crisis, and the likelihood of a Renminbi (RMB) devaluation.

The Retail Price Index (RPI) has fallen every month since October 1997 (except in August and September 1998 when the heavy flooding disrupted supplies in several heavily-populated parts of the country) with the biggest drop in 1998:3Q at the annualized rate of negative 3.5 percent. The broader Consumer Price Index (CPI) did not decline until 1998:2Q, and it fell an anuualised 1.4 percent in 1998:3Q. These price developments are in line with the steady decline in inflation since 1994, which is, in turn, in line with the steady decline in real GDP growth since 1992. GDP growth was 8.8 percent in 1997 down from 9.5 percent in 1996. The just-released news from the State Statistical Bureau that the 1998 growth rate was 7.8 percent is the latest confirmation of the severity of the economic downturn in China. Many observers are expecting further decline in the growth rate, for example, the IMF predicted in December 1998 that the 1999 growth would be 6.6 percent.

Since the 1996 growth rate of 9.5 percent about equalled the average annual growth rate of the 1978-95 period, we could say that the Chinese stabilisation program that started in mid-1993 had achieved soft-landing in 1996, with CPI inflation in 1996 down to 8.3 percent from 24.2 percent in 1994. Some observers have termed the continuation of tight macroeconomic policies until early 1998 to be a case of "macroeconomic policy overkill" because of the fall in the level of retail prices since 1997:4Q, and the precipitous plunge in reserve money growth from an inflationary 43 percent in 1993 to 17 percent in 1997, and then to 11.2 percent in 1998:3Q. We do not dispute the macroeconomic consequences of the tight monetary policies and the tight controls over investment spending, but we note that these restrictive policies had succeeded in forcing considerable restructuring in the inefficient state-owned enterprise (SOE) sector. Because most loss-making SOEs did not receive their accustomed allotments of credit to continue production (a large portion of which went straight into inventory), the default outcome was that many were taken over by new owners who reorganised the firms and changed the output mix. Our point is that a temporary slowdown in growth is many times necessary in order to force resources to move to a new growth path that would lead to a more competitive economy in the future. We have to recognise in the so-called "macroeconomic policy overkill" the audacity of the top Chinese leadership which has chosen dislocating reforms which would produce sustained dynamic growth in the future over Brezhnev-style maintenance of the comfortable status quo which ensures a dismal future.

The "macroeconomic policy overkill", in short, reflected a deliberate decision to accept growth rates that are somewhat lower than the 9.5 percent average growth rate of the 1978-95 period in order to ensure an acceptable rate of economic restructuring and to moderate the boom-bust cycles of the last two decades. The implicit growth range that policymakers appear to think is compatible with achieving the restructuring and stabilisation objectives seems to be about 7.0 percent to 8.5 percent. When the Asian financial crisis hit in 1998, causing China’s exports to stagnate, and hence rendering growth lower than intended, it was only natural that the government undertook stimulation of domestic demand to reflate the economy.

The government responded to the onset of price deflation in 1997:4Q by cutting the average lending rate from 10.1 percent to 8.6 percent. However, the anticipated surge in credit expansion did not occur. This is largely because of the new-found reluctance of the state commercial banks to extend more credit to its traditional clients, the SOEs -- especially the loss-making SOEs.

This reluctance by banks to extend credit had at its origin in the determined efforts of Zhu Rongji to improve the balance sheets of the state banks and to promote restructuring in the SOE sector since he took over as economic czar in mid-1993. By the end of 1997, the twin facts that Zhu Rongji would be promoted to become the Prime Minister in 1998 and that he had peremptorily dismissed bank managers when the proportion of NPLs in their banks had gone up had instilled a new sense of prudent lending in the entire state bank system. The slowdown in economic activity after 1997:3Q, and hence the decline in profitability, has only heightened this new sensibility of prudent lending among state bank mangers. So despite several more reductions in interest rates, money growth (at least until the middle of 1998) has continued to drop in line with the decline in GDP growth.

By early 1998, in the wake of the collapse of several important Pacific Asian economies, Chinese policymakers recognised that stronger reflation was required to offset the coming collapse in external demand. Furthermore, the SOE reform program announced at the 15th Party Congress in September 1997 was beginning to take bite and firms would soon begin to shed excess workers. So, stronger reflation was also desirable in order to induce the establishment of new urban enterprises to soak up the newly released SOE workers.

The reflation program that is now in effect seeks to boost aggregate demand by trying to:

(a) increase investment by approving faster the backlog of investment applications;

(b) increase government spending;

(c) loosen monetary policy; and

(d) stimulate private spending through housing reform.

The State Planning Commission was literally put on an over-time schedule in early 1998 to speed up the approval of investment projects. "Increased economic openness" was a fortuitous byproduct of this measure. Approval was given to a number of large foreign projects that had been held up for several years because of concern either about the possible domination of these particular lines of business by foreign firms or about the possible competition that they might provide to domestic firms of national strategic importance.

One unexpected check on approval acceleration as a reflation tool was that many local governments had not bothered to turn in local investment plans for 1998 because of the across-the border rejection of local investment plans since the earnest implementation of the stabilisation program in 1994. Of course, the greatest obstacle to the effectiveness of investment approval as an economic stimulus is that approval does not necessarily translate into realization. The translation of approval of investment into realization of investment is especially uncertain in times of low aggregate demand like the present. Firms, foreign and domestic, might well choose to postpone the actual investment until sustained economic recovery seems imminent. Partly, because of the low aggregate demand in China and abroad, but mostly because of the panic in international credit markets, actual FDI in 1998 was 11 percent (US$5 billion) lower than in 1997 despite the "increased economic openness" noted above.

Vice-Premier Li Lanqing created a stir at the 1997 Davos meeting of the World Economic Forum when he announced that China would spend US$750 billion on infrastructure spending over the 1998-2000 period. It was clarified later that the US$750 billion figure included investments of state industries and other types of state-related investments that are not normally classified as infrastructure spending. Conventionally-defined infrastructure projects would amount to US$370 billion in the 1998-2000 period. This would raise infrastructure spending to 12.5 percent of GDP in 2000 up from 7.0 percent in 1996, and the proportion of infrastructure spending in total fixed asset investment to 30 percent from the existing 20 percent.

In July 1998, the government announced the issuance of RMB 100 billion in bonds to finance new infrastructure investment by the central and local governments. (It seems that these bonds had been purchased mainly by state banks.) This announcement was quickly followed by new spending plans on telecommunications, railways, and roads.

A natural question raised by the recent expansionary fiscal policy is whether the level of public debt in China is still at a level that would not be too heavy a burden in the future. The issue is what should be counted as "public debt" when so much of the economy is still state owned. If public debt is defined to be the stock of government bonds that has been issued to finance budget deficits, (and held by both domestic and foreign agents), then the public debt-GDP ratio was 7.3 percent of GDP in 1996 and 8.1 percent in 1997.

It has been argued, however, that since the government is the guarantor of the state banks, the nonperforming loans of the state banks ought to be counted as public debt. Estimates of the extent of nonperforming loans range from 20 percent to 50 percent of total bank loans. If we take the NPL ratio to be 33 percent, then the broader definition of public debt would put the "broader public debt"-GDP ratio at 37.0 percent of GDP in 1996 and 41.1 percent in 1997.

What about the debt of SOEs and other state institutions (for example, the regional trusts and investment companies, TICs)? The government could be construed as being responsible for these bad debts just as they were construed to be responsible for the bad debts held by the banks. Since the bulk of the domestic borrowing of SOEs and state institutions is from the state banks, the inclusion of nonperforming loans of the state banks in the broader definition of public debt has already taken into account the bad debts of SOEs and state institutions that are owed to domestic agents.

Foreign debts of SOEs and other state institutions may deserve different treatment from their domestic debts because of the government’s great concern about China’s continued access to international financial markets at favorable interest rates. In order to arrive at the "broadest" definition of public debt, we take into account all the bad debts that SOEs and other state institutions could potentially owe to foreigners. We constructed the "maximum" public debt as the sum of the broader public debt plus the entire foreign debt of SOEs and public institutions. The "maximum public debt"-GDP ratio was 50.1 percent of GDP in 1996 and 55.1 percent in 1997.

Is a debt-GDP ratio of 55.1 percent too low or too high? Compared with the Italian, Swedish and U.S. situations where central government debt (after deducting intragovernmental debt) to GDP ratios were, respectively, 117.6 percent in 1995, 70.8 percent in 1995, and 50.5 percent in 1996, it might appear that there is still substantial room for the Chinese government to increase its borrowing to finance its expansionary fiscal policy without causing serious debt problems in the future. However, such conclusion would be overly optimistic. This is because China raises much less state revenue (as a share of GDP) than these other countries, and hence has a much lower capacity to service its public debt. The revenue-GDP ratio was 11 percent for China in 1995, 30 percent for Italy in 1995, 38 percent for Sweden in 1995, and 21 percent for the U.S. in 1996. The point is that until China increases its tax collection, there is a real tradeoff between restructuring the state financial sector and increasing infrastructure investment to stimulate the economy. And it is important to note that increasing tax collection is as much a political challenge as it is an administrative challenge.

The People’s Bank of China has cut interest rates several times since price deflation became obvious. For example, the bank lending rate on loans with maturity from 3 to 5 years was reduced from 11.70 percent to 9.90 percent in October 1997, to 9.72 percent in March 1998, then to 7.65 percent in July 1998, and most recently to 7.15 in December 1998. Furthermore, the bank reserve ratio was lowered to 8 percent from 13 percent in March 1998. Throughout 1998, various high-level officials (including directives issued by the People’s Bank of China) have called upon the domestic banks to expand their lending; and the banks, in turn, have pledged to increase credit to various segments of the economy. However, until the second half of 1998, money growth has continued to be slow, prompting some Chinese economists to be alike their Japanese colleagues in postulating the existence of liquidity traps.

As explained earlier, the main reason for the slow credit growth lies on the supply-side. Given the personally severe consequences from an increase in the ratio of NPLs, the typical bank manager is justifiably wary of increasing credit to his traditional SOE clients. However, the state banks are also reluctant to lend to new clients that are non-state enterprises, partly because the latter’s non-standard accounting makes risk assessments difficult. More importantly, a banker knows that while a NPL to an SOE is financially undesirable, a NPL to a private enterprise is more than that, it is also politically undesirable. The outcome presently is that the loans that state banks are most willing to make are infrastructure loans guaranteed by the central government.

Hence, the practical short-run solution to this "liquidity trap" is for the government to undertake new infrastructure spending financed by the state backs (and ultimately by new reserves from the central bank). However, a larger sustained increase in credit is possible only if the state commercial banks would use the new deposits (new reserves) to extend new loans, i.e. only if banks act according to the standard "money multiplier" process. As the banks’ willingness to lend depends now on finding truly economically viable projects, the government has sought to create new safe lending opportunities to the banks by announcing housing reforms, including privatisation of the housing stock. The hope is that the banks would then expand mortgage lending on the basis that the household debt would be fully (and, presumably, also safely) backed by a marketable asset, and hence boost aggregate demand.

The majority of the urban population lives in virtually free housing supplied by their employers. In early 1998, the government announced that SOEs and other state institutions would stop providing free housing after July 1 and that the housing stock would be privatised. The marketization of housing is now in full swing, marking another significant milestone on the way to a market economy. The marketization of housing will enhance labor mobility and free the SOEs to focus on production and distribution of goods.

The China Macroeconomic Analysis (1998:3Q issue) has estimated that, with a functioning mortgage system in place, the marketization of housing would increase the annual demand for housing by 20 to 30 percent. Since housing investment is presently about 4.3 percent of GDP, the housing reform would increase GDP growth by 1 percentage point.

However, in our assessment, the short-run result of the housing reform was a decrease in aggregate demand even though the new steady-state level of housing demand under the market regime is higher than the old steady-state level of housing demand under the entitlement regime. First, the demand for new residential construction by SOEs stopped abruptly on July 1, 1998, and because it takes time for private agencies to appear to intermediate between the builders and the millions of disparate buyers, the immediate impact was more likely to have been a drop in housing demand than an increase.

Second, the mortgage system is not yet in place. The banks need time to build up its expertise in mortgage lending, and the certification/registration system of house ownership is usually not standardized province-wide. More importantly, at the moment, only the richest 5 to 10 percent of the urban population can qualify for mortgage loans; and these well-to-do folks are likely to have already acquired the most of the housing that they want.

The reflation package is working much better than expected by most observers. When GDP grew only 7.0 percent in the first half of 1998, most observers revised their forecasts of 1998 growth to be below 6.5 percent. When the strong third quarter growth pulled the annual growth rate up to 7.2 percent, the IMF revised its prediction of 1998 growth to 7.2 percent from 5.5 percent. The announcement by the State Statistical Bureau on December 30, 1998 that actual growth was 7.8 percent in 1998 has confounded many observers (see Table 1).

(EIU=Economist Intelligence Unit, IMF=International Monetary Fund)

EIU 1997:2Q 8.6

EIU 1997:3Q 8.4

EIU 1997:4Q 8.0 8.5

EIU 1998: 1Q 7.3 8.2

EIU 1998:2Q 6.7 7.6

EIU 1998:3Q 6.1 7.0

EIU 1998:4Q 7.8 6.7

IMF October 1998 5.5

IMF December 1998 7.2 6.6

Exports in 1998 were almost the same as in 1997, but imports fell by 3.8 percent to produce a record trade surplus of US$45 billion, an increase of $4.7 billion from 1997. The primary cause of the high GDP growth in the second half of 1998 was the 26 percent surge in state investment in the third and fourth quarters.

It must be mentioned that a number of observers believe that the official growth numbers are wrong and that actual growth in 1998 was between 3 to 5 percent. The basis of this skepticism is the low usage of electricity, the low volume of goods being transported, and the continued fall in the level of retail prices. The well-known Chinese economist, Mao Yushi, was quoted as saying that: "The GDP figure is still dubious… There must be some local government trying to please the central government by reporting inflated statistics." There is credibility in Mao Yushi’s statements because Premier Zhu had criticized provincial leaders in early December for each reporting a provincial growth rate greater than 10 percent in the first half of 1998 when the national growth rate was only 7.2 percent.

In other work, we had found that the annual GDP growth rate in the 1985-93 period could have been overstated, on the average, by as much as 2 percentage points; and, after taking various factors into account, we had suggested a downward correction of about 1 percentage point. In light of our work, and the skepticism expressed in the preceding paragraph, we think that the actual 1998 GDP growth rate could plausibly be around 6.5 percent.

We think, however, that even if the official 1998 growth rate is overstated, there can be little doubt that the reflation program has indeed arrested the decline in growth rate. It is likely that the growth momentum will carry on to 1999, producing growth of about 7 to 8 percent.

The Asian financial crisis is typified by:

1. a collapse of the exchange rate because of heavy capital outflow, and
2. a collapse of the domestic financial system causing a shortage of working capital that, in turn, caused output to collapse.

A dramatic speculative attack on the RMB can be ruled out simply because the RMB is not convertible for capital account transactions in financial assets. It is difficult for a person to borrow RMB from a Chinese bank to buy US dollars to speculate against the exchange rate because the purchase of US dollars requires documentation to prove that the transaction is trade-related.

Capital outflow by foreign private agents has not occurred because most of the foreign private investments in China are foreign direct investments, and there is very little short-term foreign debt. The total foreign debt of China was US$129 billion in 1996, and short-term foreign debt was only US$25 billion, less than 20 percent of total foreign debt. The fact that China also has US$143 billion in foreign exchange reserves make defense of the exchange rate feasible even if all short-term foreign debts are recalled.

Furthermore, foreign participation in the Chinese stock markets is limited to transaction in B-shares. Only foreigners can own B-shares, and B-shares are denominated in US dollars and transacted using US dollars. In short, an abrupt withdrawal by foreigners from the Chinese stock markets can affect the value of the yuan-dominated A-shares (that only Chinese can own) only if their withdrawal would cause Chinese speculators to revise their expectations of future Chinese growth downward.

Of course, capital flight can occur through channels like over-invoicing of imports and under-invoicing of exports. A successful speculative attack on the RMB via large and pervasive mis-invoicing is theoretically possible, but difficult to prove because the paper trial would point to trade imbalance rather than portfolio adjustment being the cause of the exchange rate collapse. An exchange rate collapse from mis-invoicing of trade requires that the government be rigidly commited to current account convertibility, but this is not credible. Any government like China that has in place a comprehensive administrative system that processes every import application to buy foreign exchange (in order to prevent capital movements) can be easily tempted to defend the exchange rate by delaying approvals of import applications. So imports could be compressed to a significant degree whenever a trade deficit threatens to materialise.

We turn now to the issue of whether China’s banking system would collapse spectacularly as in the countries experiencing the Asian financial crisis. To a first approximation, when the won, baht, and rupiah went into free fall, many Korean, Thai and Indonesian banks were rendered insolvent through a combination of the following channels: the sudden increase in the value (measured in domestic currency) of their foreign liabilties; the default on bank loans by domestic corporation bankrupted by the soaring of their external debts; and the default on bank loans by exporters who could not get short-term credit from their foreign suppliers of inputs. Many of the Korean, Thai and Indonesian banks were already financially fragile before their collapse because of undercapitalisation, and because of considerable NPLs that had been hidden by accounting gimmicks. And the exchange rate shock pushed these fragile banks over.

Much alarm has been raised in recent months about the amount of NPLs in China’s banking system, with estimate for NPLs ranging from 20 to 50 percent of total bank loans. It has even been raised as a serious possibility several times, that a run by depositors is almost inevitable, causing a banking collapse that would trigger a general output decline.

We find the likelihood of either a bank run or a collapse of the banking system to be minimal. Admittedly, there have been bank runs in China since 1978, e.g. in 1988. But these bank runs were motivated by aniticpations of high inflation caused by imminent lifting of price controls, and not by anticipations of bank failures. Whenever the government began indexing interest payments to the inflation rate, the bank runs reversed themselves. In the present time of falling prices, inflation-induced bank runs will not occur.

It is true that there is no depositor insurance in China but this in itself is unlikely to cause a bank run induced by fear over the large amount of NPLs. This is because all but one of the banks are state-owned and the government has repeatedly pledged to honor all deposits in the state banks. This pledge is credible because the government is in a position to make good its promise. As pointed out earlier, the government can easily borrow to cover the NPLs; and assuming an NPL ratio of 33 percent, the borrowing would raise the public debt-GDP ratio to just 40 percent. Alternatively, the government could always raise taxes to cover the NPLs.

Even if a bank run does occur, there need not be a collapse in bank credit because the central bank could just issue currency to the state banks to meet the withdrawals. This expansion of high power money cannot be easily translated into a loss of foreign reserves because capital controls are in place. This expansion of high power money will also not have much impact on inflation because this is mainly a shift out of bank deposits into cash, and not a shift into goods.

Simply put, even if the state banks are truly insolvent as has been alleged, and even if the insolvency does induce bank runs, a collapse in bank credit does not have to follow. It is well within the technical ability of the government to accommodate the bank runs, and it is also well within the financial ability of the government to recapitalise the state banks. Furthermore, these two government actions would not cause much damage (if any) to the economy, like lower growth and higher inflation.

While China can prevent the NPLs of the state banks from maiming the payments system and crippling production, we recognise that the NPLs have imposed real costs on the economy. With NPLs accounting for a third of total bank loans (our estimate), bank loans accounting for about a fifth of fixed investments since 1985, and fixed investments at about 35 percent of GDP, this means that about 2.3 percent of GDP has been wasted annually in the last decade. Moreover, since most of the bank loans are extended to SOEs with little going to the more efficient non-state sector, the performing loans are not in investments with the highest rates of return. In short, the productive capacity of the economy could be higher than what it is.

Of course, we also recognise tha the NPL problem might be even worse at the non-bank financial institutions (NBFIs) like the regional trust and investment companies (TICs). However, because NBFIs consitute only a small part of the national credit system, their failure is not capable of bringing down the payments system. The biggest dangers from the collapse of NBFIs are social instability (especially when the base of NBFIs is small depositors), and reduction in foreign credit.

In the recent closure of the financial arm of the Guangdong International Trust and Investment Company (GITIC), the central government assumed responsibility for all properly-registered foreign debt. Since trade-related credit with maturity of less than three months and foreign debts of GITIC’s branch in Hong Kong did not require official registration, it is likely that a very substantial amount of GITIC’s foreign debt will not be assumed by the Chinese government. The latest report put GITIC’s debt to foreign banks at US$10 billion.

As discussed earlier, this assumptionof all the properly-registered debt of state institutions and SOEs would raise the public debt-GDP ratio to 55 percent – still a very low level when compared with the public debt-GDP ratios of most Western European countries. As a general principle, the government’s decision to let NBFIs fail is important to reducing the moral hazard problem inherent in supervision of the financial sector. Both domestic depositors and foreign creditors have to be encouraged to assess and mange risks better.

As things stand at the end of December 1998, it looks unlikely that China will succumb to a financial crisis in 1999 marked by bank runs, capital flight, a severe shortage of working capital, and a deep recession. This being said does not imply that the present macroeconomic policies can and should be maintained in the medium-term, and this is especially true in the management of the exchange rate.

China had a dual exchange rate regime until January 1, 1994. There was an official exchange rate and a swap exchange rate. The swap exchange rate was determined by the tranactions of trading companies that were authorized to buy and sell the portion of foreign exchange earnings that exporters were exempted from selling to the central bank at the official exchange rate. The proportion of foreign exchange transactions that went through the swap market was allowed to increase steadily in 1992-1993, producing incentive effects similar to a RMB devalutaion. When the two rates were unified in 1994 at the then-existing value of the swap rate of RMB 8.4 per US dollar, the official exchange rate was RMB 5.8 per US dollar.

Since the Malaysian, Thai and Korean currencies had either stayed pegged to, or appreciated slightly against, the US dollar in the 1992-96 period, China’s exchange rate adjustments had increassed the competitivenesss of Chinese exports vis-a-vis these countires. However, the dramatic fall of the ringgit, baht, won and rupiah in 1997-1998 has more than reversed China’s gain in export competitiveness. The question that is now haunting many Asian capitals is whether China would devalue in 1999 to regain export competitiveness.

China’s exports in 1998 are almost the same as in 1997, and 1997 exports were 20 percent higher than 1996 exports. However, the fact that 1996 exports were only 1.6 percent higher than 1995 with the exchange rate unchanged the whole time should caution one against an overly alarmist reading of the 1998 export stagnation.

Premier Zhu certainly does not seem alarmed. He announced in late November 1998 that "China will not devalue its currency in 1999." Is this a credible promise?

It is a practical fact of life that (beside love) there are few unconditional committments. We think that China’s committment is conditional upon a number of calculations:

• small effects on China’s trade balance upon normalcy returning to the credit markets and recovery beginning in the crisis countries;
• no large depreciation of the Yen and the New Taiwan Dollar;
• China’s capital controls would continue to be effective;
• the possibility of renewed financial panic.

The first calculation is based on the expectation that the international financial markets would calm down substantially in 1999 and that the crisis countries would begin recovery soon. There would not only be full resoration of short-term trade credit to Korea and the Southeast Asian economies but also the return of some long-term capital. These financial inflows would enable exporters to import the inputs they require, investment to increase, and the strengthening of the won, baht, and rupiah that started in 1998:4Q to continue into 1999.

The above developments have two opposites effects on China’s trade balance. China’s exports to the crisis countries had plunged in 1998, and they will probabaly rebound rapidly with recovery in these countries. However, China’s exports to the U.S. and other developed markets would be displaced by cheaper goods from the crisis countries. But this displacement of China’s exports would be limited by the appreciation of the currencies of these countries (albeit not back to the pre-crisis level), and by the vertical integration between the foreign-owned and joint-venture production facilities in China and the distribution networks of the parent companies in the developed markets. According to the calculation of some Chinese economists, the overall outcome from these two opposite effects could be a smaller trade surplus for China, but certainly not a huge trade deficit.

The second calculation which the no-devaluation pledge is conditional on is that there would be no large depreciations of the Japanese Yen and the New Taiwan Dollar. Large depreciations of these currencies would mostly likely cause the Korean and Southeast Asian currencies to depreciate (instead of appreciating with the end of financial panic). A collective devalutaion of this sort would decimate China’s exports.

The third calculation is that China’s capital controls are likely to be more effective than before. There were considerable unregistered capital flow in 1997 and 1998 as indicated by the "errors and omissions" line in the balance of payments. The balance of payments surplus was US$67 billion in 1997 but the stock of international reserves (minus gold) increased only US$36 billion that year. The trade surplus was US$45 billion in 1998 but there was less than a US$2 billion increase in international reserves.

The businesses owned by the military have long been rumored to have been major players in capital flight and smuggling. The government’s order to the military to hand over the business it owned represented a decisive tightening of capital controls. The handover was reportedly completed by the end of 1998.

China’s tight control over capital flow means that changes in the overall balance of payments position reflect primarily change in the trade balance, and that the direction that the exchange rate would move is determined by whether the balance of payments is in surplus or in deficit. When the "errors and omissions" category is large, the best indicator of the true balance of payments position is the change in the stock of foraign reserves. Since China's foreign reserves has been increasing since 1993, one could argue (as the U.S. trade negotiators have) that a freely floating RMB would have appreciated against the US dollar, instead of depreciating in 1994 and then staying unchanged since then. China would almost certainly be accused of exchange rate manipulation if it were to devalue when its balance of payments was still in surplus.

The fourth calculation behind the no-devalutaion pledge is that a RMB devaluation in 1999 when the financial markets have just calmed down may reignite the financial panic and cause the Korean and Southeast Asian currencies to depreciate anew. There is no assurance that the final outcome from this new round of depreciations would leave the RMB sufficiently depreciated against the currencies of the crisis countries.

Of the above four calculations, the one – in our opinion – that has the highest probability of being falsified is that there would not be substantial displacement of China’s exports in the markets of the developed countries. China’s export growth would also be reduced if a lot of originally China-bound FDI would now go to Southeast Asia instead. We would like to suggest, however, that even if the first calculation is wrong, China may still not choose to devalue. China could very well give in to the reflexes from its centrally planning past and use the import application processing system (which is designed to check capital outflow) to reduce imports. The efficiency costs of this new protectionism has to be not only compared with the expected costs of a possible renewed financial panic sparked off by a RMB devaluation, and also has to be weighed against the political benefits of being a responsible economic power that has the stability of international financial markets at heart.

Official Chinese thinking appears to be that a RMB devaluation cum resumption of current account convertibility in late 1999 or early 2000, after financial panic has totally leached out of the credit markets, are the best course of action for China. China’s pledge not to devalue in 1999 appears to have been most influenced by its fourth calculation - that a renminbi devaluation in 1999 would most probably spark off renewed region-wide devaluation. For good reasons, China does want to jeopardise the recovery of Pacific Asia, especially when the net result of the renewed regional devaluation may not result in a significant gain in competitiveness for China.

While we understand the over-arching importance of the fourth calculation in China’s decision on exchange rate policy, we have considerable skepticism about the validity of this fourth calculation. Our reading of past experiences suggest that financial panics resemble tropical storms. The immediate aftermath of tropical storms is usually marked by bright sunshine rather than by lingering showers. Similarly, the aftermath of financial panics is marked more by near-perfect calm rather than by investors jittery over their own shadows. It almost seems like investors generally heightened their rationality after the crisis to compensate for the irrational collective pessimism of the immediate past.

A postponed devaluation could wreak great havoc. FDI into China would certainly be postponed to after the devaluation, exporters would increase inventories than export their total production, importers would bring forward their future imports, and the increase in import protectionism in order to control the incipient trade deficit would invite the ire of US trade negotiators. There is thus a lot of merit to the view that if the first calculation were to fail, a devaluation of RMB should follow in short order. While there is even greater economic merit to an anticipatory devaluation (devaluing once it is clear that the trade account would deteriorate significantly), it cannot be done without risking a storm of criticisms from US trade negotiators.

From our reading of the tea leaves, we hazard the following propositions:

• GDP growth in 1999 will probably be about 7.0 percent to 8.0 percent;
• China is unlikely to experience a bank run, and certainly will not experience a meltdown of its banking system, and if a bank run were to occur China has the technical means to contain it;
• a RMB devaluation is likely to occur in response to the large devaluations by the crisis countries, but China might be able to postpone it (using administrative means if necessary) until 2000; and
• as long as the government is relying on massive infrastructure investment to reflate the economy, significant recapitalisation of the state banks is unlikely to occur because the government will not have the financial resources for the additional debt-servicing.

We want to highlight one possible negative long-run result from the present reflation package. There have been claims that the larger growth in the third and fourth quarters was achieved only after implicit assurances were given to bank managers that they would not be held responsible if the NPL ratio were to increase. A temporary deviation from the firm policy of cleaning up the balance sheets of the state banks may be defensible in the midst of the Asian financial crisis, but a prolonged deviation would underline the credibility of the commitment to reform the state banks and mean a return to the traditional socialist boom-bust cycle.

The long-term answer to the NPL problem goes beyond punishing bank managers who experienced increases in the NPL ratio, the long-term answer lies in changing both the supply-side and the demand-side of the credit market. Many changes are required on both sides of the credit market, and the most fundamental changes include the transformation of the state banks and the SOEs into shareholding corporations to make profit-maximisation their primary objective, the establishment of a modern legal framework to promote transparency and reduce transaction costs, and the creation of a prudential regulatory body to reduce excessive risk-taking by banks.

The above complex institutional changes that are necessary in order to address the NPL problem adequately illustrate that most of China’s economic problems cannot be individually addressed, success depends on systemic reform. This brings us to the basic point that while President Jiang and Premier Zhu deserve much credit for their competent handling of the current macroeconomic problems so far, their position in Chinese history will depend more on their success in addressing the many and varied long-term development challenges facing China. These challenges include the slowdown in agricultural productivity growth, the decline in job creation in the rural enterprise sector, the acceleration of losses by state-owned enterprises (SOEs), the relentless growth in nonperforming loans (NPLs) at the state banks, the inability of the legal system to meet the demands of an increasing sophisticated economy, and the inadequacy of social safety nets to cope with the temporary dislocations that are characteristic of a fast-growing economy. The ability of China to maintain its international competitiveness after the Asian financial crisis is over is conditional upon the resolution of the above problems.

We must say that, at the minimum, the audacity of the economic policies adopted so far by President Jiang and Premier Zhu – which included laying off one-third of the central bureaucracy and committing to making shareholding the principal form of ownership -- makes us (at least initially) optimistic and hopeful of China successfully overcoming the many obstacles on the way to becoming a competitive industrialised modern market economy.