American Economic Association

China's Exchange Rate Policy Dilemma Author(s): Morris Goldstein and Nicholas Lardy Source: The American Economic Review, Vol. 96, No. 2 (May, 2006), pp. 422-426 Published by: American Economic Association Stable URL: http://www.jstor.org/stable/30034684 Accessed: 12/12/2009 22:53
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=aea. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The American Economic Review.

http://www.jstor.org

ASIANCURRENCY MATTERSt

China'sExchangeRate Policy Dilemma

AND NICHOLASLARDY* By MORRISGOLDSTEIN

This papersummarizes aspectsof China's key exchangeratepolicy, outlinesthe problemsit creates for both Chinaand the global economy, and proposesa feasible policy compromise. I. China's CurrencyRegime On July 21, 2005, China announced a 2.1percent appreciation of the renminbi (RMB) against the U.S. dollar, a move to a managed Most of float, and a numberof other "reforms." these reforms simply reiterated long-standing since 1994, China has identified arrangements; its currencyregime as a managed float and has set a 0.3-percent, per day, fluctuationlimit (in either direction) for the RMB against the dollar (vis-a-vis the central parity). The July 2005 announcement,however, did pledge two potentiallyimportantalterations:(a) the RMB was, henceforth,to be managed"with reference to a basket of currencies"ratherthan being pegged to the dollar;and (b) the exchange rate was to become "more flexible," with its value based more on "market supply and demand." In practice, the July 2005 reforms have had little visible effect. As of mid-December2005, the RMB-dollar rate was 8.07, a furtherappreciation of only 0.5 percent. There is also little evidence of pegging to a basket; rather, the RMB continues to trackthe U.S. dollar closely. And the centralbank's monthly interventionin the foreign exchange marketin August and September remained huge, at $18 billion per month-only slightly smaller than the $19 bilt Discussants: Eswar Prasad, International Monetary Fund; Peter Kenen, PrincetonUniversity. * Goldstein: Institutefor International Economics, 1750 MassachusettsAvenue, NW, Washington,DC 20036-1903 (e-mail: mgoldstein@iie.com; Lardy: Institute for International Economics, 1750 MassachusettsAvenue, NW, Washington, DC 20036-1903 (e-mail: nlardy@iie.com). 422

lion per monthin the firsthalf of 2005. In short, China's exchange rate system remainsa heavily managed peg to the dollar-and at a dollar exchange rate very close to the level prevailing before the July reform. II. Is the RMB Undervalued? There are several approaches to evaluating the misalignmentof the RMB. The "underlying balance" approachasks what level of the real, effective (trade-weighted)exchange rate would produce an equilibriumin which the "underlying" current-account position is approximately equal, and opposite in sign, to "normal"net capital flows. During the 1999-2002 period, before there was any expectationof a currencyappreciation, China ran an average capital-account surplus equal to 1.5 percent of GDP. Take this as our measure of normal net capital flows. currentaccount Next, define the "underlying" as the actual current-account position, adjusted for both cyclical movements that make the demand for importsunusuallyhigh or low, and the effects of recent exchange lagged trade-balance rate changes. China's 2005 current-accountsurplus will probablybe in the neighborhoodof 7 to 9 percent of GDP. The underlying current-account surplus should be somewhat lower-say 5 to 7 percent of GDP-because (buoyant GDP growth rate notwithstanding)domestic demand

1 Goldstein (2004) provides fuller discussion of this approach. There could be cases where some increase in reserves was desirable, or where only some share of the externalimbalancewas to be eliminatedwithin a given time period. Nevertheless, the basic premise is that a large discrepancy between normal capital flows and the underlying current-account position signifies a misalignmentof the real equilibriumexchange rate.

VOL.96 NO. 2

ASIAN CURRENCY MATTERS

423

growth has slowed and in 2005 the RMB has appreciatedin real trade-weightedterms. If China's underlyingcurrent-account surplus is now in the range of 5 to 7 percent of GDP, while what is requiredto offset normal capital flows is an underlying current-accountdeficit equal to 1.5 percent of GDP, then China's current accountneeds to deteriorateby a whopping 6.5 to 8.5 percent of GDP to restore (strict) balance-of-paymentsequilibrium.If one does a set of simulations with a small trade model to calculate what size real, effective appreciation of the RMB would be necessary to producesuch a large turnaround the currentaccount-takin ing into account the high import content of China's exports, using a plausible range for price elasticities of demand and supply, and making alternativeassumptions about the second-round effects of income changes on the demandfor imports-the answerscongregatein the upper part of the 20 to 40 percent range. This is somewhat larger than the estimated undervaluationfor 2003-2004.2 We would not want to go to the stake for the precision of our estimates of RMB undervaluation: after all, there is uncertainty about the underlying parameters, other methodologies yield somewhat different estimates, and China is in the midst of a nontrivial revision of its GDP accounts. Still, with huge reserve accumulation in each of the past three years; with persistent surpluses on both the current and capital accounts; with the real, trade-weighted RMB showing a cumulative depreciationsince the dollar peak in February2002; with the Chinese economy runningat full steam, or close to it, for most of the past three years; and with Asian currency appreciationneeded to reduce defthe excessively large U.S. current-account icit, we submit that it is difficult to arrive at a credibleestimatethat shows anythingbut a sizefor able undervaluation the RMB.3

III. Problemswith the CurrentRegime China's currentexchange rateregime has created problems for China and the global economy. First, the (de facto) fixed-dollarexchange rate limits the independence of China's monetary policy and has contributed,thereby,to the pronounced macroeconomic fluctuationsof recent years. In 2003-2004, when investment was booming and the rate of price inflation was accelerating, the central bank was reluctantto raise domestic lending rates, in part because it feared that, despite controls,higher rates would attractcapital inflows. As a result, the real rate of interest to corporateborrowers (proxied by the one-year bank-lending rate minus the change in the corporategoods price index), fell from 8 percent in 2002-Q3 on the eve of the investment boom to negative 4 percent during April through September 2004; that, in turn, contributedto an excess demand for loans. Second, China's undervalued currency has contributedto growing trade surpluses and, at least in some years, to very large portfolio capital inflows, which appearmotivated by an expectation of appreciation.Mammoth exchange market intervention,amounting to 11, 12, and 14 percent of GDP in 2003, 2004, and the first half of 2005, respectively, has been necessaryto prevent currencyappreciation. The central bank sterilized much of this reserve accumulationthroughincreasesin reserve and requirements open marketoperations.Since mid-2004, the central bank also used administrative controls to limit bank credit creation. This approach is costly in several respects. There have been episodes when the barn door was locked only after the horses had left. In 2003 and the firsthalf of 2004, therewas a bank credit blowout, with the ratio of the increase in bank credit to GDP hitting an all-time high. This led to an investmentboom, with long-term consequences for excess capacity in a number of sectors, for downwardpressureon operating margins in these sectors, and potentially for nonperformingloans in banks that lent heavily to supportsuch projects.In 2005 we again witnessed a substantialbuildup of bank liquidity, reflectedin growing excess reserves and declining money marketrates. Increased use of administrativecontrols on

2 Goldstein (2004) and Lardy (2005). To estimate the undervaluationof the RMB with respect to individual currencies, one needs a multi-countrymodel. William R. Cline (2005) does such an exercise and estimates that as of November 2005, the RMB was undervaluedwith respect to the U.S. dollar by 43 percent. 3 Taking a broad view of the evidence across several methodologies, it looks to us like the RMB is undervalued in real, trade-weightedterms by 20 to 35 percent.

424

AEA PAPERSAND PROCEEDINGS

MAY2006

lending is inconsistent with the government's long-term policy of requiring state-owned banks to operate on commercial principles. When the centralbank specifies lending ceilings and sectoral lending priorities,the development of a credit culture is slowed. And when money market interest rates are depressed by excess liquidity, bank profitabilitysuffers, since banks are net lenders in the money market. Third, a highly undervalued RMB encourages excess investmentin tradablegoods. Eventually, the real exchange ratewill appreciateand will lower profitabilityin tradablegoods industries, again with adverse implicationsfor banks. Fourth, China is accumulatinga large exposure to potential capital losses. China's foreign exchange reserves will, at year-end 2005, be in the neighborhood of 40 percent of GDP. A 20-percent revaluationof the RMB against the majorreserve currencieswould impose a capital loss equivalent to 8 percent of GDP. Finally, China's prolonged, large scale, oneway interventionin the foreign exchange market-along with its large and growing global current account surplus-fuels protectionist pressure in the United States and elsewhere. One reflection is the Schumer-Graham bill in the U.S. Senate, which could lead to a uniform 27.5 percent tariff on all Chinese imports. Another is the ill-considered congressional action subjecting a potential China National Offshore Oil Corporation(CNOOC) purchaseof Unocal to a special congressional review. Since over half of China's exports go to the United States, Euroland, and Japan, and since China has a long-terminterestin investing abroad,such protectionist threatsought not be taken lightly. In sum, the costs to China of maintainingan undervalued currency are already significant and are likely to rise over time. Moreover, unlike some others (e.g., Michael Dooley et al., 2004), we do not see the growth and employment benefits of an undervaluedRMB as exceeding its costs (Goldstein, 2006; Goldstein and Lardy, 2005). China's inflexible currency regime and its undervalued RMB also handicapthe adjustment of global imbalances and increase the risk of a hard landing for the dollar and the U.S. economy, with adverse global spillovers. The U.S. current-account deficit is now runat about 6.5 percent of GDP-an all-time ning

deficit is not high. If the U.S. current-account reduced, U.S. net foreign debt relative to GDP will rise to a worrisome level and U.S. assets will eventually lose their attractivenessrelative to alternativeinvestment opportunities-resulting in some combination of a fall in capital flows to the United States, higher interest rates on U.S. liabilities, and a decline in the dollar. If these adjustments were large and disorderly, they would depress U.S. growth significantly, with adverse spillovers for U.S. trading partners. As Michael Mussa (2005) has emphasized, the inevitable correction of the U.S. external imbalance will necessarily involve a depreciation of the dollar against most currencies, and domestic demand growing more slowly (faster) than domestic output in the United States (rest of the world). The first channel is the expenditure-switchingchannel, while the second is the expenditure-changingchannel. Both are needed. That is why both a more ambitiousprogramof fiscal consolidation and a tightening of U.S. monetary conditions are essential-along with policies that stimulate domestic demand growth in major U.S. trading partners. The real, trade-weightedU.S. dollar needs to fall by an additional 15 to 25 percent from its current level to support external adjustment. The currenciesof Japan,China, and the rest of emerging Asia have a combined weight of roughly 40 percent in the Federal Reserve's dollar index. If the Asian currencies do not participate,eitherthe overall depreciationof the dollar will be too small or the appreciation burden will fall too heavily on economies where a further, large appreciation would not be warranted. In the wave of dollar depreciation starting in February 2002, the euro, the Canadian dollar, and the Australian dollar, among other market-determined currencies, experienced strong, real appreciation, whereas the Asian currencies-with the notable exceptions of the Korean won and Indonesian rupiah-did not. China remains a prime candidatefor leading wider Asian currency appreciationbecause of its large external imbalance and because the RMB's behavior can influence -for better or worse-the attitudethat other Asian economies have toward pressuresfor appreciationin their

VOL.96 NO. 2

ASIAN CURRENCY MATTERS

425

own currencies.4In this sense, it is important that China be part of the internationalsolution to the adjustment of large external imbalances-and not part of the problem. IV. Policy Constraints and a Compromise Solution China's willingness and ability to deal with the shortcomings in the current exchange rate regime are heavily constrainedby at least three importantfactors. First, China's banking system, although improving, remains fragile. Households, which have bank deposits equivalent to almost 100 percent of GDP, have had little opportunityto diversify the currency composition of their financial assets and could be susceptible to capital flight if they thoughtrisks of repaymenthad worsened. Thus, any adjustmentof China's exchange rate regime will have to maintainmost existing capital controls until the domestic banks are furtherstrengthened. Second, the degree of undervaluationof the RMB is now so large that it would take a long time to correct if it were done in a series of small steps. But this is apt to exhaust the patience of the international community.Also, the the expectationthat each RMB apprecigreater ation will be followed by another,the larger is the scale of capital inflows and the larger the sterilizationproblem. And third, China's current leadership, even more than its predecessor,appearsto prefer an incremental approachto dealing with the current exchange rate dilemma. They highly value growthand stability.They worrythatlettingmarket forces freely determine the exchange rate couldresultin too muchRMB appreciation, slowing exports, employment, and growth and, to thereby,contribute increasedsocial instability. How then to square the circle? Absent the aforementionedconstraints,the policy solution would be straightforward: China could simultaneously and immediately remove the restrictions on capital flows and let the market determine the value of the RMB. The con-

straintsmean the search is for second-best policy options. All things considered, we believe the preferred course of action is a compromise along the following lines.5 China would make a credible down payment on removing the large undervaluation of the RMB by immediately revaluing it by 10 to 15 percent, relative to the currencybasket;it would widen the bandto 5 to 7 percent on each side of the new parity; it would simultaneously implement fiscal expansion; and it would leave in place most capital controls.The 10- to 15-percent revaluation the of RMB would help to reduce China's very large current-account those surplusandshouldpersuade in theU.S. Congress,andin otherG-7 legislatures, that China was finally getting seriousaboutcontrollingits growingexternalimbalance.A 10- to 15-percentrevaluationwould be a much more tangible sign of progressthan the meager 2.5thus far. percentrevaluation registered A significant widening of the band would give China some increased independence for monetary policy, would allow scope for some further appreciation of the RMB, and would give China some practical experience in managed increased flexibility. It would also give some substance to China's repeatedclaim that "market supply and demand" would increasingly determinethe RMB's value. Fiscal expansion is needed to offset some of the contractionary effects of the RMB revaluation. The fiscal expansion needs to be expenditure-based, because direct tax receipts in tax China-like the agricultural and the income tax-are currently far too small to provide a meaningful boost from a tax reduction. Increased government expenditures should strengthenthe social safety net because one of the main factors behind China's very high household savings rate is precautionary savings against potentiallyhigh out-of-pocketexpenses for health care, education, and retirement.6 Keeping in place the bulk of existing controls measureagainstpoon capitalflows is a prudent tential capital flight, should there be bad news aboutthe stateof the still-fragile bankingsystem.
5 This is close to our original formulationof the "twostage" approachto China's currencyreform (Goldstein and Lardy, 2003). 6 Olivier J. Blanchardand Francesco Giavazzi (2005).

In contrast, Barry Eichengreen (2004) has argued that other Asian currencies would not necessarily follow the RMB upwardif China revalued.
4

426

AEA PAPERSAND PROCEEDINGS

MAY2006

True, this would leave a sizeable, real appreciation of the RMB to be carriedout later, over, say, the next two years, with all the expectations-cum-capital flow problems that such a phased adjustmententails. Indeed, that is why we pushed for China to remove, in one step, the undervaluation 2003-2004.7 But thatwindow in of opportunity passed-now that the underhas valuationof the RMB is largerand domesticdemand growth has slowed. While speculative inflows have recently declined from the record pace of 2003-2004, one cannotbe confidentthat they would not revive;in such a case, the authorities would need to choose between an acceleration of the RMB appreciation and a temporary recourseto tightercontrolson capitalinflow. In the final stage of currencyreform-when China's banking system is on firmerfootingChina would float the RMB and remove the remaining controls on capital flows. Appropriate sequencing of capital-accountliberalization would preventcurrencyreformfrom undermining the continued strengtheningof the banking system. Admittedly, this is not such an elegant plan. But if it avoids the internationaltrain wreck that would otherwise take place, it merits consideration. REFERENCES Blanchard,Olivier J. and Giavazzi, Francesco. "Rebalancing Growth in China: A ThreeHanded Approach." Massachusetts Institute
Goldstein and Lardy (2004).

of Technology, MIT Departmentof Economics Working Paper:No. 05-32, 2005. Cline, William R. "The Case for a New Plaza Agreement." Institute for InternationalEconomics, IIE Policy Brief: No. PB05-4, 2005. David and Dooley,MichaelP.; Folkerts-Landau, Garber,Peter. "The Revived Bretton Woods System." International Journal of Finance and Economics, 2004, 9(4), pp. 307-13. Eichengreen, Barry."ChineseCurrencyControversies." Center for Economic Policy Research, CEPR Discussion Papers: No. 4375, 2004. Morris."AdjustingChina's Exchange Goldstein, Rate Policies." Institute for International Economics, IIE Working Paper: No. 04-1, 2004. Goldstein,Morris. "RMB Controversies."Cato Journal, 2006, 26(2). Goldstein, Morris and Lardy, Nicholas R. "China'sRole in the Revived BrettonWoods System: A Case of Mistaken Identity."Institute for International Economics, IIE Working Paper:No. 05-2, 2005. Goldstein,Morrisand Lardy,NicholasR. "TwoStage Currency Reform for China," Asian Wall Street Journal. September 12, 2003. Lardy, NicholasR. "ExchangeRate and Monetary Policy in China." Cato Journal, 2005, 25(1), pp. 41-47. Mussa, Michael. "Sustaining Global Growth While Reducing ExternalImbalances,"in C. Fred Bergsten,ed., The UnitedStates and the world economy: Foreign economicpolicy for the next decade. Washington, DC: Institute for International Economics, 2005.