Chinese Surplus Reversal: A Bitter Pill

China’s current account surplus reflects one of the most troubling imbalances in the global economy. This surplus, which has grown rapidly since the start of the decade, signifies the nation’s status as both a net supplier and net lender on the international stage, and has become a point of major international contention. In fact, China’s export-driven growth model makes it the world’s most active supplier and lender by a wide margin. In the third quarter of 2010, the surplus inflated to a staggering $102.3 billion, or 7.2% of Chinese GDP.[1] This figure represented an increase over the $72.9 billion surplus from the second quarter of 2010 and an even greater increase over the much more reasonable surpluses from 2009. Recent historical values are presented in Figure 1. Many, most notably the United States, have clamored for a reevaluation of the renminbi and reduction of the surplus, calling it an unfair economic aid, but China will find itself hard-pressed to make any significant  cuts to the surplus without undermining domestic political stability.

Economists disagree about the source of this large and growing surplus. The most popular explanation relies on exchange rate distortion. Analysts agree that the Chinese government, by trading massive amounts of renminbi for dollars at a fixed ratio, actively pegs its currency at the artificially low rate of 6.8 renminbi per dollar. This low rate makes Chinese exports cheap and attractive to foreign consumers while making imports relatively expensive and unattractive to domestic consumers, thereby lobbing the trade pendulum in their favor. While American politicians insist that Chinese currency manipulation is the source of global trade imbalances, other factors may be equally important. China’s extraordinary savings rate, for example, serves to explain its major capital outflows (net lending), which in turn can explain the export gap. Government policies promoting growth in China’s export sector are also to blame. Finally, economists point to relocation of industries from other East Asian economies to China and to various factor market distortions.[2]

Whatever the cause, the massive surplus represents a dangerous instability in the Chinese economy. The nation’s status as a net global lender has led to a massive influx of debt for its clients—particularly the United States. Indeed, most estimate that currency manipulation alone has led to a Chinese stockpile of over $1 trillion in US Treasury securities (see Figure 5).[3] Such massive holdings of foreign exchange reserves are fundamentally unstable, making Chinese asset values susceptible to fluctuations of a currency the state does not control. Perhaps more importantly, China’s reliance on exports leaves it vulnerable to fluctuations in global economic conditions. The recent financial crisis, for example, led to a sharp contraction of global demand that served as a “wake-up call” for China.[4] Economists also warn that massive expansion of the money supply caused by the surplus could impose inflationary pressures on the renminbi. Finally, the surplus has led to troubling political tensions with trading partners such as the US, who blame China for rising trade imbalances.[5]

Government intervention is clearly necessary in order to cure Chinese current account woes, and the state has a number of tools at its deposal. The most obvious is currency revaluation—if China allows the renminbi to appreciate to its natural level, exports will decrease while imports increase, allowing a shift towards stability. Another important tool is economic restructuring: finding a way to shift the economic balance away from exports and towards consumption. This will likely center around suppression of the private savings rate, which will stimulate consumption (and thus imports) while controlling net capital outflows (and thus exports). Finally, some economists suggest simple fiscal stimulus in the hopes that increased government spending can be another driver of income and consumption.[6]  However, this paper will not explore these options in depth, rather it will take as its fundamental assumption that, by the year 2010, China is able to achieve current account neutrality through an even balance of these reforms. Instead of evaluating the likelihood or effectiveness of such reforms, this paper will take them as given and explore the consequences for China.

Such a reversal may not come cheap.  As net exports fall, so might Chinese GDP. This blow to growth will be particularly salient given the Chinese political and economic landscape. China has been enjoying extraordinary growth over the last 30 years, with annual real GDP growth rates above 8% each year since 2000 (Figure 6). Critics, however, point out that such fabulous growth is actually necessary to quell political unrest in the communist nation. China’s “social stability imperative” suggests a threshold of 8% annual growth is necessary to prevent massive political uprising.[7] This imperative of consistently high growth rates has handcuffed Chinese leadership as it considers reforms aimed at curbing major economic instabilities. Thus, financial stability may come only at the cost of social stability, and if the   foregone growth of current account reversal proves too severe, the government may be unwilling to enact reform for fear of compromising social stability.

This paper takes aim at a crucial question: will it be possible for China to achieve a current account reversal through balanced reform while maintaining the growth demanded by its social stability imperative? The paper will begin by exploring the theoretical evidence. Specifically, it will examine the Mundell-Fleming model of open-economy macroeconomics to reach conclusions about the effect of surplus reversal on growth rates. Next, the paper will examine the empirical literature on the subject, evaluating whether the theoretical predictions have proven historically accurate. Finally, it will take all the evidence together, examining what variables affect the relationship between reversal and growth and how those variables apply to the current situation in China. The paper will conclude by using all the available evidence to arrive at the central conclusion: that it will not be possible for China to achieve a surplus reversal through balanced reform while maintaining growth above the 8% social stability threshold.

Theory

The Mundell-Fleming model is one of the most popular for describing the relationships between key macroeconomic variables in economies that engage in international trade. The model presents two particular advantages in analyzing the Chinese situation. First, the model is Keynesian, meaning it describes short-term relationships between variables. While long-run implications for growth are important, this paper is concerned exclusively with short-run performance. In fact, the Chinese government’s social stability imperative necessitates high growth rates in the extreme short-run—a single year of lagging income could prove catastrophic. Second, the model can be used to describe large open economies. Other models, in contrast, often focus on macroeconomic relationships in small open economies. As home of the world’s second-largest GDP, China has certainly achieved large economy status.

The Mundell-Fleming model revolves around the IS curve (a basic macroeconomic relationship which represents equilibrium in the goods market), the LM curve, which represents equilibrium in the money market, and the balance of payments identity. The IS curve dissects annual GDP into consumption, investment, government spending and net exports; the LM curve focuses on the supply of money and price level (exogenous factors) as well as interest and income; and the balance of payments which relies on the relationship between net exports and net capital outflows.

When its components are assembled, the Mundell-Fleming model holds that the interest rate and GDP of an economy will settle at the unique values that put both the goods and money markets in simultaneous equilibrium, and this equilibrium interest rate, in turn, determines the level of net capital outflow. Finally, this equilibrium level of net capital outflow gives the level of net exports, as they are equivalent according to the third equation. Because net exports is a function of the exchange rate, the equilibrium exchange rate will be that which equilibrates net capital outflows and net exports.

This model and the current situation in China differ in that China fixes its exchange rate at a target below the equilibrium level. In practice, it does this by selling renminbi for dollars at a fixed low rate. By selling renminbi, it expands the real money supply, which leads to a reduction in the interest rate, an increase in net capital outflow, an increase in net exports and a decrease in the exchange rate. Thus, while there is consistent “upward pressure” on the exchange rate as it attempts to float back towards its equilibrium level, government sales of renminbi actually artificially suppress it.[8]

The model can be used to predict the impact of a current account reversal, or a large decrease in the balance of net exports. One way the Chinese government might attempt to decrease net exports is through partial or full revaluation of the renminbi. To achieve this, the government would raise the peg at which it trades renminbi for dollars (or perhaps abandon the peg altogether), contracting the real money supply and thereby triggering a higher interest rate, which would in turn lead to lower net capital outflows. The lower net capital outflows would mean both a decrease in net exports and an increase in the exchange rate, reducing output.[9]

A second method for decreasing net exports hinges on expansion of domestic spending. This can be achieved either through structural policies that increase consumption (such as increasing wages or lowering the private savings rate) or through direct fiscal stimulus. Such policies would all have the effect of raising interest rates, and, as before, this would reduce net exports and increase the exchange rate. However, this policy would have an expansionary effect, increasing national income.[10]

One factor complicating this fiscal approach is that it would increase the exchange rate, which could cause discomfort in the Chinese government. If the government remained committed to an artificially low exchange rate, it would have to fight the increase by selling more renminbi. This would also have the effects of a decrease in the exchange rate, an increase in national income and an increase in net exports. This policy, however, would be counter-productive, as it would reverse the effect of the original fiscal expansion on net exports. Because this paper assumes a balanced reversal that includes partial or full revaluation of the renminbi, it is safe to assume that the government would simply let the exchange rate appreciate following fiscal expansion.

Unfortunately, the Mundell-Fleming paints a picture that looks hazy at best. A reduction in net exports achieved through revaluation of the renminbi would result in a contraction of national income, but a reduction in net exports achieved through either restructuring policies or direct fiscal stimulus would result in an expansion of national income. Thus, the effect on growth of a balanced approach that combines the two is ambiguous. In fact, it is possible that various policies could be combined to achieve massive reductions in net exports with zero net effect on national income. In the end, the combined effect on growth (as well as net exports) depends on the relative sizes of the different policies and on the elasticity of the Mundell-Fleming components. As these factors are extremely difficult to determine and even harder to predict, the net effect on growth must be deduced by examining historical current account reversals. With this in mind, this paper now turns to empirics.

Empirics

The bulk of empirical literature on current account reversals focuses on the impact of deficit, rather than surplus, reduction (situations directly opposite to China’s). Most of this work on deficit reduction arrives as similar conclusions. First, most authors find negative impacts on growth. Second, most find that deficit reductions were linked to currency crises (wherein overvalued currencies “crashed”). Third, most find that currency devaluations and decreases in net exports resulted in massive increases in net capital outflows. These conclusions are all plausible under the Mundell-Fleming model.

Melecky (2005) found fairly typical results. The author studied deficit reversal events in central and Eastern Europe from 1993-20008 and found that, on average, a current account reversal brought a 1.1% growth rate contraction in the following year. He also found that, in most countries, growth eventually recovered as domestic capital was substituted for reduced capital inflows. Finally, his results suggest that the rate of growth recovery depended on the size of the initial shock. Edwards (2001) reported nearly identical results in a study of 120 countries over 25. He found a negative impact on growth around 1% in the year following a reversal, most of which resulted from diminished capital inflows.[11]

Milesi-Ferretta and Razin (2000) examined the effects of various macroeconomic variables on the size of growth reduction following deficit reversals. The authors examined 105 middle- and low-income countries, scanning for “reversal episodes.” Interestingly, in the 100 episodes identified, the authors found that the median change in output growth was zero. However, there were a number of events that did result in significant contraction or expansion. In these events, growth was positively related to openness to trade and less-appreciated exchange rates (deficit economies tend to feature overvalued exchange rates). Growth was negatively related to debt and official transfers. The authors also found that growth had no relation to current account level prior to the reversal, the foreign interest rate, prior GDP or the level of investment.[12]

While the impacts of deficit reduction are well-documented, there is almost no empirical evidence available on the impact of surplus reduction on growth. While it is tempting to simply “reverse” the findings on deficit reduction and apply them to surpluses, the anatomy of the two types of reversal are different (deficit reductions are often traumatic, unintended episodes that follow currency crashes) and thus require different treatment. This paper now examines the bulk of the cannon on surplus reversals.

Crichton (2004) presented a joint analysis of both deficit and surplus reversals. The author began by examining a theoretical model, developed by Chari, Kehoe & McGrattan (2006), which describes a small open economy. The model predicted that surplus reversals have a negative impact on growth. He then turned to empirics, examining the effects of surplus reversal on developing economies. The author found that real GDP growth does tend to fall following reversals. Specifically, growth in sampled economies averaged 4.76% in the year before a reversal but fell to 3.67% in the year of the reversal event. Growth then tended to slowly recover, averaging 3.55%, 3.69% and 4.04% in the following three years. The author also examined the effect of reversals on net capital inflows (the opposite of outflows). The results suggest that inflows tend to increase (equivalent to decreasing outflows), as predicted by the Mundell-Fleming model. Specifically, the year before reversals saw average inflows of 2.8% of GDP, but the year of the reversal saw average outflows of 1.4% of GDP, a 4.2% swing. Outflows then averaged 1.3%, 2.2% and .8% of GDP, respectively, in the three years following the reversal.[13]

Meissner (2010) applied an anecdotal approach, presenting a case study of surplus reversals in France and Great Britain in the interwar period. While the effectiveness of anecdotal analysis is limited (particularly in reference to events so far in the past), the French case bears striking resemblance to the current situation in China. In 1926, the French currency had been stabilized through a peg against the British sterling pound, but most thought it considerably undervalued. To prevent appreciation, the French central bank sold large amounts of domestic currency in exchange for sterling at a fixed rate (France at the time, like China today, denied this). From 1926-1931, France experienced considerable current account surpluses. During these years, the economy enjoyed rapid growth (at 6% or more annually), much like China today. In 1931, however, things began to change. The US and Great Britain both devalued their currency, leading to exchange rate appreciation in France. At the same time, depression abroad led to decreases in exports (much as the current financial crisis has led to a slowdown of Chinese exports), and eventually, France enacted expansionary fiscal policy to help fight depression. As a result of these effects, the current account gradually reversed. At first, however, the expansionary fiscal policy offset the losses imposed by the exchange rate appreciation, and the growth rate remained stable. It wasn’t until 1936, when export markets began to dry up completely, that growth started to fall.[14]

The most recent study on surplus reversals, Abiad et al. (2010), is by far the most exhaustive. The authors examined data from 1960 to 2010 in search of policy-driven surplus reversals. These policies included exchange rate appreciation, fiscal stimulus and structural rebalancing. Maintaining specific technical criteria for a policy-driven reversal, the authors identified 28 examples in the past 50 years and then examined the “anatomy” of current account reversals. Their data showed policy-driven reversal to be remarkably effective, with average surpluses dropping from 5.5% to just .4% of GDP. Increases in imports usually drove these adjustments, with the average economy experiencing import growth of 4.2% of GDP and export reduction of only .1% of GDP. On average, reversals led to a drop in savings of 2.1% of GDP and an increase in investment of 3.0% of GDP. Finally, reversals often accompanied an appreciation of previously-undervalued exchange rates, as the average nominal exchange rate appreciated 9.2% in the course of the reversal.[15]

They then examined the effect of surplus reversal on growth. Their findings suggest that, on average, reversal does not bring lower growth (Figure 16, Top Panel). In the short-term and middle-term, changes in the growth rate were not statistically different from zero. To control for fluctuations in the global economy, the authors also measured growth relative to global growth and obtained the same result (Figure 16, Bottom Panel). Their results indicated that growth following stimulus-driven reversals slightly outpaced growth following exchange rate appreciation-driven reversals, again in line with the predictions of the Mundell-Fleming model. Typically, while reversals reduced net exports, these were offset by increases in domestic consumption (Figure 17). Although the effect of surplus reversals on growth was found to be negligible on average, the authors found substantial variation in outcomes (Figure 18). Changes in the growth rate ranged from -5.1% to 9.4% and were somewhat skewed to the right.[16]

Finally, Abiad et al. attempted to account for the diversity in growth outcomes by examining the effects of various macroeconomic variables on post-reversal growth. The authors concluded that countries enjoying higher growth rates before a reversal tend to experience lower growth afterwards (specifically, a 1% increase in growth before a reversal results, on average, in a .65% decline after reversal). They also found that reversals occurring during periods of slow global growth lead to greater declines in post-reversal performance: 1% decrease in global growth accompanies a roughly .5% decrease in domestic growth. Countries more exposed to global demand (as measured by the export/import ratio, or “terms of trade”) have particular susceptibility to this effect. Next, the authors found that countries with larger pre-reversal surpluses experience greater declines in growth. This result was driven in particular by low growth among countries with high savings rates. The authors also showed that exchange rate appreciation (by economies with previously undervalued currencies) exacerbates slowdown of demand: a 10% appreciation of the real exchange rate brings, on average, a 1% contraction of growth, with effects increasing over time. This has a more pronounced effect in more undeveloped economies, as measured by per capita income and export quality. All told, variation in these inputs accounted for 70% of the observed variation in growth rate changes following reversals.[17]

Conclusions

The results of the empirical analysis match the predictions of the theoretical analysis. In general, the evidence analyzed in this paper seems to agree that surplus reversals have ambiguous effects on growth is ambiguous, but both increases and decreases are possible and do occur. These diverse outcomes depend on a number of key macroeconomic variables. Reversals tend to feature both exchange rate appreciation and net capital outflow decreases. Balanced reversals generally see growth decreases associated with exchange rate revaluation but compensating growth increases associated with fiscal expansion.

However, the observed macroeconomic trends in growth outcomes do not bode well for Chinese surplus reversal. Specifically, Abiad et al. (2010) report that higher pre-reversal growth rates lead to lower post-reversal growth, and Chinese growth rates have been at unprecedented levels for years (Figure 6). They also found that a slow global economy can lead to lower post-reversal growth, particularly for economies heavily exposed to external demand. The recent financial crisis has left the global economy reeling, and global demand—to which China is heavily exposed—appears positioned to lag in the coming decade. The authors also point out that larger pre-reversal surpluses lead to slower growth, and the current Chinese surplus (7.2% of GDP) is well above the study’s average (5.5%). The effect was particularly pronounced for economies with high savings rates, and China’s savings rate has reached historically unprecedented levels (Figure 3). Finally, the study found that economies requiring greater exchange rate appreciation saw bigger decreases in growth, with especially severe effects on developing economies. China, of course, has both a vastly underappreciated currency and a developing economy. Thus, every single macroeconomic indicator seems to point to a considerable reduction in China’s growth rate following a potential reversal. Given all of the variables in consideration, forecasting the magnitude of such a reduction is quite difficult. However, because China comes up short in every single indicator, a growth contraction of at least 2% seems likely. This value is only twice the average found by Crichton (2004) and less than half of the maximum reported by Abiad et al. (2010), who found contractions between 0% and 3% in 50% of all surplus reversals.

Such a growth contraction may be an impossible pill for the Chinese leadership to swallow. Growth rates have hovered around 9% for the past 2 years. Since 2000, they have mostly fluctuated between 8% and 10%, with a brief period of higher growth between 2006 and 2008 (Figure 6). Li et al. (2010) present a simulation of the Chinese economy and predict growth rates through 2030 which holds that, even with optimistic assumptions about development of the Chinese economy, growth rates will not venture far beyond 8% in the coming decade as China becomes increasingly developed.[18] Given these growth trends, it seems unlikely that the Chinese economy would be able to absorb a 2% blow and maintain consistent growth above the 8% target. Growth rates below this critical threshold that last even a single year could spell disaster for Chinese social stability. Thus, it looks unlikely that China would be able to achieve a current account reversal while maintaining the growth demanded by its social stability imperative.