For the past ten years the global balance of payments has been dominated by the trade and investment relationship between China and the US. This relationship is now undergoing a major shift. Most large economies will be affected, and to the extent that their economic policies do not accommodate this shift, they are likely to fail, in much the same way that economic policy failed in the 1930s.
China runs a massive current account surplus with the US and, in recycling this surplus, an equally large capital account deficit. This recycling has been both the main source of the global liquidity that has engulfed the world until recently and a constraining factor in the global economy. Given their magnitude it is impossible for either country to adjust without a major counterbalancing adjustment from the other, but it is far from clear that policy-makers on either side, especially in China, have a clear grasp of how big the necessary adjustments must be. The result is likely to be a steep drop in global growth, much of it borne by China, and possibly even a collapse in global trade.
Other countries have played a role in this imbalance too, of course, but with a few important exceptions (OPEC, for example) they have fallen broadly into two camps whose characteristics are typified either by China or the US. One set of countries, like the US, has had booming domestic consumption and high and rising trade deficits. Their highly sophisticated financial systems were able to intermediate the surge in underlying liquidity into the consumer loans that permitted the consumption binge. The second set of countries, like China, have excessively high rates of saving that have been systematically funneled into domestic investment, resulting in a huge and rising surplus of production over consumption, the balance of which is exported abroad.
Until recently excess US demand and excess Chinese supply were in a temporarily stable balance. As part of running a trade surplus, China necessarily accumulated dollars, which had to be exported to (invested in) the US. This capital export from China to the US did not occur in the form of private investment – indeed it was actually exacerbated by the fact that China was a net importer of private capital – but rather occurred as forced accumulation of central bank reserves, which were recycled back to the US in the form of purchases of US Treasuries and other US dollar assets. Since China had effectively pegged its currency, its central bank, the People’s Bank of China (PBoC) had no choice but to accumulate reserves in this manner as long as China ran a surplus.
The recycling process also functioned as a great liquidity generator for the world, converting US consumption into Chinese savings, which were then recycled back into the US financial markets through purchases by the PBoC of highly liquid US securities. There are several self-reinforcing aspects to this system that pushed it to the extremes it ultimately took. In the US the torrent of inward-bound liquidity boosted real estate and stock market prices. As markets surged, substantially increasing the wealth of US households, these same households began diverting a steadily increasing share of their income to consumption rather than savings. At the same time rising liquidity always forces financial institutions to adjust their balance sheets as they attempt to accommodate the expansion in underlying money, and one of the most common ways they do so is by increasing outstanding loans. With banks eager to lend, and households eager to monetize the value of their assets in order to fund consumption, it was only a question of time before household borrowing ballooned.
Meanwhile in China, the country’s currency regime locked it into self-reinforcing trade surpluses. As foreign currency poured into the country via its trade surplus, the money was purchased by the PBoC, whose accelerating accumulation of foreign reserves was paid for by creating money or by issuing central bank bills, a close substitute for money. In China most new money creation ends up in the banking system, whose main purpose is to fund investment (consumer lending is a negligible part of bank lending). As investment surged, industrial production inevitably grew much faster than private and public consumption. A country’s trade surplus is the gap between its production and its consumption, and as this gap grew, so did China’s trade surplus, which resulted in even more foreign currency pouring into the country, thus reinforcing the cycle.
In this highly instable balance, sometimes dubbed Bretton Woods II, Chinese overcapacity was matched with American over-consumption, and Chinese official lending was matched with US household borrowing. This ensured that the current account flows were matched with the capital account flows, and any change in one of these accounts required equal and opposite changes in the other three. This is a fundamental requirement of the global balance of payments – it must balance.
The great imbalance
Many analysts think of the US economy as the engine that drives the rest of the world, but this is not always true. Sometimes large changes or distortions in one part of the world can force adjustments elsewhere, and as the world’s largest and most open economy, with an astonishingly flexible financial system, it is often the US that absorbs adjustments originating elsewhere.
We saw this in the trade numbers. For most of last fifty years the US current account has fallen within a range of plus or minus 1% of GDP. There have been at least two exceptions. The first occurred in the mid-1980s when the US current account deficit rose to nearly 3.5% of GDP in 1986-87 before declining sharply and running into a small surplus in 1990. The second began in 1994, around the time of the Mexican crisis, when the US current account deficit climbed to around 1.6% of GDP, declined for two years, and then took off in 1997-98, after which time it raced forward in straight line to peak, in 2006, at 6.2% of GDP.
If the US trade deficit were driven simply by a US consumption binge, as is often claimed, it is a little hard to see why it would have followed a pattern of general stability over many decades marked by two surges – a small one from 1984-1988 and a very large one after 1997. If it was driven by changes in Asian savings, this pattern becomes easier to understand. The 1980s surge was driven largely by Japanese trade policies and domestic savings and is a fascinating case study in itself, but it is the post-1997 surge that is much more interesting and relevant to the current crisis.
1997 was, of course, the year in which several Asian countries experienced terrifying financial crises and viciously sharp economic contractions, and the crises profoundly impressed Asian policy-makers to this day. The crises seemed to be caused by the sudden reversal of current account surpluses into deficits, along with inverted balance sheets in which large external obligations were mismatched against domestic assets. Central bank reserves normally act as a hedge against this kind of currency mismatch, but the most affected Asian countries were precisely those countries with very low levels of foreign currency reserves. When the market worried about external debt sustainability, the currency mismatch proved to be the biggest concern, and as investors fled they set into play a series of events that caused plunging asset prices, collapsing currencies and, as a consequence of the latter, a surge in the external debt burden. The result was economic chaos from which most of the affected countries have still not emerged.
One of the main lessons policy-makers learned from the crisis was the risk of a currency mismatch between external obligations and domestic assets – too much dollar debt and not enough dollar reserves. To protect themselves from a repeat of the disastrous 1997 crisis many Asian policymakers engineered current account surpluses by managing trade policy and the value of their currencies. As a necessary consequence they began amassing large foreign currency reserves.
This resulted in what some have called a global capital flow “paradox” – the fact that in recent years developing countries have been large and growing net exporters of capital to rich countries. This is a paradox because, historically, capital-poor developing countries have generally been net importers of capital. By accumulating foreign currency reserves they are often net exporters of official capital, but for most of the last fifty years official capital exports on average were significantly less than net private capital imports.
In 1998 official capital exports in the form of foreign currency reserve accumulation among developing countries began to take off, and by 1999 net official capital exports exceeded net private capital imports. This is when the “paradox” begins. Since 1998 except for a small decline in 2001 net capital exports from developing countries surged almost in a straight line to around $700 billion annually (combining $1.2 trillion of reserve accumulation versus $0.5 trillion of net private flows).
But the global balance of payments must balance. As Asian trade surpluses and net capital exports surged, some other part of the world had to equilibrate these adjustments by running large trade deficits and importing capital. The US did exactly this, and the US trade deficit soared after 1997 while at the same time US household savings collapsed.
This process was enabled by two related factors. First, the massive recycling of the US trade deficit into the US securities markets set the stage for the surge in real estate and stock market prices which, by raising the market value of accumulated US savings, encouraged households to consume increasingly larger shares of their income. Second, as financial institutions accommodated themselves to the surging liquidity, their response – as it has always been during every liquidity surge – was to expand credit rapidly. This included intermediating flows to home-owners and consumers via new mortgages, credit card loans and consumer loans, and as banks made nearly unconditional lending offers to American consumers, inevitably debt-financed consumption rose. As part of the adjustment in the global balance of payments, US households took advantage of laughably easy lending conditions to engage, much too willingly, in the greatest, gaudiest spending spree in history.
Now, however, the party is over. The global balance of payments of the last ten years has finally broken down, and the resulting financial crisis and economic contraction has occurred as the world lurches to find a new stable balance. One necessary consequence of the financial crisis will be an increase in US household savings rates. Collapsing real estate and stock markets have caused household wealth to decline sharply, and households must save more than ever out of current income to replenish their wealth. But even if consumers wanted to continue spending, banks – caught in one of the worst credit crunches in recent history – are no longer willing to lend for consumption. The US household savings rate has nowhere to go but up.
By how much will US household savings increase? For most of the past fifty years until the early 1990s US household savings rates have varied between 6% and 10% of GDP (except for a few brief periods during the economic crisis of the 1970s when household savings went as high as 13% of GDP). In the early 1990s, the savings rate began declining, virtually collapsing after 1997 when household savings fell to well under 2% of GDP, just as central banks in developing countries, not coincidentally, began engineering large trade surpluses and accumulating vast amounts of foreign reserves.
Although we can’t say for sure, it is probably safe to argue that US savings rates will climb back to earlier average levels, or even temporarily exceed those levels, as American households rebuild their shattered balance sheets. If they return only to the mid-point of earlier savings rates, this implies that US household savings must rise by some amount equal to roughly 5% of US GDP, or, to put it another way, that all other things being equal US household consumption must decline by that amount.
But since the balance of payments must balance, something else must happen to equilibrate this decline in US household consumption. Either consumption in other sectors of the US economy (i.e. the government) must expand by that amount, or consumption by China (by which we mean all foreign countries, with China bearing the brunt) must expand by that amount, and as it does so its savings must decline. To the extent that neither happens, global overproduction – which consists mainly of Chinese overproduction – must decline by that amount. This is just a way of saying that if net American consumption (the excess of consumption over production) declines, either consumption must rise somewhere else, or production must fall.
In the best possible world Chinese consumption would rise by exactly the same amount as US consumption drops, and a new stable balance would quickly be achieved with one major difference: the US trade deficit would decline, and the amount of capital exported by China to the US would decline by exactly the same amount (the PBoC would accumulate fewer reserves). But if it doesn’t, total global consumption must decline, and the world economy slow – in fact as it slows global income will decline with it, so that both savings and consumption could decline, trapping the world in a downward spiral.
By how much must Chinese consumption rise to prevent a global slowdown? Given that the US economy is about 3.3 times the size of China’s, and consumption accounts for less than 50% of China’s income, Chinese consumption will have to rise by nearly 40% in order to accommodate a 5% increase in US savings. This is clearly unlikely.
It’s 1929 again
Although there are great differences between 1930 and 2008 that should not be papered over, the global payments imbalances that led up to the current crisis were nonetheless similar in many ways to the imbalances of the 1920s – with a few countries, dominated by one very large one, running massive current account surpluses and accumulating, in the process, rapidly growing central bank reserves. In the 1920s it was the US that played the role that China is playing today. The U.S. economy had been plagued in the 1920s with overcapacity caused by substantial increases in US labor productivity. This was a consequence of significant investment in the agricultural and industrial sectors and the mass migration from the countryside to the city.
Although US capacity surged, domestic demand did not rise nearly as quickly. As a consequence the US ran large trade surpluses, which stayed high as long as domestic production grew more rapidly than domestic consumption. US overcapacity didn’t matter when there was sufficient demand from abroad. It could be exported, mostly to Europe, while foreign bond issues floated by foreign countries in New York permitted deficit countries to finance their net purchases. But as the US continued investing in and increasing capacity, without increasing domestic demand quickly enough, it was inevitable that something eventually had to adjust.
The financial crisis of 1929-31 was part of that adjustment process. It was not just the stock market that fell – bond markets collapsed too, and bonds issued by foreign borrowers were among those that fell the most. This, of course, made it impossible for all but the most credit-worthy foreign borrowers to continue raising money, and by effectively cutting off funding for the current-account deficit countries, it eliminated their ability to absorb excess US capacity.
The drop in foreign demand required a countervailing US adjustment. Either the US had to increase domestic consumption, or it had to cut back domestic production, but there was unfortunately more to the crisis than simply the drop in foreign demand. With the collapse of parts of the domestic US banking system, domestic private consumption looked set to fall, rather than rise.
The slack in demand should have been taken up by US fiscal expansion, but instead of expanding aggressively, as Keynes demanded, President Roosevelt expanded cautiously. When the credit crunch came and the world was awash in American-made goods that no consumer was willing or able to buy, it was unreasonable, as Keynes argued bitterly, to expect the rest of the world to continue demanding US goods, especially since the financing of their consumption had been interrupted.
Since US production significantly exceeded US consumption (the balance consisting of the trade surplus), the need for demand creation most logically rested in the US. In 1927 and 1928 there had already been unemployment pressures and the 1929 collapse in foreign and domestic demand exacerbated those pressures. This prompted US senators to respond with the notorious Smoot-Hawley Tariff Act, which they did in 1930, in order to boost demand for domestic production. They attempted, in other words, to create additional demand by shifting US demand for foreign goods to US goods – basically attempting something akin to exporting their overcapacity problem – and in so doing force the brunt of the adjustment onto their trading partners. Their trading partners, not surprisingly, retaliated by closing their own borders to trade, causing international trade to decline by nearly 70% in three years, and thereby shifting the brunt of the adjustment back onto the US.
By reducing trade, each country was forced to adjust domestic supply to domestic demand. For countries with excess demand, that meant expanding production, whereas for countries with overcapacity, that could easily mean rising inventories followed by a slashing of production. There is an important lesson in here for us. In an overcapacity crisis current-account surplus countries are likely to be more vulnerable to trade war than are current-account deficit countries, because trade war implies an expansion of production in the latter and a contraction in the former. In the 1930s it was noteworthy that the current account surplus countries like the US suffered more deeply from the crisis than did current account deficit countries, especially, it seems, once barriers to trade were imposed.
What would Keynes say?
The US suffered a deep crisis in the 1930s, and its imposition of trade tariffs made things worse, not just because impediments to trade are costly to the global economy, but rather because it set off a trade war in which other countries too forced it into trade balance. The US trade surplus with Europe declined from just over $1 billion in 1929 to just under $400 million in 1934, and the total decline in the country’s trade surplus exceeded 10% of GDP. US excess production over consumption had to be resolved largely within the US, and that meant that either the US engineer a substantial increase in domestic demand by fiscal means, as Keynes demanded, or that it adjust via a costly drop in production and employment. It did the latter.
China today may be facing a similar problem. Today it is China who is exporting overcapacity and it is the US who is consuming too much, fed by Chinese financing. With the collapse of bank intermediation US households and businesses are cutting consumption and raising savings. This is a necessary adjustment. Most analysts, perhaps thinking they are echoing Keynes’ analysis of the problem in the 1930s, call on the US government to engage in massive fiscal expansion to replace lost private demand, but this is not what Keynes would have recommended. If declining US private consumption is met with increasing public consumption, the world will simply continue playing the game that has already led into so much trouble. The only difference would be that instead of having one side of the global imbalance accommodated by private US over-consumption and rising debt, it would be accommodated by public US over-consumption and rising debt.
But like in the 1930s, if there is a drop in global demand, it is countries with too little demand, the non-OPEC current-account surplus countries, who will need to adjust more than countries with too much demand. Because of the importance of the export sector to domestic growth and employment, if their exports drop quickly there may be significant political pressure for these countries to engineer moves to expand exports.
However since most of them lack large domestic markets, the result isn’t likely to be direct import tariffs. What we are more likely to see is direct and indirect export subsidies and competitive devaluations. Already in China policy-makers are raising export rebates, providing cheap financing to suffering exporters, and even discussing currency depreciation. Other Asian countries eager to boost exports are also considering ways, including currency depreciation, to strengthen their share of the declining export market, and all of this is happening even before the real impact of the global slowdown begins to appear.
If Keynes were around today he would probably make the same point he did over 60 years ago. Demand must be created by the current-account surplus countries who have, to date, relied on net exports to protect themselves from their overcapacity. They must force demand up quickly in order to close the gap, and since expecting private consumption to rise quickly enough is unrealistic, it has to be public consumption – a large fiscal deficit. But above all they should not try to grow their way out of overcapacity problem by reducing imports or increasing exports. In 1930 the US foolishly tried to dump capacity abroad by creating import restrictions (which have the effect of expanding domestic production), but the furious, and hardly unexpected, reaction of its trading partners caused the strategy to misfire and the US suddenly found itself forced to bear almost all of the adjustment on its own.
Might China repeat the US mistake? Perhaps. It already seems to be in the process of engineering its own Smoot-Hawley-with-Chinese-characteristics. Although there has been an attempt to boots fiscal spending, most analysts argue that this so far has been too feeble to matter much. On the other hand it has tried to protect and strengthen its export sector (resulting in rapidly growing exports and three record trade surplus months in a row).
This cannot work for long. The world clearly suffers from overcapacity, and as the US reduces its demand and increases its savings, which it must do as part of its own adjustment, this overcapacity will only rise. The proper place for new demand to originate is, as in the 1930s, in current-account surplus countries. They should be engaged in demand creation, not supply creation. If they continue trying to export their way out of a slowdown, there will almost certainly be a trade conflicts, as there were in the 1930s, in which case the full force of the adjustment will be borne by the current-account surplus countries, again as in the 1930s. The current-account deficit countries know this, and as the world’s economy contracts, their domestic tolerance for rising trade surpluses, or even just a continuation of trade surpluses at the current level, is likely to decline.
It is not hard to see why this could easily end in trade war. In order to avoid this possibility, the world’s major economies must engineer a joint program of fiscal expansion, in which the current account deficit countries expand moderately so as to slow down the adjustment period and to give maximum traction to fiscal expansion on the part of the current account surplus countries, who must inevitably bear most of the burden for demand creation. Fiscal expansion on the part of the current account deficit countries should occur with the clear understanding that it is a temporary measure aimed only at assisting the transition among China and other major current account surplus countries from an over-reliance on exports to absorb capacity.
The world cannot support indefinitely continued debt-financed overconsumption on the part of the US, whether this consumption takes place at the private or public level, and it cannot support continued growth in Chinese capacity without more rapid growth in Chinese consumption. To continue in this way almost certainly means little more than to postpone a larger and more difficult adjustment on the part of both countries, and will probably eventually lead to a collapse in international trade.
Michael Pettis is a finance professor at Peking University and the author of The Volatility Machine (Oxford University Press, 2001)