by Michael Pettis, China Financial Markets
Since the beginning of the global crisis in 2007-08 I have argued that the crisis was a consequence primarily of global trade imbalances generated by structural features that led to significant saving imbalances in China, the US, and within Europe. I describe this model in more detail in my recent book, The Great Rebalancing: Trade, Conflict, and the Perilous Road Ahead for the World Economy (Princeton University Press).
In that sense the current crisis shares a lot of characteristics with nearly every other global crisis of the past 200 years, as I point out in my book, and is likely to be resolved in similar ways: with a series of sovereign defaults or debt restructurings including, but not limited to, a number of European countries. None of the “globalization” cycles of the last 200 years has ended without widespread sovereign defaults except the one that ended in the First World War, and in that case the war caused soaring commodity prices and sharp constraints in Europe’s manufacturing exports, both of which were a tremendous help to developing countries. This probably why this was the only “globalization” cycle that did not end in massive sovereign defaults.
For those who are interested, by the way, I would argue that the first “modern” global debt crisis probably began in Britain in the mid-1820s. It spread to southern Europe and Latin America by 1825 (the first in a long series of Latin American sovereign debt crises), and continued on into the US in the 1830s with the default of a number of US states including, most shockingly, Pennsylvania, which was at the time one of the richest of the US states and among the richest economic entities in the world.
Others argue that the crisis that began in Vienna in May 1873, and spread to the rest of the world, most famously in the form of the railway crisis in the US by September of that year, qualifies as the first truly global debt crisis. The crisis of the 1930s which involved much of Europe and Latin America and a number of countries elsewhere was clearly a global crisis and in many ways the one that most closely resembles the current crisis. In addition I would add the crises of 1890s and the 1980s to my list of global debt crises. I don’t want to sound in this issue of the newsletter as if I am on a book-peddling mission, but I discuss all of these waves of financing followed by crises in one of my earlier books, The Volatility Machine: Emerging Economies and the Threat of Financial Collapse (Oxford University Press, 2001).
In the first three or four chapters of my book I examine the history of each of these periods, and try to show that it was not just, or even primarily, fundamental changes in the recipient economies that drove economic growth (or the political process of economic reform) but rather that exogenous changes in global liquidity determined the timing of the process. Among other things, I argued, this suggests that in order to predict the performance of individual countries, including the direction and pace of economic reform, it may be at least as important to understand external liquidity conditions as it is to evaluate domestic economic polices – something the much-discussed Fed tapering may be about to remind us again.
If my model for thinking about the global imbalances was an accurate description of the source of the crisis, there would obviously be a number of “predictions” arising from the model, which I had discussed in a number of newsletters and which, two years ago, I summarized in the form of 12 “predictions”. These predictions were:
- BRICs and other developing countries have not decoupled in any meaningful sense, and once the current liquidity-driven investment boom subsides the developing world will be hit hard by the global crisis.
- Over the next two years Chinese household consumption will continue declining as a share of GDP.
- Chinese debt levels will continue to rise quickly over the rest of this year and next.
- Chinese growth will begin to slow sharply by 2013-14 and will hit an average of 3% well before the end of the decade.
- If the PBoC resists interest rate cuts as inflation declines, China may even begin slowing in 2012.
- Any decline in GDP growth will disproportionately affect investment and so the demand for non-food commodities.
- Much slower growth in China will not lead to social unrest if China meaningfully rebalances.
- Within three years Beijing will be seriously examining large-scale privatization as part of its adjustment policy.
- European politics will continue to deteriorate rapidly and the major political parties will either become increasingly radicalized or marginalized.
- Spain and several countries, perhaps even Italy (but probably not France) will be forced to leave the euro and restructure their debt with significant debt forgiveness.
- Germany will stubbornly refuse to bear its share of the burden of the European adjustment, and the subsequent retaliation by the deficit countries will cause German growth to drop to zero or negative for many years.
- Trade protection sentiment in the US will rise inexorably and unemployment stays high for a few more years.
Finally one additional “prediction” which was not included in this list but which belongs here, was that the US would be the first major economy to emerge from the crisis of 2007-08 and China probably the last, although Europe might give China a run for its money.
Evaluating the predictions
It is still too early for all of these predictions either to have materialized or to have failed, but I thought it might be useful to review them to see whether or not they have been reasonably accurate in describing unfolding events and, if not, how my model for thinking about global imbalances should be revised. My reason for doing this is not so much to keep score but rather that these predictions were almost necessary or logical outcomes of the savings imbalance model I implicitly use to understand the world, and so to the extent that my model is valid it should show up in the evolution of these predictions.
Obviously a truly clever economist never makes a verifiable prediction, and if he does he should never refer to that prediction subsequently, so it is with a sense of trepidation that I do so, but here goes:
1. BRICs and other developing countries have not decoupled in any meaningful sense, and once the current liquidity-driven investment boom subsides the developing world will be hit hard by the global crisis.
For many analysts, this was one of my most surprising predictions at the time, but my reason for this was pretty straightforward. As I saw it the world had experienced policies, primarily in Germany and China, which had automatically forced up their respective savings rates by, in different ways, subsidizing rapid growth through hidden transfers from the household sector. These hidden transfers from households effectively constrained consumption growth, so the savings rates rose not because German or Chinese households became thriftier but rather because the median household retained a smaller share of GDP.
In Germany, as I have argued many times before, the constraint on consumption growth occurred as the combination of “voluntary” wage constraints around the turn of the century and a weak “German” euro caused by monetary union with countries that have structurally weaker currencies (and remember that an undervalued currency is effectively a tax on consumption that subsidizes the tradable good sector). In China it was a combination of slow wage growth (relative to productivity growth) financial repression and an undervalued renminbi.
In either case the net result was that the household share of total production in both countries declined, and with it of course the consumption share also declined, which is just another way of saying that national savings rates were forced up. There are only three ways the world can accommodate polices that force up the savings rates in one part of the world. Either global investment must rise by the same amount to balance the rise in savings, or the consumption rates in the rest of the world must rise so that global savings remain the same, or, if neither of these can happen, then growth must slow and unemployment rise to force down total savings (fired workers cause production to fall faster than consumption, which reduces savings).
To put it a little more schematically, in a two-country world, if the savings rate in Country A rises, then one, or a combination, of the following three things must happen:
- Investment in Country A and Country B must rise
- Consumption in Country B must rise
- Unemployment in Country B must rise
I have explained elsewhere why these savings imbalances led to the crisis, but the main point is that until the crisis, the counterbalance to the structural rise in savings in parts of northern Europe and Asia was in part an increase in real estate investment in the US and peripheral Europe (much of it, it turns out, perhaps not surprisingly, to be excessive) and, with booming stock markets partly a consequence of excess liquidity creating a strong wealth effect, a collapse in savings in the US and peripheral Europe, which is simply the obverse of a rise in consumption.
The emerging markets benefited, as they always have in previous similar cases, from both soaring consumption in the US and peripheral Europe and soaring investment in China which, as I have explained elsewhere, was part of the same set of policies that drove up the savings rates and which required excess consumption in the north to accommodate the excess of Chinese production over total Chinese demand. Once this excess consumption dropped, as it had to when the financial crisis eliminated both the booming markets that created the wealth effect and the easy availability of credit, it would eliminate the demand that created such rapid growth in the emerging markets.
But at first this seemed like it wasn’t happening, thus sparking over-excited claims that the developing world had finally decoupled from the developed world and was capable of sustaining its own growth. Anyone familiar with what happened in the mid-1970s should have treated these claims with extreme skepticism. In the 1970s the developing world also seemed to have decoupled from crisis in the developed world, only to suffer a much worse crisis in the 1980s.
It turned out that the contraction in demand from the developed world was counterbalanced by a surge in investment – once again fed by cheap and abundant capital, just as it was in the 1970s with the recycling of petrodollars – in the developing countries. China of course led this investment surge, but because of its ferocious demand for hard commodities to satisfy the building of near-infinite amounts of real estate developments, infrastructure, and most bizarrely of all, additional manufacturing capacity, countries like Brazil, Peru, Australia Chile and other commodity exporters also significantly increased investment in commodity production.
But the purpose of investment today is to serve consumption tomorrow, and with the long-term slowdown in consumption demand from the developed world, the surge in investment only took investment levels in countries like China, which were already much higher than their ability to absorb these investments productively, to even higher and harder to justify levels. Instead of bringing investment down sharply, which is what economic rationality would have required, in other words, there was a sharp increase in investment.
Of course this was clearly not sustainable. It had to end, and when it ended, the emerging markets would face the double whammy of
a) sharply reduced consumption growth in the developed world no longer counterbalanced by surging investment in China and the commodity exporters and
b) a huge increase in debt that exceeded the marginal increase in debt servicing capacity. This meant that rather than decouple from the developed world which was undergoing a deleveraging process, emerging markets simply continued leveraging up until the point where they could no longer do so, in which case they would have an exacerbated reaction to the deleveraging process in the developed world.
There was no emerging market decoupling, in other words, except in the writings of some overexcited sell-side analysts. This was always a completely incoherent claim and two years ago I warned that
“by the middle of this decade the whole concept of BRIC decoupling will seem faintly ridiculous”.
We are already beginning to see this, but I would argue that we are only at the beginning of the process. The next three or four years are going to see much greater pain among the emerging markets and probably more than one sovereign default or restructuring.
Given the huge change in sentiment in the last year that has already taken place about prospects for the emerging economies, and the sharp slowdown in growth in China and other developing countries, I am tempted to declare victory and say this prediction was correct, but of course it is much too early to say anything of the sort. If I am right, it must get worse. We have only just started to recognize the impact of slower investment growth.
2. Over the next two years Chinese household consumption will continue declining as a share of GDP.
I expected consumption to continue declining as a share of China’s GDP because the low consumption share was a consequence of the low household income share, which was itself an automatic consequence of China’s growth model. The consumption imbalance, in other words, could not change until the model was basically abandoned, which I did not expect to see until the new administration came in and recognized how the growth model was leading inexorably to a credit problem.
The next banking crisis will have to be paid for by transfers from the state sector.
In fact the unsustainability of the surge in debt became apparent much earlier than I expected, and it seems that the consumption imbalances may have stabilized. I say “may have” because debt has continued to surge in the past two years and with it, problem loans almost certainly have too. There is a relationship between bad loans and the consumption share of GDP, of course, and this is because in the past bad loans were resolved in the form of repressed interest rates that effectively passed their costs onto the household sector, which was the single most important reason, in my mind, for the astonishing drop in the consumption share of GDP during this century.
We don’t know how the next banking crisis will be resolved. The old way – forcing the household sector to pay in the form of negative returns on their deposits – is I think still the default thinking among policymakers, but I don’t think it can possibly work again. Not only is the household share of GDP much to low to tolerate the huge transfers needed to resolve the next banking crisis, but with Beijing determined to rebalance the economy away from investment and towards consumption, it should eventually become pretty clear, if it isn’t already, that the next banking crisis will have to be paid for by transfers from the state sector.
This means that the next few years are critical. Nominally the consumption share has stabilized, which, I am glad to say, means my prediction may have been too pessimistic, but we won’t really know until we see if there is indeed a surge in bad loans in the next two to three years and how Beijing responds.
3. Chinese debt levels will continue to rise quickly over the rest of this year and next.
I don’t think I need to say too much about this prediction – no one doubts anymore that debt is growing much too quickly. Over the past three days the Financial Times has published a series of very interesting if alarming articles about debt in China and it is hard to pick up a business periodical nowadays that doesn’t discuss the topic.
But there really should never have been any surprise about China’s unsustainable debt path. I argued that given the systematic tendency to investment allocation that has driven growth for the past several years, an unsustainable increase in credit is a necessary condition for GDP growth levels much above 3-4%, not an accident caused by spates of irresponsible lending, and whether or not you agree with me I think it would be hard to disagree that credit growth has been astonishing.
For my model to be right, either GDP growth must slow significantly in the next year or two, or if it doesn’t, credit will continue to surge dangerously and perhaps reach debt capacity constraints within two or three years. The consensus seems to be that growth won’t slow. On Monday, for example, Credit Suisse, sort of in the middle of the pack, wrote in their research note that
“We call for growth to stabilize, without much upside momentum. We revise up our forecasts for 2013 GDP growth to 7.6% from 7.4% and 2014 to 7.7% from 7.6%.”
It is only if the growth forecasts by analysts like those at Credit Suisse are right, and, instead of surging, credit growth slows substantially, that this aspect of my interpretation of China’s growth model will have been proven wrong.
4. Chinese growth will begin to slow sharply by 2013-14 and will hit an average of 3% well before the end of the decade.
Again I don’t need to say too much about this prediction. The first part obviously turned out to be dramatically right – in fact growth began to slow sharply in 2012 for reasons I discuss in the next “prediction”. Whether the second part will also be right is exactly where the debate is now, and I stick to my prediction for the reasons I discussed in my August 8 newsletter. The key, as I note above, is whether China can engineer GDP growth rates much above 3-4% without even more rapid growth in credit.
5. If the PBoC resists interest rate cuts as inflation declines, China may even begin slowing in 2012.
This was a pretty good prediction in spite of the fact that only the most ferocious of bears didn’t agree, and to its credit (or, as rumors have it, to the credit of Li Keqiang even before he became premier), the PBoC did not cut interest rates. Growth indeed began slowing sharply in 2012. The same story stands. If the PBoC begins to cut interest rates sharply, growth will pick up but debt will explode.
6. Any decline in GDP growth will disproportionately affect investment and so the demand for non-food commodities.
I argued at the time that as a result of the disproportionate impact of a GDP slowdown on Chinese demand for hard commodities, the price of hard commodities would drop by over 50% in the next five years. So far this seems to be happening, perhaps even faster than I predicted. According to an article three weeks ago the Wall Street Journal “copper prices have dropped 30% from their 2011 peak, and iron ore is down 32%,” while the Economist shows an interesting graph on iron ore prices in a recent issue which I cannot reproduce here.
The price of hard commodities would drop by over 50% in the next five years [from 2011].
According to the graph iron ore traded around $190 at the time I made my prediction and has dropped since then to around $120. The only additional comment I would make is that I think the decline in the prices of hard commodities from their peaks will turn out to be much greater than 50%. I wouldn’t be surprised at all if iron traded well below $50, for example, within the next three years.
7. Much slower growth in China will not lead to social unrest if China meaningfully rebalances.
In a recent piece I wrote for the Financial Times I argued that what matters to ordinary Chinese households is not GDP growth but rather growth in household income. In the article I said:
Simple logic shows that it is nearly impossible for China’s GDP to grow at current rates while rebalancing away from its dangerous over-reliance on exports and debt-fuelled investment. Consider what it means for China to rebalance. Household consumption, at an astonishingly low 35 per cent of GDP, is just over half the global average.
Attempts to engineer a rebalancing that lifts consumption over the next 10 years to, say, 50 per cent – which will still leave it with the lowest consumption share of any large economy in the world – would require consumption growth to exceed GDP growth by close to 4 percentage points every year. So an average annual GDP growth rate of 6 or 7 per cent requires average growth in consumption of nearly 10-11 per cent for a decade for China to rebalance meaningfully.
Consumption growth rates of 10-11% for the next decade are simply the arithmetical implications of a rebalancing China growing at 6-7%, and it is surprising to me that few analysts who think China will indeed continue growing at 7% while rebalancing its economy have explained how it would achieve this level of consumption growth for a decade. China was not able to achieve such high consumption growth rates even in the best of times, when it and the world were growing much more briskly, and it seems to me that, absent a massive transfer of resources from the state sector to the household sector, it will prove near impossible for China to manage such high consumption growth under much weaker Chinese and global conditions.
The consumption rate in China is low, as I have written many times, mainly as a consequence of policies that systematically transferred resources from the household sector to subsidize rapid growth. This forced down the household income share of GDP, which, at about 50%, is among the lowest ever recorded in the world. There is no sustainable way to boost household consumption without boosting household income.
This suggests that consumption growth of 10-11 per cent requires similar growth in household income. In principle China could have this by paying workers much higher wages, sharply revaluing the currency, and sharply raising the deposit rates paid by banks, but since low wages, an undervalued currency and cheap capital are at the heart of China’s growth model, raising wages, the currency and deposit rates enough to rebalance the economy could spread financial distress and cause growth to collapse too quickly. Only a continued, and ultimately self-defeating, surge in debt can get household income to grow quickly enough to accommodate both high GDP growth rates and a rebalancing economy.
This is why GDP growth rates must drop further. But after many years of annual GDP growth above 10 per cent, it would seem that a sharp drop in GDP growth rates to below 6-7 per cent would clash with the rising expectations of ordinary Chinese. Won’t slower growth lead to social unrest and perhaps political chaos? Not necessarily. My Financial Times article concluded by saying:
For China successfully to rebalance towards a healthier and more sustainable model without unrest, the growth rate that really matters, as a number of prominent Chinese economists have already noted, is that of median household income. Ordinary Chinese, like people everywhere, do not care about their per capita share of GDP. They care about their income.
In recent decades real disposable income has grown at well above 7 per cent a year on average. To ensure social stability, it should continue growing at this rate or close to it. But growth in household income and household consumption of about 6-7 per cent implies that, if China is to rebalance meaningfully, GDP must grow by “only” 3-4 per cent. This much lower rate is consistent, among other things, with almost zero investment growth.
China’s GDP, in other words, does not need to grow at 7 per cent or even 6 per cent a year in order to maintain social stability. This is a myth that should be discarded. What matters for social stability is that ordinary Chinese continue to improve their lives at the rate to which they are accustomed, and that the Chinese economy is restructured in a way that allows it to tackle its credit bubble.
If household income can grow annually at 6-7 per cent, income will double in 10 to 12 years, in line with the target proposed by Premier Li Keqiang in March during the National People’s Congress. What is more, if China can do this while the economy is weaned off its addiction to credit, it will be an extraordinary achievement, even if it implies, as it must, that GDP grows far more slowly that the growth rates to which we have become accustomed.
8. Within three years Beijing will be seriously examining large-scale privatization as part of its adjustment policy.
The only relatively quick to rebalance the Chinese economy (and they probably don’t have time to do it gradually), which by definition means that households must retain a higher share of GDP and the government a lower share, is to transfer assets from the state sector to the household sector. This will not be easy.
Privatization is the most efficient way to do it, but there will be tremendous political opposition to doing so. I am pessimistic about the likelihood of a serious privatization program, but there is no question that it is being increasingly discussed, for example in a recent article in the New York Times by Chinese banker Joe Zhang, in which he says:
Thus, to avoid long-term stagnation, China will soon have to embark on a second wave of privatization, as the government is running out of options to fund further growth. Its fiscal deficits have risen quickly, banks are overextended and land sales have become untenable. Moreover, there are growing calls for the government to replenish the inadequate social security fund to meet the needs of an aging population.
This inevitable wave of privatization will bring with it the added benefit of unleashing a corresponding wave of restructuring in what has become a grossly inefficient and bloated economy. Luckily, the new leadership seems to recognize that this is necessary. A consensus appears emerging in the policy making circles that, as fiscal stimulus runs out of steam, structural reforms can provide huge potential for the next leg of development.
I expect these discussions to get more heated in the next two years.
9. European politics will continue to deteriorate rapidly and the major political parties will either become increasingly radicalized or marginalized.
There certainly has been some deterioration in European politics, but I confess I am surprised by how little has happened so far. Still, for the reasons discussed in the 12th prediction, below, I do not expect conditions to improve.
10. Spain and several countries, perhaps even Italy (but probably not France) will be forced to leave the euro and restructure their debt with significant debt forgiveness.
I continue to be quite certain of this but, as I said in my original prediction, we will probably have to wait another two or three years before the debate about leaving the euro becomes central to the political discussion (I think only in Portugal and Greece is this debate are taken seriously). As an aside it is worth remembering that every time a weak country leaves the euro, the currency is likely to strengthen and so the overall burden of adjustment will, at the very least, remain the same while being concentrated on a smaller group of countries.
11. Germany will stubbornly refuse to bear its share of the burden of the European adjustment, and the subsequent retaliation by the deficit countries will cause German growth to drop to zero or negative for many years.
So far, as I expected, we have not seen a serious reflation of German demand that will allow Europe to adjust without either punishingly high unemployment for many years or an attempt to force European imbalances abroad in the form of a huge European trade surplus. I discuss this in much greater detail in the May 9, 2013, issue of the newsletter, and in a Voxinterview earlier this year. Without a German reflation, which they are unlikely to do because of their own debt concerns, I don’t really see many other options for Europe.
If Germany does not take dramatic steps to push its current account surplus into deficit, the brunt of the European adjustment will fall on the deficit countries with a sharp decrease in domestic demand. This is what the world means when it insists that these countries “tighten their belts”. If the deficit countries of Europe do not intervene in trade, they will bear the full employment impact of that drop in demand – i.e. unemployment will continue to rise. If they do intervene, they will force the brunt of the adjustment onto Germany and Germany will suffer the employment consequences.
For a few more years the deficit countries will try to bear the full brunt of the adjustment while Germany scolds and cajoles from the side. Eventually they will be unable politically to accept the necessary high unemployment and they will intervene in trade – almost certainly by abandoning the euro and devaluing. In that case they automatically push the brunt of the adjustment onto the surplus countries, i.e. Germany, and German unemployment will rise. I don’t know how soon this will happen, but remember that in global demand contractions it is the surplus countries who always suffer the most. I don’t see why this time will be any different.
12. Trade protection sentiment in the US will rise inexorably and unemployment stays high for a few more years.
Trade policy in the next few years will continue to be about deciding who will bear the brunt of the global contraction in demand growth. The surplus countries, because they are so reliant on surpluses, will be very reluctant to eliminate their trade intervention policies, and in fact the current European plan, as I discussed in my June 11 newsletter, is to force the cost of the European adjustment onto its trading partners by running a large surplus. In principle this mostly means the US. Because they are making the same mistake the US made in the late 1920s and Japan in the late 1980s – thinking they are in a strong enough position to dictate terms – the surplus countries will refuse to take the necessary steps to adjust.
Surplus countries … are making the same mistake the US made in the late 1920s and Japan in the late 1980s.
But in fact in this fight over global demand it is the deficit countries that have all the best cards. They control demand, which is the world’s scarcest and most valuable commodity. Once they begin intervening in trade and regaining the full use of their domestic demand, they will push the adjustment onto the surplus countries. Unemployment in deficit countries will drop, while it will rise in surplus countries.
13. The US will be the first major economy to emerge from the global crisis, and China will be the last.
The US, and deficit countries generally (unless they are suffering from serious financial distress), tend to adjust fairly quickly and I expected that this time would be no different. I expected the adjustment to take much longer in China because it was always likely to be politically very difficult and because the system was more rigidly locked into the policies that underlie the imbalances.
Although this was another one of the predictions that seemed to cause a lot of surprise and disbelief when I first proposed it in 2009, and even two years ago, I am pretty sure that by now no one is surprised by the claim that the US is in the process of recovery (although it can still be derailed). Nor are many surprised by the claim that China has only just begun the adjustment process (although there are still China bulls who believe China has already bottomed out).
As an aside, I have been told by one of my friends that an English economist at a Chinese university, who found this to be the most preposterous of my predictions, is now bizarrely enough arguing that the proof that China has emerged from the crisis sooner than the US is that Chinese growth rates are still higher than US growth rates. If China is growing faster than the US, he claims, this proves that China has turned the corner of the crisis before the US has. I suppose that if Spain was growing faster in 2008 than the US, he would have argued that Spain had already put the crisis behind it.
Aside from the fact that we don’t really know what China’s real growth rates are without correctly writing down bad debt, as my Spanish example shows, this claim is completely irrelevant and more than averagely muddled, especially given that my prediction for the upper limit of Chinese growth during this decade – 3-4% annually on average – is likely to be higher than US growth rates anyway.
If US growth has more or less bottomed out, and if savings rate are on a long-term upward path, and if Chinese growth has not yet bottomed out and the savings rate is forced further downward, as I expect, the US will have emerged from the crisis much sooner than China. It doesn’t matter if Chinese growth rates are currently, or ever, higher than that of the US. This should be obvious.