It wasn’t enough that we started 2016 with one of the worst weeks in the recent history of Chinese and global markets, but the panic continued into the following weeks and wreaked a great deal of damage to confidence. A lot of the reflexive China bulls are cautioning against misinterpreting the implications of the stock market collapse, and of course they are right, but the fact that the plunging Chinese markets can easily be misinterpreted should not in any way suggest that things are fine. Two weeks ago in the FT Alphaville blog (which is the best place to read regularly about China’s vulnerabilities, in my opinion, especially in their relentless focus on the changes in the various components of the balance of payments) Peter Doyle discussed one of the standard set of responses that we’ve seen repeated regularly since 2011 and 2012. The bull refrain has been, in his words: “things really aren’t that bad or surprising, and there’s considerable willpower and ammunition left in Beijing should it be necessary.”
“This makes good copy,” Doyle suggests, and adds, more diplomatically than I might have, “but is not persuasive”. It certainly isn’t, and he discusses some of the reasons why. On the same day George Magnus published an OpEd in the Financial Times that makes a point that too many people, as he points out, are still overlooking.
China’s chief vulnerability, and the main factor driving the tendency towards the increasingly fragile balance sheets that underlie the series of interrelated financial-market disruptions that began seriously in June 2013, is the inexorable rise in debt, and any analyst that fails to come to grips with the problem of excess credit is simply wasting his time. In the article Magnus writes:
Important economic reforms to the real economy and state monopolies have stalled, or succumbed to inertia and pushback. Policies designed to develop new sectors have not been matched by those needed to tackle problems in larger ones, such as poor productivity, chronic overcapacity and now a fourth consecutive year of producer price deflation. Tellingly, China’s most serious problem — the relentless accumulation of debt — received passive attention at most.
I will return to his point about stalled reforms, but Magnus is right about the debt. China’s most serious problem is “the relentless accumulation of debt”, and economic conditions will continue to deteriorate until Beijing directly addresses the debt. In fact it doesn’t really matter if China is able to report growth rates for another year or two of 7%, or 6%, or even 8%. If the only way it can do so is by allowing debt to grow two or three times as fast, there will have been no improvement at all, the economy will not have adjusted, and China’s longer-term outlook will be worse than ever.
It is only when credit growth begins to decelerate much more rapidly than nominal GDP growth that we can begin to talk hopefully about China’s moving in the right direction, and it is only when credit growth falls permanently below the growth rate of the economy’s debt-servicing capacity that China will have adjusted. The astonishing ability of the China bulls, both foreign and Chinese, to celebrate every unexpected decline in growth and every new surge in debt as if they somehow justified nearly a decade’s worth of denials of the urgency of China’s rebalancing has done so much damage to China that the sooner Beijing’s leaders finally turn against the bulls, as I believe they might finally have done, the better for the Chinese people and the Chinese economy.
Beggaring thy neighbors
Before I explain why I think Beijing has decided that it has been misled in recent years, I should point out that what worried me most about the events of the past weeks was not the stock markets themselves, nor even the change in how investors and businesses inside and outside China perceive Beijing’s ability to manage the economy and the markets (I have already said many times that just as most people systematically over-rated the quality of Chinese policymaking in the past, they are likely now to be overly harsh in accusing Beijing of mismanagement). I am far more worried about how other countries might misinterpret the rapid decline in the RMB, accompanied by what seems like another surge in capital outflows.
Contrary to some of the muttering out there, I don’t think Beijing is planning competitive devaluations in order to strengthen the tradable goods sector, in the hopes that surging exports will revive growth. Certainly if the PBoC ever were to stop intervening, and to let the RMB depreciate to some imagined fundamental “equilibrium”, we would quickly see that there is no such equilibrium level. In a speculative market, the market does not tend towards some stable value, with self-dissipating movement in any one direction reducing pressure for further movement in that direction. Price movements instead are self-reinforcing, and can quickly overshoot fundamentals.
Beijing is more likely to believe that the economic slowdown was caused by been weakness in domestic real estate and infrastructure construction, and not because exports are weak, and the latest trade dataconfirms the relatively strong export performance. Although manufacturing overcapacity is certainly a problem, much of it is in areas in which global demand has simply collapsed, and isn’t coming back, and so a cheaper currency would have little impact beyond temporarily reducing excess inventory, which is not enough of a benefit to justify the many costs of a weaker currency. Production facilities would still have to be closed down.
I think the real reason for the recent RMB weakness lies elsewhere. Beijing is trying to boost domestic liquidity in the hopes that this will generate stronger domestic demand, but expanding liquidity fuels capital outflows, and these put downward pressure on the currency, while increasing PBoC concerns about the monetary impact of money leaving the economy which, as an article in last week’s FT argues, might be worse than we think. Last week’s People’s Daily reports that prominent Tsinghua professor and former member of China’s Monetary Policy Committee, Li Daokui, claimed at Davos “that at least $3 trillion foreign exchange reserves in China is required to prevent foreign debt default risk”, for reasons that elude me, but if this reflects official views, after dropping $513 billion in 2015, current PBoC reserves of $3.33 trillion might suggest that two or three more months of continued strong outflows might prompt further steps by the PBoC to limit outflows.
The biggest risk created by the weaker RMB, as I see it however, is not a Chinese risk but rather a global one. The rest of the world may view recent Chinese RMB weakness as a signal for a new round of competitive devaluations. I have already said that I expect 2016 to be another bad year for trade, and I am worried that it seems as if every major economy in the world has implicitly decided to use US demand to bail out its own faltering economy. This will very likely derail the US recovery in 2016 or 2017 unless the US, too, decides to step in and intervene in trade. If that happened, of course, the impact on Europe and China would be terrible, but it seems to me a matter purely of logic that if the hard commodity and energy exporters are nearing the limits of their absorption capacity, either the major surplus nations or the US are going to have to absorb a bigger share of the demand deficiency created in Europe, China, and Japan.
The nine-point summary
But all this is preamble. Rather than add to the mass of coverage that the recent market events in China have generated, or to continue expressing my concern about the intractable arithmetic of global demand imbalances, I plan to discuss the process of Chinese reform and adjustment in this issue of my blog. While these may at first seem unrelated, in fact financial market disruptions are tightly tied into the self-reinforcing processes of rising debt, capital flight and slowing growth that recent reforms were supposed to untangle and address – and for which they have clearly failed.
I will argue that most economists have an incoherent understanding of China’s rebalancing needs, and for this reason many if not most of the reform proposals of the past few years, about which economists widely agree and even celebrate, are in many if not most cases largely irrelevant. This is going to be a long post, so for those who want the 9-point summary:
China’s economic growth is not decelerating as a natural consequence of the aging of China’s growth model. It is decelerating for three reasons. The first reason is the reversal of the growth process by which China’s imbalances have reached their systemic limits.[1]
The second reason is that during the phase of rapid growth, China’s balance sheets, as occurred in every similar case, evolved to become highly inverted, and just as this automatically caused growth to be higher than expected during the expansion phase, it must cause lower-than-expected growth during the contraction phase.
Finally, the economy must shift, one way or another, from one of rising leverage to one of declining leverage, and with rising debt the only thing propping up growth levels, deleveraging cannot help but cause growth to drop.
This means that regardless of trends in underlying productivity, growth must slow sharply, and it will, either smoothly and continuously, or in the form of higher growth early in the adjustment period and a collapse in growth later.
The growth deceleration can be temporarily countered by rapid increases in debt, but ultimately this will only increase future deceleration, with a rising chance that the shift will be disruptive. Every growth miracle in history has been followed by an unexpectedly difficult adjustment, and it is unreasonable to have expected that China would be any different.
The only way to minimize the costs of the adjustment is to take steps to speed up the rebalancing of demand and the repayment of debt. This must be the direction of reforms if Beijing is going to reduce the costs of adjustment and the risk of a disruption.
Repaying debt simply means allocating debt-servicing costs, either directly or indirectly, to specific sectors within the economy. This will either occur in ways targeted by policymakers, or if postponed for too long, it will occur in unpredictable ways determined by circumstances. For example default allocates the costs to creditors, inflation allocates the costs to household savers, economic collapse and high unemployment allocates the costs to workers, etc.
By far the most efficient ways for Beijing to minimize the adjustment costs for the economy and reduce the risk of a debt-related disruption is to allocate debt-servicing costs to local governments by forcing them to liquidate assets directly or indirectly to pay down debt, and to increase household wealth by transferring wealth directly or indirectly from local governments to the household sector. Successful reforms must be consistent with these two goals.
Beijing has already tried to address its growth problems by implementing the productivity-enhancing reforms beloved of orthodox economists, but while these might be a good idea in normal times, they will have almost no effect on reducing the cost of China’s economic adjustment.
Evaluating Beijing’s policies
It might seem exceptionally contrarian to say that most of the reform proposals of the past few years, about which economists widely agree and even celebrate, are in many if not most cases largely irrelevant, but I think that senior policymakers in Beijing are beginning to agree. I say this because with the end of the Central Economic Work Conference last December, Beijing has announced with some fanfare that it plans to design and implement a new reform program consisting of what are being called “supply-side” policies.
A new reform program would seem only to be necessary if the old one has failed or is failing. It does seem to have failed. Chinese businesses and investors are very obviously increasingly concerned about both Chinese and global prospects, and so it is fitting that stock markets have been so accident prone this year. Like everyone else I have often cautioned against reading too much about China’s economic fundamentals into stock market performance. To the extent that there is informational content in the price behavior of stocks, however, we are more likely to see it expressed in the volatility of the markets than in its actual price level.
While opinion is still very split about the outlook for the Chinese economy, there is clearly a growing sense of unease about progress to date on the successful management of China’s economic adjustment, and this unease is at least part of the reason for market volatility. That is why the important news for me has been the announcement of the new reforms and the form of the announcement. Analysts are still very uncertain about what this new package of supply-side reforms that we’ve been hearing about since at least November may entail, but they way in which the reform package was announced suggests, at least too me, that the leadership is no longer satisfied that the policies Beijing has been pursuing during the past three years are having the intended effect.
I will discuss why later, and in spite of the limited information available, I also plan in this blog entry to try to place the package of reforms in some sort of useful context and to evaluate whether or not in principle these reforms are likely to be consistent with the rebalancing process. This might not be as difficult a task as it may at first seem, because rather than try to guess what Beijing will or won’t do, I will instead try to specify the conditions, albeit very abstractly, under which various policies will, individually or in the aggregate, be consistent with the rebalancing process.
It helps of course that China’s development model is not unique, and that its experiences are “different” only in the sense that with its powerful and rigid institutions (most importantly the extensive government involvement in the economy and its control of the banks) it has pushed the typical imbalances associated with its growth model often to levels that are unprecedented. There have otherwise been dozens of other countries that have experienced investment-led “miracle” growth in the past century, and their histories have been remarkably consistent.
Based on these histories we should be able to make fairly reasonable predictions about some of the problems that China faces today. Indeed we should have been able to do so a decade ago, but because very few economists seem to be familiar with the history, no matter how predictable they were each new reversal or systemic shift always seemed to come as a surprise. This is an important point to stress, especially if we want to understand what kinds of policies are likely to prove useful over the next several years. China’s extremely successful growth model was always likely to generate sustainable and rapid if unbalanced growth under certain easily-specified conditions.[2] It was also always likely to cause the financial sector and the various government and business balance sheets to evolve in a specific direction and imbed certain risks.
This is why nearly a decade ago – even if none of the economists or analysts writing about the Chinese economy understood why, with the exception of a handful, mostly Chinese academics – it was clear that the Chinese growth model would have to be reformed, and that these reforms were generally fairly easy to specify. Imbalances can persist for many years if the institutional constraints preventing adjustment are very strong, but all unbalanced systems tend towards rebalancing, and eventually the institutional sources of the imbalances are reversed, and they do.
Must everything cause a crisis?
Not everyone sees things this way. In the standard economic framework the economy is always at or close to equilibrium, and when exogenous shocks or policy distortions push it away from equilibrium it is only temporary.[3] For that reason most economists seem to assume that the longer what looks like an imbalance persists, the less likely it is really to be an imbalance that must ultimately reverse itself, when in fact the opposite is true. Imbalances can persist and get deeper year after year, but that only means that the reversal is more certain and likely to be more disruptive.
For the same reason most economists also seem to assume that if anyone thought that the Chinese economy was deeply imbalanced, and that debt was growing at an unsustainable pace, he was necessarily predicting an imminent crisis. This belief is so powerfully embedded in the standard equilibrium models most economists use that, strangely enough, even those of us who described the imbalances in one paragraph and in the very next paragraph insisted that a crisis was unlikely – in China’s case because of the government’s very high credibility and its role as financial guarantor – were automatically assumed to be predicting an imminent crisis.
Earlier this year, for example, a strategist at a Chicago-based fund by the name of Brian Singer said:
“Everyone is aware of China’s horrid debt levels and [Chinese financial markets analyst] Michael Pettis has done some great work, but he is China’s ‘Chicken Little’,” Singer said, referring to his panic-style predictions. “He discovered through his research that China has built up a lot of debt and he is right. China’s debt to gross domestic product (GDP), however, is pretty close to the United States’ and Germany’s.
“If we’re all so terribly concerned about China’s debt to GDP levels, why aren’t we equally as concerned about the US?…Even if [China’s] growth is slowed from double digits down to 4 or 5 per cent, they would still be absorbing that debt a lot faster the US could.”
I have already explained many times why comparing US and Chinese debt is nonsensical, and should be very obviously so, but to the extent that Singer, who seems an intelligent person, finds it impossible to accept that countries whose already-high debt levels are rising unsustainably (which simply means that debt is rising faster than debt-servicing capacity) do not necessarily default or collapse soon afterwards, it can only be because the economic model he implicitly uses is incoherent and that he is unfamiliar with financial history. Not only can an unsustainable rise in debt persist for many years, in some cases even decades, as Japan may one day prove, but in fact most unsustainable debt burdens are not resolved by crisis or collapse.
They are always resolved by mechanisms in which debt-servicing costs are explicitly or implicitly allocated to some sector of the economy, usually unwillingly, and while default or some form of financial crisis is one of the usually-explicit ways (the losses are assigned to creditors, obviously enough), it can often take many years before it happens, and it does not even happen in the majority of cases, as I will explain later in this entry when I discuss some of the ways in these debt costs have been allocated.
So while I have never predicted a crisis, panic-style or otherwise, I certainly have pointed out very early on that Chinese growth had become dependent on an unsustainable relationship with debt. While it was always possible that after many more years this could lead to a crisis, I always noted that a crisis was unlikely and most certainly not imminent. So why would Singer (and, to be fair, most other economists who use conventional equilibrium models) have found it impossible to see the sentences in which I said crisis was unlikely, once they read sentences in which I said imbalances were deep? It is because the two statements are incompatible in their models, which exclude ordinary balance sheet dynamics. The systemic creation and reversal of imbalances are not formally captured in these models.
My point is not simply to correct any misperceptions about what I did or didn’t say – in fact I have to confess that it hasn’t been all bad: in the past year because of similar mistaken models I have received a huge amount of credit from very generous people for correctly predicting market crises that in fact I did not predict. My point is rather about the incoherence of conventional thinking about unbalanced economies. The wide-spread inability intuitively to understand disequilibrium explains at least in part the misplaced confidence so many people have in the role Chinese reforms of the past few years would have in resolving China’s economic vulnerabilities (and those of peripheral Europe, for many of the same reasons) because it made improving productivity, rather than the rebalancing process itself, the main objective of the reforms. Refocusing on the rebalancing process will allow us to see not just why many of the old reforms simply didn’t matter, but also which of the new supply-side reforms might in principle work and which cannot.
Debt isn’t irrelevant
China’s unsustainable rise in debt is part of a self-reinforcing dynamic involving the consumption imbalance, and the important point is that because imbalances necessarily must reverse themselves eventually, any useful reform must be consistent with China’s economic rebalancing. In fact it should have been obvious years ago that Chinese policymakers only ever had two choices. They could proactively implement the reforms aimed at reversing the imbalances, however costly these reforms might be (and the longer they were postponed, the more costly they would become), or they could try to postpone the necessary reforms indefinitely – as many if not most countries in similar circumstances had done unsuccessfully in their attempts to sidestep the costs of rebalancing.
In the latter case, however, they would have almost certainly discovered, as their predecessors always discovered, that as the debt burden increased, the increasing impact of the distortions caused by the imbalances – and it is important to remember that the two are self-reinforcing – would eventually break through the institutional constraints that had prevented adjustment earlier. Rising balance sheet fragility makes an economy more sensitive to imbalances and to disruptive shocks, by which I mean that and balance sheets become increasingly fragile, their disruptive unraveling can be caused by progressively weaker random shocks – i.e. “triggers” – so that in very extreme cases it takes deceptively minor shocks to trigger major disruptions.[4] The risk in that case is that the imbalances would reverse themselves disruptively and force an unwanted resolution of the excessively high debt burden – of which the most obvious, but not only or even most likely, way would be in the form of a financial crisis of some sort.
I will give concrete examples later of how this has happened in previous cases, but put in starker terms, when for structural reasons economic growth in any country depends on an unsustainable increase in the debt burden, then simply as a matter of logic policymakers have two choices. They must either move aggressively to bring debt under control before the economy reaches its debt capacity limits, in which case they will cause growth to slow sharply over a difficult adjustment period during which balance sheets are rearranged. Or if they wait too long – usually because they mistakenly believe there is a set of efficiency-improving reforms that can cause productivity to rise faster than debt – at some point, as the debt burden continues to rise, either creditors will refuse to lend, in what economists call a “sudden stop”, or debt capacity limits will otherwise be reached, after which, in either case, growth will collapse. China must prevent its debt burden from reaching these levels.
Although it would have been much better for China if policymakers had recognized the urgent need to rebalance and implemented the necessary changes a decade earlier during the administration of Hu Jintao, it seems that Xi Jinping’s administration acknowledges the risk of continued credit expansion and wants to implement the necessary reforms before the economy is forced into a disruptive rebalancing. This, in effect, was what had been recognized during the October, 2013, Third Plenum. What is less clear to me, however, is how much time policymakers believe they have in which to force through the necessary reforms, and whether their assessment is realistic.
Right and wrong reforms
However much time they have – and in my opinion they are unlikely to have much more than 2-3 years in which to get credit growth under control, but there is no science to this so I cannot know for sure – as Beijing moves forward in its struggle to rebalance the Chinese economy, we should keep three things in mind:
The rebalancing process is in fact fairly straightforward and easy to understand or describe, albeit only conceptually and at a very abstract level. Both the logic of China’s existing growth model and the almost uniform experiences of the many historical precedents reveal clearly the vulnerabilities China faces, what it must do to address them, and why the necessary reforms will be difficult.
But while we should know what must happen, the specifics of the rebalancing process can be extremely complex, and depend very much on institutional conditions, at both the national level and the local level, and which include very powerful vested interests created by the development model. This is why even if it is clear in principle what must be done, choosing and implementing the actual reforms will be neither easy nor straightforward, and will involve significant experimentation and political maneuvering. However if we understand the rebalancing process in principle, we can at the very least distinguish between reform policies that are consistent with the necessary rebalancing process and policies that are not.
But there is a conceptual problem. The historical precedents suggest – with ample support from Chinese experience of the past 3-4 years – that like very small whales with very large blowholes the economists advising policymakers, with the agreement of outside analysts, will offer a profusion of reform proposals which confuse two very different sets of reform programs, largely because of the mistaken model used by most economists and to which I referred earlier on the section on Singer. One set consists of what I call “asset-side” policies that are designed to improve China’s economic efficiency, and these are the only reforms that are meaningful according to most consensus economic models. The other set consists of what I might call “liability-side” policies – although these involve more than the liability side – that address the rebalancing process directly, and while finance specialists, or economists who have been influenced by the balance sheet approach of people like Hyman Minsky, easily recognize these kinds of reforms, in general they are not well understood. I have discussed before, including most recently in the September 1 entry of this blog, the difference between the two kinds of policies.
This third point is important and probably explains a feature of history about which economists seem unfamiliar. There have been dozens of cases in the past couple of centuries in which sovereign debt levels were seen as being excessive. In some of these cases the debt was subsequently repudiated, but in most cases policymakers at first sought to reassure their creditors that the country was facing a short-term liquidity shortage, and promised to design and implement a package of policies that would improve the country’s economic efficiency, increase growth, and restore confidence. In some cases these promises were perhaps never credible, but in many cases they were, and the markets were prepared to give policymakers enough time for the measures to work and for the country to begin growing its way out of its debt burden.
Efficiency versus rebalancing
But of these dozens of cases, few, if any, sovereign borrowers were in fact able actually to grow their ways out of their debt burdens until, either explicitly or implicitly, the debt was substantially written down and assigned to an unwilling sector.[5] Until policymakers take on the debt burden directly, the historical precedents seem to tell us very firmly, sovereign borrowers have never been able to implement reforms that improve efficiency enough to allow them to grow out of their debt burdens.
I have explained elsewhere some of the reasons that determine whether a country’s debt is “excessively high”, and I hope formally to list these reasons more fully in my next book, but the key is the gap that is created between projected debt-servicing costs and the projected revenues earmarked to service the debt when an economic entity suffers an unexpected surge in debt or an unexpected decline in growth. When this happens and as debt levels rise relative to debt servicing capacity, at some point the major stakeholders — including businesses, creditors, household savers, workers and so on — became uncertain enough about how this gap will be allocated that they take steps to protect themselves from this uncertainty.
As this happens investors, recognizing that stakeholder actions are likely to undermine the economy further and worsen balance sheet fragility, express their worry about the risk of default in the form of high required yields. They also make it difficult for the sovereign borrower to issue new forms of debt. Finance specialists will recognize this condition as very similar to, but more general than, the trigger that sets of financial distress costs.
For most economists this seems implicitly to be a surprising statement: excessive debt is a balance sheet problem, and not an efficiency or productivity problem (although of course inefficient or unproductive economic activity is usually, along with badly-designed balance sheets, the main cause of excessive debt). That is why, because policymakers will rely on advice from economic advisors who are unlikely to understand balance sheet dynamics and who are almost completely unfamiliar with historical precedents, it was almost inevitable that at first Beijing would prioritize the wrong set of reforms aimed at raising the equilibrium growth rate of the economy, even sometimes at the expense of further balance sheet deterioration.
And they have. This process is the typical end of the expansion phase of every investment-driven growth “miracle” in history. In each case after many years of investment misallocation, both unexpectedly high debt and unexpectedly low growth create the gap between debt servicing costs and expected revenues, with each driven by and exacerbating the other (as has clearly been the case in China today).
As the gap widens, it creates rising uncertainty about how excess debt servicing costs will ultimately be allocated, and at the point at which this uncertainty is high enough to alter materially the behavior of economic agents, and so lower the net asset value of the economic entity, the borrowing country has “excessive” debt. Once it does, the process of deleveraging, like rebalancing, is inevitable, and it too can occur in many different ways, all of which involve forms of “debt forgiveness”, usually involuntary.
Some forms of debt forgiveness are explicit. The devastating LDC debt crisis of the 1980s, which began in August 1982 when the Mexican government announced that it was unable to service its obligations to foreign banks, ended only in 1990, when these loans were exchanged for a nominal amount of Brady bonds equal to only 65% of the original notional amount of outstanding loans. In the subsequent years, one after another of the indebted LDCs obtained notional debt forgiveness of 30-50% in the subsequent Brady-bond restructurings.
Partial debt forgiveness has been a formal part of nearly every sovereign default or debt restructuring in modern history, although usually not until there has been a long and painful period of angry posturing and one or more partial restructurings. During this time we often also see informal kinds of partial debt forgiveness, for example when sovereign borrowers have repurchased their obligations in the secondary market at steep discounts, often secretly, or exchanged their obligations for other assets at a discount, for example the famous debt/equity swaps in several Latin American countries in the 1980s (see footnote 3).
The ways of debt forgiveness
Most forms of “debt forgiveness”, however, are implicit, and nearly always involuntary. German’s excessive debt burden after the Great War, for example, was “forgiven”, unwillingly, mainly by middle- and upper-middle-class households and civil servants, whose fixed income portfolios withered to nothing in the hyperinflation that began in mid 1921 and ended in early 1924. China’s huge portfolio of NPLs at the end of the 1990s (perhaps as much as 40% of total loans) was resolved by a decade of severe financial repression, so that lending rates of around 7% – in an economy in which GDP grew nominally by 18-20% and the GDP deflator usually exceed 8% – implied substantial debt forgiveness.[6]
Because these loans were funded by even cheaper deposits, debt was forgiven at the expense again of household depositors in the banking system, and of course it is no coincidence that during this period the household income and household consumption shares of GDP, which began the decade at already very low levels, plunged to levels that are historically unprecedented. There are many other ways of allocating a significant portion of the debt-servicing cost to unwilling agents in the economic equivalent of debt forgiveness: to creditors when debt is repudiated, to workers when wages are suppressed in order to increase net revenues for debt servicing, to small business owners when assets are expropriated to pay down debt, and so on.
In the current global environment this problem, by the way, is not one just limited to China. For example the same thing is happening in many European countries which – for all the urgency some, like Spain under President Rajoy, have implemented efficiency-enhancing reforms – have been unable to grow their economies faster than the growth in their outstanding debt. I would argue that so far the policy advice Beijing has been receiving has also mistakenly assumed that China’s slowdown was caused mainly by various regulatory and institutional inefficiencies and rigidities, and that as these are removed by regulatory reforms, or countered by the appropriate fiscal and monetary policy, the Chinese economy will naturally move towards a stable equilibrium with much higher productivity levels. This they think will ensure sufficiently rapid growth and will give Beijing additional time in which to resolve its debt problems.
But this is not how the adjustments have ever worked and until now it does not seem to be working in China. A rapid slowdown in growth is imbedded in the adjustment process, and will inevitably occur with or without the proper reforms. For now the only way to keep growth from dropping sharply is to allow debt to rise as rapidly as is needed to meet growth targets, currently at least two to three times as fast as the growth in China’s debt-servicing capacity, but as debt rises growth will continue to decelerate more quickly than predicted.
How never to be wrong
In fact growth in China has already slowed sharply, and by far more than was permitted in most of the standard models, but bizarrely enough the rapid deceleration in growth has not caused economists to reject their models. Instead, with great nimbleness and intellectual flexibility, and often without ever missing a beat, they have simply applied the same framework to explaining lower growth. Like Kevin Keegan famously, they always know what’s around the corner but they never know where the corner is.
A typical case might be Stephen Grenville, a former Australian central banker, whose growth forecasts are regularly and quickly undermined by reported growth data with no appreciable effect on his analysis. It may be unfair to single him out, but I have become familiar with his work mainly because several of my investment and academic friends in Australia seem to delight in sending me his articles and making witty comments about how economists can take data that confounds their forecasts and use it to confirm their analysis. He has long been an optimist on Chinese growth and a little over two years ago he assured Australian investors that Chinese government debt was quite low, and that keeping it low was easily achievable. Now he accepts that it is high and poses risks, but it was apparently never likely to be otherwise.
He even cheekily wondered at the time about economists predicting a slowdown “How wrong do you have to be before you lose your expert status?”, he asked. The question is going to prove purely hypothetical, the investor from Sydney who sent me the article assured me, given that there is no amount of slowdown and no strain in credit that will do the trick for Grenville, but he was nonetheless fairly annoyed that a former Australian government official who claimed to understand what was happening in China would have done the unforgiveable and never warned that iron prices were about to drop from over $190 to test $50, even though that turned out to be a fairly easy prediction.
Perhaps he forgot. As an aside, one of my PhD students, Hao Yang, helpfully explained to me early last year that if I ever needed to write these kinds of academic papers he could do it for me because that is precisely what PhD programs train students to do. I suspected he was being a little cynical, but certainly it has been widely noted within the investment industry that while there have been profound disagreements during the past decade about the Chinese economy, and its performance seems to have thoroughly confounded the expectations and forecasts of majority opinion, so far it is hard to find economists who haven’t been proven right, and not just among Australians. Something so extraordinary almost certainly could not happen without a great deal of highly specialized training, so perhaps Hao Yang was not being cynical at all.
The most common argument used by analysts like Grenville to justify the recurrence of “unexpectedly” low growth while maintaining the validity of the conventional equilibrium models is usually that slower growth is a natural consequence of China’s rising capital intensity. As the level of investment increases, and so becomes less scarce, the return on capital naturally must drop, and it is this drop in the marginal return on capital that explains the deceleration in growth.
This explanation however verges on the nonsensical, and it is useful to consider why:
While it is true that the return on capital should decline as capital intensity rises, and that the capital intensive component of growth in China should be declining precisely because Chinese investment has surged, in fact surging investment is a three-decade-old story, whereas growth only began to decelerate rapidly over the past 3-4 years, just as China began to rebalance. This cannot just be coincidence.
The deceleration in Chinese growth moreover has been far too rapid to be explained by any normal decline in marginal returns on capital as investment rises, even if capital intensity were uniform throughout China, and in fact it isn’t. Capital intensity varies tremendously from province to province, so while investment saturation might conceivably explain growth deceleration in the most advanced provinces in China, it cannot do so to anywhere near the same extent in most provinces because of the tremendous variation in capital intensity across China’s 31 provinces and provincial-level entities. In fact a mathematician looking at a map of China listing provincial per capita investment rates would immediately see that such rapid deceleration, compressed over so short a time period, would require extraordinary mathematical convolutions to support an argument based on investment saturation.
Most damning of all, if growth deceleration were caused only, or mainly, by declining returns on capital as capital intensity rises, the most rapid deceleration would occur in the most advanced provinces, whereas there should be no slowdown in the least advanced. In fact across provinces the opposite is true.
Andrew Batson recently posted an interesting and related entry on his blog that is worth reading in full. Aside from investment saturation and unexplained invoking of the “middle-income trap”, I am not aware of any other plausible explanation for the sharp decline in reported GDP growth (and I ignore the very active debate about whether the real decline in economic growth is fully described by the reported decline in the GDP data) that might serve as an alternative to the model which posits sharp slowdown as the inevitable consequence of the rebalancing process.
The two important points if I am right, then, are, first, that during the adjustment a sharp slowdown in growth is inevitable, and is embedded within the rebalancing process, and second, that until the debt is specifically addressed, efficiency-improving reforms will never be enough to resolve the rebalancing process. Growth in other words will continue to decline substantially no matter what Beijing does.
By how much? I have argued since 2009 that the upper limit of average growth during the rebalancing period, which was always likely to be with the advent of the new leader, during the 2013-23 period, is likely to be 3-4% and that the longer Beijing succeeds in postponing the decline the greater the risk of disruptive “catching up” of the necessary deceleration in growth. It is not clear how quickly China has been growing in recent years because of the huge discrepancies between the reported growth data, which even Premier Li has questioned as being “for reference only”, and nearly every attempt by investment banks and independent economists to calculate growth independently. While it is hard to know what numbers to trust, most independent estimates already range from 1% to under 5%, but any higher growth rate over a longer rebalancing period is extremely unlikely and can only happen if implicit transfers from the state sector to the household sector very implausibly average more than 2-3% of GDP annually.
Do we care about the amount of economic activity?
Unfortunately the locus of the debate among those who recognize that the growth slowdown is not incidental and part of a “normal” deceleration of growth has shifted primarily to the accuracy of the published data. The Economist has typically been among those publications least convinced that China was facing a very difficult adjustment – in fact the FT’s Alphaville and the Economist have tended to bracket either side of the debate during the past few years – but in an article last week they worriedly note just how uncertain things have become:
Increases in indebtedness of that magnitude have been a forerunner of financial woes in other countries. Cracks are beginning to appear in China: capital outflows have surged, bankruptcies are occurring more frequently and bad loans in the banking sector are rising. It is all but certain that more pain lies ahead, though quite how much and how it will play out are matters for debate.
I think the cracks may have appeared a while ago, but in the same article they make reference to the debate about how accurate are the GDP growth data:
Judging by the eerie stability of key indicators recently, China’s statisticians appear to have been doing just that. In year-on-year terms, growth over the past six quarters has been 7.2%, 7.2%, 7%, 7%, 6.9% and 6.8%. Such a tight clustering is improbable.
I am not sure this is always as fruitful a debate as many seem to think it is. Of course it matters to anyone who wants to understand the economic cost of the adjustment, but arguments about whether the reported data are overstated, and by how much, have become part of the bull vs bear debate about whether Chinese growth is merely slowing temporarily, and not as part of a major economic reversal of the growth model. If you agree you are meant to accept the accuracy of the reported data. Otherwise you would question the data and assert that real GDP growth is substantially lower.
I don’t think this part of the discussion is especially useful. I have long argued that as long as China – or indeed any other country – has the debt capacity, it can get pretty much generate any amount of economic activity it wants. What is important is not how much growth there has been in economic activity, which is what the GDP numbers measure, but rather how much growth there has been in economic wealth, or in debt-servicing capacity, which is much the same thing, and how that compares with the growth in debt. In fact there probably hasn’t been much growth in the former, whatever the reported GDP data tell us, and there has been a lot in the latter.
Only two things matter
So what kinds of reforms are consistent with China’s rebalancing? Improvements in efficiency matter in the long term because as long as the economy does not suffer a major disruption they will tend to close the gap between the growth in economic activity and the growth in debt-servicing capacity, but they will have little effect on rebalancing. If the new set of reforms are to be truly effective, in other words, they must be designed directly to eliminate the balance sheet constraints on the economy. They must, in other words, accomplish the following:
One of the two goals must be to rebalance demand. The distribution of resources must be rebalanced in a way that it can generate debt-free demand for the economy. In principle one way to do so is to reform the financial sector so that it is able to identify productive investment opportunities and channel credit in that direction. This has been one of the most regularly proposed “solutions” available to Beijing.
In practice this almost certainly cannot happen. As I have discussed elsewhere, if we understand the political and institutional constraints that drive the evolution of a country’s banking system we would see why the necessary reforms are likely to be impossible to implement, and for those who are interested, my former Columbia University colleague, Charles Calomiris has written excellently on the subject. “A country does not choose its banking system,” he and his co-author Stephen Haber point out, “rather it gets a banking system consistent with the institutions that govern its distribution of political power.”
The historical lesson here is fairly unambiguous, although as always it is disappointing that economists who do propose such a solution for China evince so little curiosity about the historical precedents. It should be no surprise that many countries in the late stages of their own investment-growth “miracles” have tried this kind of transformation, but none has ever managed so radical a change within its financial sector quickly enough, at least in part because the capital allocation decision is at the heart of distributional politics. Because China begins the process with the highest investment level in history, the extent of the transformation must exceed that of any other case, and it must occur at a time when weak Chinese demand is compounded by weak global demand, thereby reducing productive investment opportunities for the private sector.
Another source of additional debt-free demand is, in principle, the external sector. China, however, is already challenging Europe as running the highest current account surplus in history, and in a world in which demand is likely to remain weak for many years, the external sector is unlikely to provide sufficient additional demand. Of course China can generate more demand by exporting more capital to the developing world, as it proposes to do with OBOR and the New Silk Road projects. It can also exploit its technology lead in high-speed rail, as a recent People’s Daily article on potential contracts for the high-speed Moscow-Kazan, Las Vegas-Los Angeles, Malaysia-Singapore, and Tanzania-Zambia lines. The total amount of development finance or rail exports it can provide, however, is tiny compared to domestic demand requirements, and if the recipients find themselves unable to repay the debt, as history suggests could easily be the case, this becomes a worse alternative to misallocating investment at home.[7]
The only certain and politically feasible source of debt-free demand is domestic household consumption, but Chinese households suffer from the same problem Marriner Eccles identified in the US in the 1930s: those who want to spend do not have the resources, and those who have the resources do not want to spend – or in this case are not able to spend productively. The solution is as obvious as it is politically challenging: China must redistribute resources from the latter, i.e. the state sector, to the former, i.e. Chinese households.
The other of the two goals must be to repair the balance sheet. Growth will not revive until the debt burden is sharply reduced. Debt can be reduced by partial debt forgiveness as part of a restructuring process following a default. It can be reduced as part of a pre-emptive restructuring. It can be reduced in the form of implicit debt forgiveness through monetization or financial repression. Or it can be paid down with funds generated from implicit or explicit taxes or from asset sales, including privatization. There are no other realistic ways to reduce debt.
Because most of China’s debt is internal debt, and directly or indirectly owed to the banks, debt restructuring with partial forgiveness is not an option at the macroeconomic level because ultimately it is a contingent liability of the government either way. But Beijing must resolve its debt burden at the national level or it risks repeating the mistakes of Japan –and Japan’s experience merely confirms what we already know: growth will not revive until the debt burden is sharply reduced.
China is also constrained from reducing the debt burden though monetization, financial repression, or taxes on households because in each case the cost is indirectly allocated to the household sector, which simply exacerbates the original imbalance. This leaves only two alternatives. First, Beijing can expropriate the wealth of small and medium enterprises directly or indirectly (in the latter case by raising taxes), although this means undermining the most productive part of the Chinese economy. Second, Beijing can liquidate government assets and use the proceeds to pay down debt. There are no other plausible options.
Multiple paths to the same outcome
One way or another China will adjust, and both of these objectives will be met. This will happen if for no other reason than because if something cannot go on forever, as Nixon’s CEA chairman Herbert Steinhelpfully reminded us, it will stop.
The logic of rebalancing is overwhelmingly corroborated, if it needed to be, by historical precedents. Every relevant country that has experienced Chinese-style growth has suffered in a similar way, and in every country the resolution has turned out to be the same, whether the resolution occurred automatically as a consequence of a financial crisis or occurred because of specific policies. But just because we can predict with total confidence that China will eventually rebalance and deleverage, it doesn’t mean that we can just as easily predict how this will occur.
There are several paths a country can follow once systemic distortions and imbalances have become deep enough. The path that China actually takes depends, of course, on the policies implemented by the government, the behavior and confidence of Chinese households, investors and businesses, and external conditions. For this reason the whole point of the reform process should be, first, to identify the distortions and imbalances that have become, or threaten to become, wealth destroying; second, to list the various plausible paths by which these distortions and imbalances will be reversed; third, to select the optimal path consistent with the country’s political, social and economic institutions; and finally, to design and implement the policies that move the economy along the least painful of the many paths.
I list what I think are only six plausible paths China can follow in Chapter 5 of my 2013 book, Avoiding the Fall, along with the associated conditions for each of these paths. Each of the various paths will take the economic system back into some kind of balance from which policymakers can expect further sustainable growth, but these various paths have very different impacts on wealth and stability.
Take rebalancing. I have already listed above some of the many ways deleveraging can take place, from default to buy backs to financial repression to the sale of assets to pay down debt, and likewise there are several ways the rebalancing of demand can take place.
Both the US in the late 1920s and Japan in the late 1980s had deeply unbalanced economies and excessively high savings rates, the consequence of which were huge current account surpluses, along with what may have been the highest and the third highest hoard of foreign currency reserves, respectively, in history (China today probably ranks second), and highly inflated domestic asset markets. The causes of their imbalances were, at least in part, high levels of income inequality and relatively low household income shares of GDP.
Both countries rebalanced in the subsequent decade, as they inevitably had to, and in both cases the savings share of GDP declined (or the consumption share increased, which is the same thing), but it declined for very different reasons. In the US, the rebalancing occurred mainly in the form of a collapse in GDP relative to household income, with the former dropping by around 35% between 1930 and 1933 and the latter dropping by “only” half that rate. This was accompanied by substantial income redistribution, much of it occurring partly because of redistributive policies under Roosevelt and mostly because of a wave of sovereign and domestic bond defaults whose losses were borne mainly by the high-saving wealthy. Japan’s rebalancing, on the other hand, occurred in the form of two decades in which GDP growth barely exceeded 0%, while the growth in household income and consumption averaged more than 1%.
The state must pay
It is worth noting, as we think about China’s options, that historically a sharp, economically disruptive rebalancing with negative consumption growth and even more negative GDP growth, as experienced by the US, and a long period of stagnation with low consumption growth and even lower GDP growth, as happened in Japan, represent the two main ways that significant savings imbalances have tended in the past to adjust. In principle you can also have moderate GDP growth and high consumption growth, but this has never happened in history, probably for obvious reasons, and the extraordinary faith many analysts have that this is the most likely outcome unless Beijing seriously mismanages the process is almost certainly wholly misplaced.
There seem to be four main mechanisms responsible for major incidences of income redistribution. The first is a politically-driven redistribution of wealth from rich to poor (sometimes disruptively, in politically unstable states, and sometimes not). The second is very high inflation or financial repression that erodes bond and bank-deposit values. The third is a wave of sovereign and domestic bond defaults. The fourth is war, although perhaps this occurs mainly because war is often inflationary.
However they occur, the many historical precedents, reinforced by logic, suggest very strongly that no reform program will be viable in China unless it is consistent with a rebalancing of demand from investment to consumption and with a reining in of credit expansion and eventual reduction in the country’s debt burden. Because the only plausible way of rebalancing demand requires that Beijing directly or indirectly redistributes resources from the state sector to the household sector, while the most efficient way to reduce the debt burden involves liquidating state assets and using the proceeds to pay down debt, it turns out that by far the most efficient and sustainable program of reform requires the wholesale liquidation of state assets to fund transfers of wealth.
And there’s the rub. While a reform program that liquidates state assets to pay down debt and to rebalance the household income share is economically the least disruptive kind of reform program, and is the one most likely to leave the economy in a position to resume rapid growth over the long term, it is unfortunately likely to be fiercely resisted by the many powerful vested interest who benefit from state ownership and control of assets. This is the big challenge faced by the Xi administration, and it has been the challenge faced by nearly every country that has undergone a similar type of growth “miracle”.
Why do we need a new program?
I have taken a very long digression before actually beginning my discussion of successful and unsuccessful reforms, among both the previous set of reform proposals and the newly proposed “supply side” reforms, because once we understand what has worked in the past, and what has never worked, it becomes much easier to evaluate the specific reforms that might work for China and those reforms that are clearly irrelevant. In fact they become almost obvious.
There are two main criteria by which to judge the usefulness of specific reforms. First, because the consumption share of GDP will rise no matter which of the rebalancing paths China takes, policies that maintain or even increase the growth rate in household consumption, mainly by maintaining or increasing the growth rate of household income, will push China along a “better” adjustment path of more rapid growth.
Second, because China will deleverage one way or the other, and because financial distress costs create a powerful inverted relationship between the size of the debt burden and the pace of economic growth, policies that reduce financial distress costs, by far the most important being those that pay down debt, will push China along a “better” adjustment path of more rapid growth. These should be the two main considerations when evaluating reform proposals, and while policies that accomplish neither may in fact benefit China’s economic efficiency in the long run, they will not protect China from a brutal and potentially disruptive adjustment.
Beijing formally promised to begin rebalancing the economy in a famous speech by Wen Jiaobao in March, 2007, although it was only in 2012 that the consumption share of GDP stopped declining and began to rise. Since then, the consumption share of GDP has expanded by perhaps one-fifth of the amount by which its GDP share would eventually have had to expand, and debt continues to rise at least twice as quickly as debt-servicing capacity, perhaps much more. Clearly Beijing plans to design and put into place what are being called “supply-side” policies mainly in response to the difficulties it has encountered so far.
I think it is safe to say that implicit in the new set of policies is the recognition that in recent years Beijing has failed to rebalance the economy by nearly as much as it should have, and that it has not made enough progress in implementing the Third Plenum reform proposals. These reforms, if implemented robustly, would have restructured the country’s economic institutions so as to consolidate the progress of the past three decades and make long-term growth sustainable, but this has turned out to be extremely difficult, largely, I suspect, because of what to many was an unexpectedly strong political opposition. And yet this opposition was very much one of the characteristic outcomes of the rebalancing process.
There a