Monetary policy under financial repression

By Michael Pettis

Following Paul Krugman’s lead I guess I can refer to this post as being “wonkish”. Much of it is based on my recent book Avoiding the Fall (Carnegie Endowment, September 2013).

In order to understand much of what is happening in China I believe it is crucially important to understand how financial systems operate under condition of financial repression. Because most of what we know about economics is derived from economists whose operating environment is the classical “anglo-saxon” economies (I stress “classical” because for much of the 19th Century, operating under the so-called “American System”, the US itself was not, in my opinion, a classic anglo-saxon economy), there is a tendency to assume that what happens in those economies is somehow the default position in economics, and this not only causes us to underrate important economists that don’t follow this tradition, like the German Freidrich List or the American Albert O. Hirschman, but it also leads us into mistaken assumptions, like the belief that higher interest rates lead automatically to higher savings rates.

We do know some things about financial repression. Two of the first important texts to discuss financial repression comprehensively are Edward S. Shaw, Financial Deepening in Economic Development and Ronald I. McKinnon, Money and Capital in Economic Development. There is, however, a lot more to it than what is generally known, and even this is largely ignored by most economists. It seems to me that many of the mistakes we make when we think about the relationship between cause and effect, for example the impact of monetary policy on China’s economy, arise because we assume that relationships that hold in the US economy are universal and must hold in the Chinese economy too. So to return to the assumption that higher interest rates must lead to higher savings rates, I would argue that this is true mainly under two unstated assumptions, neither of which holds for China.

First, higher interest rates must must have a negative wealth effect, which they mostly do in the US. In China however the wealth effect is positive. Second, changes in interest rates should have a minimal impact on the household share of GDP. In the US, where there is a wide variety of alternative investments and where interest rates move broadly in line with changes in inflation, this may be true, but in China where most households have few alternatives to bank deposits, and where interest rates are set independently of changes in inflation, this isn’t true.

In fact distortions in domestic interest rates may be the single most important explanation of why the household share of GDP has plummeted in China, especially over the decade ending around 2010-11. So while higher interest rates in the US are typically (although certainly not always) associated with increases in the savings rate, in China they are typically associated with reductions in the national savings rate. There should be nothing mysterious about these opposite reactions to higher interest rates – both are fully explainable with our current economic tools as long as we are clear about the assumptions, often hidden, that we  make. And notice that I refer to “national” savings rates, not “household” savings rates, which are commonly confused because in the anglo-saxon economies changes in the national savings rates closely follow changes in the household savings rate, whereas in China they do not.

To understand the Chinese economy we must understand how financial repression changes the relationships between variables, many of which we implicitly and incorrectly assume are fixed and permanent. Financial repression, in other words, is not only at the heart of both China’s rapid growth and China’s economic imbalances, but it also explains a number of otherwise puzzling aspects of the Chinese development model. A repressed financial system will seem to operate in a fundamentally different way than a market-based (“anglo-saxon”) financial system, but in fact the principles under which it operates can be explained using what we already know about the operations of monetary policy in a market-based financial system.

One of the apparent puzzles about China’s growth trajectory, especially in the past decade, is the seeming disconnect between rapid monetary growth and relatively stable domestic inflation. It is well known in economic theory that countries that have open capital accounts are forced to choose between managing domestic monetary policy and managing the currency regime.  When a central bank chooses to intervene in the currency to maintain a desired exchange level, the amount of money it creates domestically is largely a function of the need to monetize net inflows or outflows. On the other hand if it chooses to manage the domestic money supply, the supply and demand for that currency in the international markets will determine the value of the currency.

In China’s case the capital account is technically closed, so in principle Beijing should be able to manage both the value of the currency and the amount of domestic liquidity. In reality, however, there are two significant limits to the country’s ability to maintain closed capital accounts. First, and most obviously, the capital account is the obverse of the current account, and any country with the volume of exports and imports that China runs necessarily will see significant activity in the capital account, especially if, as widely believed, Chinese nationals evade capital controls by over- and under-invoicing exports and imports. What is more, although much of the trade-related capital inflow and outflow is controlled by the central bank, an increasing share of capital flows occurs outside the central bank.

Second, China has extensive trading borders, a great deal of local corruption, and a long history both of capital control and capital control evasion.  Throughout history countries with large trading borders, a long history of capital controls, and wide-scale corruption have rarely been able to control capital flows as these factors undermine the ability of financial authorities to manage them, and China is not an exception. In fact during the past decade by most accounts China has experienced significant amounts of both speculative inflows and capital flight, measuring probably in the hundreds of billions of dollars, neither of which is compatible with strict enforcement of capital controls.

For all practical purposes, in other words, and in spite of formal capital flow restrictions, China is also forced to a greater or lesser extent to choose between managing its currency regime and managing domestic money creation. Clearly it has chosen to manage the currency regime, and the enormous changes in central bank reserves, which at over $3 trillion are the largest hoard of central bank reserves ever amassed by a single county, are a testament to that.

Monetary expansion and inflation

Monetary policy from the point of view of the balance of payments is pretty clearly a consequence of the central bank’s need to monetize an enormous amount of net inflows. China’s current account surplus began surging around 2003-04 to levels that are almost unprecedented in history, with the country at its peak running a current account surplus of up to 10 per cent of China’s GDP, giving it, with the surplus equal to just over 1 percent of global GDP, one of the highest current account surpluses as a share of global GDP ever recorded.

The impact of the current account surplus on capital flows tended to reinforce monetary creation in at least two ways. As money poured into the country as a consequence both of its current account surplus and its net surplus on the capital account (among other things China is been the largest recipient of foreign direct investment in the world), it helped ignite a credit-fueled asset boom, especially in the real estate sector, that encouraged additional speculative inflows looking to take advantage of soaring prices.

In addition the massive current account surplus fueled speculation about the trajectory of the renminbi.  As investors expected the value of the renminbi to rise as it adjusted to current account inflows, even more speculative inflows poured into the country seeking to benefit from any appreciation. The result was that until late 2011 substantial net capital inflows, added to the already very high current account surplus, drove up central bank purchases to extraordinary levels. 

As a share of global GDP the only comparable hoard of foreign currency reserves occurred in the United Sates in the late 1920s, a period distorted by the destruction of much of Europe’s manufacturing capacity in World War 1 and by the impact political uncertainty in Europe had in driving capital to the relative safety of the United States. During this time, when the US experienced both massive current account surpluses as well as massive private capital account surpluses that generated its huge central-bank reserve hoard, the US share of global GDP was roughly three to four times the current Chinese share, which gives a sense of just how extraordinary the Chinese accumulation of reserves has been.

With so much money pouring into the country, the People’s Bank of China was forced regularly to monetize an amount equal to a substantial share of its existing money base. Normally central banks would try to sterilize this money creation, and the People’s Bank of China did try to mop it up, but most measures of money nonetheless continued to increase rapidly, and there is anyway a real question about the effectiveness of sterilization with highly liquid and credible instruments that are already a close substitute for money. The tools used to sterilize inflows, mainly short term bills issued by the central bank, are themselves forms of money, and the more extensively they are employed, the more liquid they become and hence the more “money-like.”

The alternative to a real and effective sterilization is for the Chinese economy to adjust in the form of a surge in inflation. As the money supply grows in response to China’s current account surplus and net capital inflows, it should cause prices and wages to surge, forcing a real appreciation in the currency, until both China’s current account surplus and net capital account inflows wither away. 

This is of course the classic currency adjustment mechanism under the gold standard. As reserves soared in China, money creation soared along with it. Rapid money creation should have resulted in a rapid rise in domestic wages and prices as demand for goods and services outstripped supply. Rising domestic prices should have in turn undermined Chinese exports, encouraged imports, and reversed capital inflows.

But this didn’t happen. In fact during the past decade, price inflation in goods and services in China has been fairly moderate, and usually driven exogenously (crop failures, high commodity prices, etc.) and wages actually grew more slowly than productivity. China’s export competitiveness not only was not eroded by domestic money creation, as it would have been under the classical adjustment mechanism, but it also had, by some measures, even increased during this period. There have been periods during which inflation seemed about to take off, but these periods tended to be short-lived and were always followed by sharp declines in inflation. 

At first this might seem to imply that sterilization was indeed effective in preventing money creation in China from getting out of hand. By selling central bank bills, transacting in the repo market, and raising minimum reserve requirements aggressively (to around 20 percent, compared to the 5-10 percent that is more common in developing countries), the central bank seems to have been successful in mopping up the money created by the monetization of current and capital account in flows and so protecting the Chinese economy from the normal consequence of maintaining an undervalued currency.

What happened, however, in fact was very different.  Generally speaking, there have been a number of countries besides China that have managed for long periods to combine tremendous capital and current account inflows, rapid growth in foreign currency reserves, and low inflation – for example Japan in the 1980s.  In nearly every case these countries also had severely repressed financial systems.

What’s more, although it was hard to find in China and other similar countries the normal evidence of rapid money creation in changes in consumer prices, other parts of the economy acted in ways that seemed consistent with rapid money creation. Credit, both inside and outside the formal banking system, grew astonishingly quickly and, as usually occurs under conditions of too-rapid credit growth, credit standards deteriorated. The stock and real estate markets experienced bubble-like behavior. Producer prices rose rapidly. Global commodity prices, spurred largely by soaring Chinese demand, also soared.

Bifurcated Monetary Expansion

So was Chinese monetary expansion excessive or not, or to put it differently, why is it that what seemed by most measures to be an extraordinary surge in money creation did not also result in significant wage and consumer price inflation? The answer, I will argue, has to do with the nature of money growth in financially repressed economies. Because the Chinese financial system is so severely repressed, money growth in China cannot be compared to money growth in a market-based financial system. Monetary growth is effectively bifurcated and affects producers and consumers in very different ways.

What does it mean to say that monetary growth was bifurcated? By this all I mean is that nominal money growth showed up as different rates of money growth for different parts of the economy. More specifically the rate of monetary growth for producers exceeded the rate of monetary growth for consumers, and this becomes clear by measuring the monetary impact on different sectors within the economy of monetary expansion under financial repression.

Countries with significant financial repression can experience periods of rapid monetary expansion with results that do not conform to normal expectations precisely because of this bifurcation in the monetary impact of credit creation.  On the production side of the economy it is easy to see in China over the past decade what looked like the consequence of rapid monetary expansion – rapid growth in credit, rising productive capacity, surging production of manufacturing goods], asset bubbles, etc.

On the demand side of the economy, however, and especially considering household consumption, one gets a very different view – monetary expansion seemed to have been very subdued. Household consumption typically grew much more slowly than GDP and its share of GDP declined steadily.  Consumer price inflation also tended to be low or moderate even in the face of what seemed like rapid monetary expansion.

So had there been too-rapid monetary expansion in China during the past decade or not?  Why do some sectors seem to indicate that there has been, and other sectors that there hasn’t?  The answer depends, it turns out, on which economic sector we examine, and whether that sector was a net borrower or a net lender.  We will see that financial repression can create a bifurcation in monetary expansion when

a)    net savers and net borrowers are two very distinct groups, in this case the former being households and the latter being producers of goods and infrastructure, including manufacturers, governments, real estate developers and infrastructure investors;

b)    the bulk of savings consists of deposits in the banking system and the bulk of corporate financing consists of bank lending or other forms of bank financing.

The experience of China (and other financially repressed economies) suggests out that when interest rates are set artificially low in such a financial system, any given nominal expansion in money supply creates a lower real expansion in money on the consumption side and a higher real expansion in money on the production side. The consequence may be rapid GDP growth, a surge in investment and low inflation for many years, but it also leads to sharply unbalanced growth in which the role of domestic demand as a driver of growth shrinks.

To see why, assume a country in which the “natural” nominal interest rate is 5% for all maturities.  For the sake of simplicity we will assume that deposit and lending rates are the same, and that the marginal reserve requirement is constant, although these assumptions do not affect our final conclusions in any significant way.

Now let us assume that there are immediately two transactions.  First, a saver deposits $100 dollars in the bank for one year at 5 percent and the $100 dollars are immediately lent out to a borrower for one year at 5 percent.  One year from now the borrower will repay $105 and the saver will receive $105.

Second, we assume that another saver deposits $97 for one year at 5% and the money is immediately lent out to a borrower for one year at 5%.  For the sake of simplicity we will round off the pennies and assume that in the second case the borrower repays one year later and the depositor receives one year later $102.

It is clear that the because of the first transaction the money supply has increased by $100, and the depositor will receive and the borrower will repay $105 in one year.  It is also clear that because of the second transaction the money supply has increased by $97, and the depositor will receive and the borrower will repay $102 in one year.

But now let us posit that the central bank decides suddenly and arbitrarily to reduce both the lending and deposit rate to 2%.  This has nothing to do with a change in inflationary expectations or the real demand for money – it is simply driven by other domestic considerations.

Following the decision a third, less fortunate saver decides to deposit $100 for one year at 2% and this $100 is immediately lent to a lucky borrower for one year at 2%.  Which of the first two transactions is closer in its monetary impact to the third transaction?

From the depositor’s point of view the present value of $102 one year from now is only $97 (for simplicity I am rounding off adjustments to the nearest dollar).  Although the nominal amount of his deposit is $100, just like that of the first depositor, the real value of his deposit is really only $97, just like that of the second depositor. If we define money so as to include deposits, did the money supply rise by $97 or $100? 

On a comparable basis it is pretty clear that the third depositor’s position, after interest rates were artificially lowered, is no different than that of the second depositor who deposited $97. Nominally the value of his deposit is the same as that of the first depositor, or $100, but his wealth is the same as that of the second depositor, or $97. Since it is real wealth, and not nominal deposits, that ultimately matters to the depositor, and which will affect his consumption and savings decisions, the third depositor is likely to behave over the long run as if he were in the position of the second depositor.  

Because in this case a $100 deposit results in a $97 increase in the real value of deposits, in other words, it turns out that the nominal growth in money as measured by deposits overstates the real growth. Under financial repression a $100 transfer from the household to the bank in the form of a $100 bank deposit results in a smaller real deposit than under conditions of no financial repression.

Transfers change the monetary impact

If financial repression distorts the balance sheet of the depositor, what does it do to the balance sheet of the borrower?  For the third borrower, who in our example borrowed under conditions of repressed interest rates, the transaction is the mirror opposite of the depositor’s transaction.  The third depositor effectively had $3 “confiscated” from his assets in the form of an arbitrary reduction in the deposit rate. This $3 is transferred automatically to the borrower, so that the third borrower’s liability more closely resembles that of the second bower, even though he receives upfront the same $100 that the first borrower receives. 

The nominal increase in money as measured by loans, in other words, understates the real increase. The third borrower receives both the $100 loan as well as a $3 “gift’ in the form of partial forgiveness of his debt.  His purchasing power has gone up not by $100 but rather by $103, even as the purchasing power of the third depositor has only gone up by $97.

Depositors in a financially repressed system may make the same initial deposits as depositors in a non-financially repressed system, and borrowers in a financially repressed system may receive the same initial disbursements as borrowers in a non-financially repressed system, but their resulting balance sheets are very different.  Wealth is effectively transferred from the depositor to the borrower under financial repression and so the purchasing power of the former is reduced relative to the nominal size of the deposit while the purchasing power of the latter is increased relative to the nominal size of the loan.

This transfer modifies the monetary impact on each of them and the effect is cumulative. Assume in the above example that the money supply consists entirely of $100 nominal of one-year deposits matched with $100 nominal of one-year loans.  If in any given year the money supply (loans and deposits) is increased by $20, or 20%, the impact on deposits and loans is very different. In effect the real value of deposits will have risen that year by only $2 (with $18 effectively transferred to borrowers), whereas the value of loans will have increased by $38. An increase in nominal money of 20% in other words, is associated with a 2% real increase in deposits and a 38% real increase in loans.

This is what it means to say that financial repression creates a bifurcation of monetary growth.  For households, and net depositors more generally, real monetary expansion is in effect much lower than nominal monetary expansion because of the implicit financial repression “tax”, and so consumption growth and consumer-price inflation will seem abnormally low.  For manufacturers, real estate developers, infrastructure investors and other net borrowers, real monetary expansion is in effect much greater than nominal monetary expansion because of an implicit financial repression “subsidy”, and so asset inflation and capacity growth seem abnormally high.

Perhaps one way of thinking about it is to consider how to make a comparable impact in a market system.  Imagine if somehow the US were to enact a law whose result was that every time the Fed expanded the money supply, a one-off tax was imposed on households, the proceeds of which were transferred to corporate borrowers. In that case monetary expansion would be much less likely to cause an increase in demand for consumer products, and so would create much less consumer price inflation, and much more likely to cause a surge in production.

This effective “tax” suggests that in a financially repressed system, it is normal that the impact of nominal monetary expansion will seem much greater in one sector of the economy than in another, with the differencing reflecting the net lending or net borrowing position of that sector. The impact of monetary expansion on the behavior of the saver is much lower than it is in a market-based financial system, all other things being the same. The impact of monetary expansion on the behavior of the borrower is much higher than it is in a market-based financial system, all other things being the same.

Under these conditions it is consequently not surprising that the economy can seem to be operating under conflicting monetary systems.  Consumption behaves as it would in an economy with much lower monetary growth, and production and asset prices behave as they would in an economy with much higher monetary growth.

I will leave it to an ambitious doctoral student to work out the full monetary implications of financial repression and to formalize a model of monetary growth under financial repression, but it is worth noting that there are several other real implications of this bifurcation in monetary policy, all of which seem to apply to the Chinese economy:

a)    Financial repression creates in effect a two-speed economy. There will normally be a growing imbalance between the net saving and net borrowing sides of the economy, and the latter should grow much more quickly than the former. 

b)    By subsidizing the production side of the economy and penalizing the consumption side of the economy, financial repression must always force up the domestic savings rate. This may seem at first counterintuitive because, as I discussed at the beginning of this entry, we normally associated lower interest rates with lower savings, but it is an automatic consequence of the very different wealth effect that changes in interest rates have on market-based financial systems and financially repressed financial systems. Savings, after all, are simply the difference between consumption and production, and any process that forces production to grow more quickly than consumption automatically forces up the savings rate.

Financially repressed systems with artificially low interest rates tend historically to have much higher national savings rates than market systems, and also much higher savings rates than financial systems in which interest rates are abnormally high, but oddly enough the higher savings rates are almost always ascribed to cultural preferences. Rather than explain differential savings rates by cultural factors, it seems far more promising to explain them as consequences of financial repression.

c)    To rebalance the two sides of the economy either policymakers must eliminate, or even reverse, the transfer created by financial repression (i.e. either nominal interest rates must rise or GDP growth must drop) or they must implement another mechanism that directly transfers wealth from net borrowers to net lenders.

d)    The more interest rates are repressed, the harder it is for consumption growth to keep up with production growth because monetary policy driving consumption is effectively much “tighter” than monetary policy driving production.

e)    Consumer price inflation is not the appropriate measure by which to gauge domestic monetary conditions.

f)      Hikes and reductions in interest rates are not expansionary or contractionary in the way we might expect in an open financial system.  A hike in interest rates may act to contract investment, but contrary to conventional wisdom it actually expands consumption because it reduces the wealth transfer from the saver to the borrower.  This allows the saver to increase consumption.

g)    For the same reason consumer price inflation in a financially repressed system can be self-correcting.  If inflation rises, but interest rates do not, the bifurcation of monetary growth will increase because the difference between the “correct” interest rate and the nominal rate increases.  In that case any given nominal monetary expansion is accompanied by an even lower (or negative) real expansion from the point of view of consumers as net savers. By lowering the real cost of credit for borrowers, it can expand production.  Increasing production while reducing consumption, of course, outs downward pressure on prices.

h)    Monetary expansion accelerates investment and asset price inflation.  If inflation rises, but interest rates do not, any given nominal monetary expansion is accompanied by an even greater real expansion for net borrowers.

i)      This may be why in financially repressed economies regulators often resort to formal or informal loan quotas.  Without loan quotas, monetary expansion for borrowers may far exceed the needs of the economy, even as monetary expansion for depositors is too tight.

j)      As long as the rest of the world can accommodate the consequent excess of production over consumption, the bifurcation in monetary policy will not seem to be a problem, but once the world cannot accommodate it, the bifurcation of monetary expansion will create deflationary pressures.

k)    As long as the rapid increase in monetary expansion for borrowers does not result in a misallocation of capital, the bifurcation in monetary policy will not seem to be a problem, but once rapid money expansion leads to increasingly wasted investment, as it eventually always must, the bifurcation of monetary expansion will create asset inflation and an unsustainable increase in debt (as debt rises faster than debt servicing capacity).

l)      Deflation or disinflation partially or wholly resolves the bifurcation by forcing real interest rates towards their “correct” level (because real deposit and lending rates rise in a deflationary environment in there is no change in the nominal interest rate).   Under deflation we would expect to see the gap between consumption growth and GDP growth narrow, or even reverse.

m)  Slower GDP growth partially or wholly resolves the bifurcation by forcing real interest rates towards their “correct” level (in a market system nominal interest rates move naturally in line with nominal GDP growth). Under conditions of much slower GDP growth we would expect to see the gap between consumption growth and GDP growth narrow, or even reverse. This, for example, is what happened in Japan after 1990.

n)    Disintermediation of the banking system, to the extent that it reduces the impact of financial repression, may create an unexpected burst in consumer price inflation. This is less true to the extent that, like in China, disintermediation is limited to the rich, since the consumption impact of higher income on the rich is limited.

Asset price inflation

To summarize, in a financially repressed system in which consumers tend to be net savers and producers net borrowers, consumers and producers experience very different monetary impacts of the same underlying monetary conditions. The latter exist in an environment in which the impact of monetary growth is much faster than do the former.

There are two problems, then, which must arise when a financial repressed system experiences long periods of rapid money growth. First, as growth in production systematically exceeds growth in consumption, absent exponential growth in investment a growing trade surplus is necessary to resolve the growing imbalance.  Once there are constraints on the ability of the trade surplus to continue growing – for example the global financial crisis has caused a collapse in the ability of the rest of the world to absorb China’s rising trade surplus – the only way to prevent a collapse in growth is to increase investment even more.

But if investment is being misallocated this simply exacerbates the second problem.  Rapid monetary expansion, exacerbated by the bifurcation created by financial repression, has a tendency to result in capital misallocation and asset price inflation because it accelerates monetary growth.  If the response by policymakers to a contraction in the world’s ability to absorb rising trade surpluses is to engineer a further increase in investment, we would expect debt to surge, more investment to be wasted, and for the debt consequently to become unsustainable much more quickly.

This seems to be exactly what happened in China during the 2008-09 global crisis.  Before the crisis, debt was already rising at an unsustainable pace thanks to many years of the combination of rapid monetary growth and monetary policy bifurcation.  China’s trade surplus also soared as production was forced to rise much more quickly than consumption.

The crisis caused a collapse in China’s trade surplus.  In order to limit the impact on Chinese growth, Beijing engineered an extraordinary increase in domestic investment.  What is more, Beijing increased both the total amount of loans and deposits outstanding while lowering even further the real interest rate.  The net impact was to increase the financial repression tax on households – a tax which when directly to subsidizing borrowers. 

This certainly resolved the problem of a sharp decline in growth caused by a collapse in the trade surplus, but it did so by exacerbating the investment bubble and accelerating the rate at which the growth in debt exceeded the growth in debt servicing capacity.  It also worsened the consumption imbalance. It is probably not a coincidence that it was only in 2010 that most analysts belatedly recognized the problem of soaring debt in China – probably at the instigation of a report by Victor Shih, then a professor of political economy at Northwestern University and until then one of only a handful of China skeptics, on the surge in local and municipal debt.

As Chinese growth rates stayed high even in the midst of the worst global economy in 70 years – a fact that was a necessary consequence of the combination of increasing financial repression and a surge in monetary growth – there were always likely to be two factors that would undermine growth.  First, if the pace of monetary expansion slowed, and second, if the financial repression tax declined. 

What does it mean to say that the financial repression tax declines?  This doesn’t simply mean that interest rates rise, but rather that interest rates rise relative to GDP growth.  In a market-based system, over long periods of time nominal interest rates are broadly in line with nominal GDP growth rates. This means that savers and borrowers fairly distribute the returns on growth in proportion to the amount of risk they take. Of course if interest rates are artificially low, savers receive a disproportionately lower share and investors a disproportionately higher share of the benefits of growth.

The greater the difference between nominal lending rates and the nominal GDP growth rate the greater the financial repression tax. In China during the first decade of this century nominal GDP growth rates have been 16-20%, depending on which period you measure and what assumptions you make about GDP growth, while nominal interest rates have been roughly 6-7%.  This gives some idea of the extent of the financial repression tax, although even this understates its extent because the spread between the lending rate and the deposit rate is set artificially high, thus lowering even more the returns to depositors (an additional tax is effectively levied on depositors to recapitalize the banks).

The important point is that beginning sometime in late 2011, both conditions were in place.  As debt continued to rise in China and as slowing growth eroded China’s trade surplus, there is evidence that beginning in 2010 capital flight from China began to surge, while beginning some time in the fourth quarter of 2011 speculative inflows into the renminbi began drying up. The combination turned China’s position from running net capital inflows to running net capital outflows (excluding changes in central bank reserves, which by definition balance out total flows to zero).

As a result, in late 2011 and 2012 we witnessed for the first time China’s reserves rise by less than the already-much-lower current account surplus. By the middle of the year, net capital outflows actually exceeded the current account surplus (reserves, in other words, declined in spite of a current account surplus).

As Chinese money creation slowed, exacerbated by monetary bifurcation, Chinese growth slowed along with it.  This had the impact of reducing the financial repression tax (the difference between nominal GDP growth and nominal interest rates narrowed).  The consequence was predictable.  GDP growth slowed far more quickly in 2012 than even the pessimists had expected.

This is part of the self-reinforcing tendencies that financial repression creates for an economy. Rapid growth increases the financial repression tax, which tends to create even more rapid growth by reducing the real cost of capital. Slowing growth reduces the financial repression tax, which slows growth even further. These self-reinforcing tendencies imbedded in the national capital structure are typical of developing countries and one of their great sources of economic volatility – one that tends to undermine long-term growth.