China: I Still Think its Money, Not Pork; Credit is Contracting Very Rapidly

In all the worry about the trade numbers I haven’t discussed another data release last week which, until recently, would have been the most important piece of economic news for me. It was only a few months ago that we were intensely debating the cause of rising inflation in China. Now, inflation is clearly receding. PPI and CPI figures for November were released Wednesday and Thursday, and they showed an unexpectedly high decrease in inflation, with month on month numbers implying actual deflation. Year on year PPI prices rose by 2.0%, compared to 6.6% in October and around 4.5% expectations. CPI inflation declined from 4.0% in October to 2.4% in November, also well below the 3.0%expectations.

What does this say about monetary conditions? First of all let me take a radically contrarian position. I think the latest numbers, far from confirming the view that Chinese inflation in the period until this summer was caused by food supply constraints, actually indicate that the money explanation was right. Inflation in China earlier this year, in other words, was caused by too much money, not too little pork, in Ken Rogoff’s famous formulation.

For those of us in the money camp, inflation was the all-too-expected reaction to explosive money creation caused by China’s currency regime, in which the country’s central bank purchased a torrent of inflowing dollars in order to maintain the value of the currency. In order to fund the purchase of all these dollars, it created RMB, which was intermediated by the formal and informal banking systems into rapid credit growth.

For the pork camp, however, this monetary explanation of inflation was wrong. The cause of Chinese inflation, they argued, was temporary food supply constraints, including harsh winter storms and blue ear disease, which seriously affected pork and grain production. These caused food prices to surge. Inflation, in their view, was nothing more than the temporary consequence of rising food prices, and would end as soon as the supply constraints ended.

While we in the money camp of course acknowledged the sharp rise in Chinese food prices because of harvest problems and pig diseases, we felt that this was as much a coincidence as anything else (especially since food prices were rising around the world). I believed that if monetary policy in China were consistent with low inflation, the brutal surge in food prices would have caused Chinese households to divert so much spending away from non-food items that there would be significant downward pressure on non-food prices. I calculated that non-food prices should have declined by 5-7% if Chinese monetary conditions really were consistent with the 1-3% inflation targeted by the PBoC.

But non-food inflation was not negative. It was low, but we were actually seeing rising non-food inflation for most of the year while food prices were soaring. There was not even disinflation in non-food items, which is the least I would have expected if Chinese inflation was not money-based. That is why I rejected the food-supply constraint argument.

Of course based on our very different analyses, each camp had different predictions for the inflation trajectory. The money camp claimed the still-rapidly growing central bank reserves would ensure that inflation continued through the end of 2008 and into 2009. The pork camp argued that as food supply came back onto line by the end of the year, the food supply constraint would ease and inflation would quickly come down to manageable levels.

As we all know, inflation stayed high through the summer but subsided very rapidly thereafter as food prices declined sharply. In fact they declined far more rapidly than anyone predicted. Year-on-year CPI fell to 2.4% in November and year on year PPI fell even more to 2.0%. So the pork camp’s analysis must have been correct, right?

Not quite. The first clue should be the speed at which inflation fell. Last summer even the most hard-core members of the pork camp didn’t believe that the official target of 4.9% inflation for 2009 would be met. I, frankly, wrote it off as silly even to pretend it was possible (note to self: nothing is impossible). The pork camp made their food price projections based on the removal of supply constraints, and the optimists among them were arguing that we might see inflation between 5% and 6% for this year and a little lower next year.

In fact although food prices declined in line with their predictions, CPI and PPI inflation fell way below what anyone expected, and what is more, now everyone is worried about deflation in China. This to me isn’t necessarily consistent with the idea that Chinese inflation was all about food.

But there is a more technical problem with the idea that the drop in inflation was simply a consequence of declining food prices. In the standard inflation model the rapid decline in food prices should have automatically caused Chinese households to spend less on food and more on non-food consumption. If monetary policy was not deflationary, this would cause the price of non-food goods to increase as the price of food decreased.

But that didn’t happen. From September to November, CPI inflation has declined from 4.6% to 4.0% to 2.4%. As expected, food inflation declined from 9.7% to 8.5% to 5.9%.

But non-food inflation also declined. It dropped from 2.0% in September year on year to 1.6% in October to 0.6% last month. If monetary conditions were stable, just as non-food prices refused to behave the way they should have when food prices were rising, they still refuse to behave correctly when food prices are declining. This suggests to me that the pork camp still has to do a lot of explaining, or else the conclusion must be that Chinese inflation has not been caused by changes in food supply, but rather by changes in the money base.

If this is the case, it has very worrying implications. China’s reserves continue to rise, so the central bank has been continuing to create yuan at a rapid pace, but somehow money supply is contracting rapidly, as indicated by the decline in inflation. What could be happening?

Before answering, let me digress. About two months ago I had dinner with an old friend of mine, Arminio Fraga, the former head of the Brazilian central bank and one of the smartest finance guys I know. During our dinner we talked about conditions in China and I wanted his advice on monetary conditions. I had always felt that monetary aggregates in China did not seem to be explaining what was really happening in underlying money. My instinct was reinforced by a piece I had read by Ronald McKinnon and Gunther Schnabl titled “China’s Financial Conundrum and Global Imbalances” in which they say:

Why didn’t China rely more heavily on domestic financial indicators? With rapid financial transformation and very high saving, the velocity of money – whether based on M0, M1, or M2 – was (is) too unpredictable for any monetary aggregate to be useful as an intermediate target. And the velocity of money, defined as GDP/M, becomes even more difficult to predict when nominal GDP itself is subject to large revisions. Indeed nominal GDP was revised sharply upward in 2006.

Since 1990, Figure 5 shows that these monetary aggregates grew faster than nominal GDP—with M2 growing twice as fast so as to approach 200 percent of nominal GDP in 2008. The high growth in M2 was largely a natural result of China’s very high saving rate when bank deposits are the principal financial asset open to Chinese savers. Thus, the authorities had, and still have, no firm idea of what the noninflationary rate of growth in M2 should be.

I asked Arminio based on his tenure at the Brazilian central bank what he thought about the monetary aggregates (M0, M1, M2 and so on) to judge money growth. He responded that from his experience McKinnon was right and the traditional monetary aggregates were not terribly useful in evaluating money conditions. It was far more useful, he claimed, to look at credit growth.

But in China there is a big problem with the credit data. We don’t have good numbers on credit growth. We know what is happening in most parts of the formal banking system because the PBoC has fairly good numbers there, but we have only a vague idea about off-balance sheet transactions, about financing involving municipalities and provinces, and about inter-company lending. Most importantly we have very little idea about the informal banking sector.

But let’s get back to inflation. During the inflation period, officially credit growth was not too high, but there is strong evidence that during the period of explosive credit growth, when the central bank limited the amount of credit creation permitted to the banking system, the market responded by accommodating explosive growth in off-balance-sheet items and within the informal banking system. Anecdotal evidence suggests that much of China’s huge mot money inflow before this summer may have ended up in the informal banking sector. This (and the explosion in reserve accumulation) is consistent with the idea that inflation in China was money based.

It is the opposite now. Banks don’t want to lend and this implies that credit is contracting. Now that banks are not willing to lend, and companies not willing to borrow to invest (they mainly borrow if they are desperate for liquidity), credit growth is mainly being caused by government pressure to maintain credit growth. But of course credit growth in China is not just a function of the size of commercial bank portfolios. These may be growing, even if very slowly, but transactions that were once off-balance sheet may be coming back on balance sheet to give the illusion of credit growth. More worryingly, the informal banking sector, which may consist of as much as one-third of total loan assets, may be contracting rapidly. That is what the anecdotal evidence suggests. As an aside, in today’s weekly meeting of the Guanghua Students Monetary Policy Committee, one of the members, Gao Ming, reported that he had been getting a huge number of phone text messages from informal lenders offering him money. The other students agreed that they had seen a big rise in these kinds of text messages. Perhaps informal lenders are having trouble finding borrowers?

Certainly if inflation is a monetary phenomenon, this would be the implication of the near-collapse of inflation, which occurred more rapidly than even the most optimistic members of the pork camp expected. So what does this mean if true? It means that money is contracting, and perhaps at an alarming rate. Just as too-rapid money expansion until this summer was showing up as inflation, too-slow expansion (or perhaps even contraction) is showing up as disinflation and even deflation.

I can’t prove it, of course, but I think Chinese inflation is a money-based phenomenon and I am guessing that if China is experiencing deflation it is also experiencing rapid contraction in outstanding credit – much more rapid than we think. In China like in most developing countries, as I have written many times before, the structure of balance sheets tends to reinforce trends, which increases underlying volatility. We may have suddenly swung from ferocious credit expansion to ferocious credit contraction. The fear of deflation next year, in that case, is well founded.

About Michael Pettis 166 Articles

Affiliation: Peking University

Michael Pettis is a professor at Peking University's Guanghua School of Management, where he specializes in Chinese financial markets. He has also taught, from 2002 to 2004, at Tsinghua University’s School of Economics and Management and, from 1992 to 2001, at Columbia University’s Graduate School of Business.

Pettis has worked on Wall Street in trading, capital markets, and corporate finance since 1987, when he joined the Sovereign Debt trading team at Manufacturers Hanover (now JP Morgan). Most recently, from 1996 to 2001, Pettis worked at Bear Stearns, where he was Managing Director-Principal heading the Latin American Capital Markets and the Liability Management groups.

Visit: China Financial Markets

Be the first to comment

Leave a Reply

Your email address will not be published.