Is loan growth in China slowing?
By Michael Pettis
Most of this week’s very long newsletter focused on the NPC meeting in Beijing, the proposals to boost consumption (which I think will greatly disappoint), and the release of data by the National Bureau of statistics. Last week, for example, new lending numbers were released, and if it hadn’t been for rumors all the previous week that they would come it at RMB500-600 billion I think we would have all been surprised by how low they were. Here is the relevant article in Xinhua:
The People’s Bank of China (PBOC), the country’s central bank, said Monday that new yuan-denominated loans stood at 535.6 billion yuan (81.52 billion U.S. dollars) in February.
The figure was 192.9 billion yuan less than February last year, said the PBOC in a statement on its website. By the end of February, the balance of outstanding yuan-denominated loans stood at 48.89 trillion yuan, up 17.7 percent from a year earlier. The rise was 9.5 percentage points lower than the rise a year earlier.
As I discussed last week, total new lending has represented a rapidly declining share of the total lending the PBoC now monitors, so we don’t really know how much new credit the banks extended in February. The table below was released two weeks ago by the PBoC (I have summarized it), and it shows what has been happening:
The total represents all of the financing that the PBoC is now tracking, and is listed as 100%. I have broken this total lending into a few especially important categories and listed the percentage they comprise of the total. By the way I have left out several categories that I didn’t think especially useful to the discussion.
Notice that at the beginning of the decade new RMB bank lending represented virtually all new financing – 92% as far as the PBoC calculates. What is striking is how quickly it came down – in 2010 new RMB loans represented only 56% of total new lending. It didn’t decline, of course, because of any restraint in new lending. On the contrary new lending has exploded, especially in 2008, 2009 and 2010.
It came down because other categories surged. We have discussed this on my blog for several years. I have argued many, many times that limiting loan growth through administrative measures (loan quotas, for example) while keeping interests low, credit risks socialized, and maintaining pressure for investment driven growth, could not help but result in an explosion of lending outside the normal channels which the PBoC and the CBRC simply would not be able to control.
When you have excessively loose monetary and credit policy, you will automatically get a rise in risky loans. Not even Japan in the 1980s was able to violate this rule of finance (for all they dismissed it as a “western” rule), and China has not been able to do so either. The PBoC tried to regulate loan growth by putting into place a RMB quota system, but the consequence was completely predictable, and in fact was predicted by several of us.
The system adjusted so as to allow more loan growth to take place outside the regulated areas. Banks, in other words, simply created alternative forms of financing to get around the rules. Now that the PBoC is monitoring this wider range of lending activities, if they start trying to control growth in all of these areas I think it is pretty safe to assume that even newer forms of lending will develop.
I think it is also pretty safe to assume that in a month or so we are going to see a renewed expansion in credit. With controls on new lending a lot of companies are struggling for financing and I suspect growth is going to slow down by more than Beijing can bear.
The “correct” exchange rate always changes
I want to turn away from recent events in China to focus on a more technical or macro discussion. My friend Robert Kapp, President of Robert A. Kapp & Associates, Inc, recently had a question about the undervaluation if the RMB. This question is about an issue the represents a pretty common confusion in the debate over the revaluation of the RMB, so I asked him if I could reproduce his question for this newsletter:
I hope someone can help me with this. In 2003, when the currency-value flap took off, the high end of the allegations of over-valuation, if I remember correctly, was about 30%. The RMB was at 8.28 to the dollar.
Today, my handy little forex-gadget tells me that $1 is equal to RMB 6.577. My hand calculator tells me that that represents an appreciation against the dollar of about 20.56 percent.
How is it that, having risen 20.6% against the dollar since the RMB first surfaced in Washington, the RMB is now claimed variously to be anywhere from 25% to (most commonly used in political rhetoric) 40% undervalued against the dollar?
“This is not a rhetorical question,” he adds, “there has to be an economic answer.”
There is. To simplify his question, a few years ago people suggested that the RMB might be undervalued by 30%. Since then the RMB has appreciated by 20-25%. And yet today people are still arguing that the RMB may be undervalued by 30%. How is it possible that so much appreciation has not seemed to affect the estimates of undervaluation?
Before answering it is worth pointing out that there is no way that anyone can determine precisely the amount of undervaluation of the RMB, or any other currency, and so any estimate can be nothing more than that – an estimate based on many moving parts. There are plausible reasons for arguing that a currency is undervalued or overvalued, but there is absolutely no way to determine with any precision by how much.
This difficulty is compounded by the fact that many analysts are simply getting the math wrong. So for example when people say the RMB is undervalued by 30%, they often mean that the dollar is overvalued by 30%. These two claims may sound like the same, but of course they aren’t. If the RMB is undervalued by 30%, it means that the dollar is overvalued by 43%, not 30%. I have seen so much confusion on this issue that I pretty much give up on trying to understand what people mean when they discuss currency changes without seeing their actual numbers.
So for example if the RMB went from 8.26 to 6.58, as Kapp argues, the RMB actually appreciated by 25.9%, and not by the 20.6% he calculated. It was in fact the dollar that depreciated by 20.6%. What’s more if we assume that the RMB was indeed initially undervalued by 30%, the subsequent appreciation of the currency by 25.9%, all other things being equal, leaves it still undervalued by 3.3%,
Price differences matter
But that still leaves the underlying question unanswered. Why do people still insist on saying that the RMB is undervalued by roughly the same amount as it was several years ago after it has already appreciated so much during that time?
It turns out that there are many reasons. The “correct” exchange rate is never correct except at a single point in time. Six years later, as in this case, the new correct level will be different because of several factors.
The most obvious factor that can change the relative valuation of the nominal exchange rate is the inflation differential in the tradable goods sector of each country. If Chinese inflation were say 2% lower than US inflation, it would imply that in real terms the RMB depreciated over six years by a further 11%, roughly.
Many people take a shortcut and just assume that what matters is the difference in CPI inflation between the two countries, but this is a mistake, and one I wrote about in my newsletter a couple of months ago. So analysts argue, incorrectly, that because in the past year the RMB has appreciated nominally by 4%, and in addition there has been a 3% inflation differential between China and the US, the RMB has actually appreciated in real terms by 7%.
It hasn’t. What matters is not CPI inflation but rather the inflation in the cost of inputs in the tradable good sector. I don’t have the numbers in front of me, but over the past few years most inflation in China has been in the food sector, and not in the tradable goods sector. If the relevant US inflation exceeded the relevant Chinese inflation during the past six years, the inflation differential argument might explain part of the reason for the continued high estimates of RMB undervaluation. There would have been a real depreciation of the RMB during these years that would have moved against the direction of its nominal appreciation.
The second obvious factor is the productivity growth differential between two countries, or perhaps more accurately, the US/China differential in the domestic differential between wage growth and productivity growth. If workers in one country become more productive at a faster rate than workers in another country, and that difference isn’t neutralized in the form of higher wages, it effectively lowers the real vale of the exchange rate.
So if Chinese worker productivity grew faster than US worker productivity by 3% annually, and if wages failed to keep up, the RMB would over a six year period depreciate in real terms by roughly 16% (again, the actual nominal numbers matter if you want greater precision, as does of course the wage/productivity differential in the tradable goods sector, not the country as a whole).
The third obvious factor in countries like China is the cost of capital in the tradable goods sector. If the cost of capital is heavily subsidized in China (which it is) and not in the US (which it isn’t, except recently at very short maturities, which is not too relevant for producers anyway), this counts as an effective depreciation of the RMB. The cost of capital is of course an important input in the production of tradable goods, and to subsidize its cost is no different than putting into place import tariffs and export subsidies. In both cases you reduce the real exchange value of the currency.
Any subsidy matters
The amount of undervaluation caused by artificially low interest rates is much harder to calculate than in the other cases, but the sheer amount of interest-rate repression in China (at least 400-600 basis points, by my calculation), suggests that this may be the biggest single reason for arguing that the RMB is still significantly undervalued in spite of the 26% appreciation in the past six years.
Finally, as all of the above factors imply, any other differential in the growth rate of subsidies, including taxes, will imply real appreciation or depreciation in a currency. If Chinese manufacturers get subsidized land, or subsidized energy, this has the same impact on the trade balance as lowering the nominal exchange rate.
Remember that an undervalued exchange rate is nothing more than a consumption tax on imports, the proceeds of which are used to subsidize manufacturers in the tradable good sector. It reduces household consumption by reducing real household income, and it increases production by subsidizing manufacturing. By putting upward pressure on the gap between the two, it puts upward pressure on the trade surplus.
There are many other ways of doing the same thing. Any time household consumers are explicitly or implicitly taxed, and the proceeds used directly or indirectly to subsidize manufacturers in the tradable goods sector, for example by forcing household depositors to lend money to manufacturers at artificially low rates, it is the same thing as devaluing the exchange rate. In either case it forces an artificial split between production growth and consumption growth, and this shows up as trade intervention.
The reasons above, and the confusion they create, is why I worry when people focus too narrowly on the currency in terms of pricing equivalence as the source of trade imbalances. The correct way to look at the causes of trade imbalances, in my opinion, is to look at policies that force up or down the savings rate, or the consumption rate (which is more or less the same thing since total savings is simply total production minus total consumption).
In that sense anything that reduces consumption and increases production forces up the savings rate, and unless there is an equivalent increase in investment, it also forces up the trade surplus. Undervalued currencies (like repressed interest rates or low wage growth relative to productivity growth) are a kind of tax on households and so reduce consumption, and they are a subsidy for manufacturers and so increase production. This is what forces up the savings rate and so forces up the trade surplus. If China were to raise the value of the RMB and simultaneously lower real interest rates, which is has done in the past year, you could have the seeming paradox of a rising RMB and even greater RMB undervaluation.
This is an abbreviated version of the newsletter that went out Tuesday. Academics, journalists, and government and NGO officials who want to subscribe to the newsletter should write to me at firstname.lastname@example.org, stating your affiliation, please. Investors who are clients of Shenyin Wanguo Securities will already receive the newsletter. Investors who are not clients but who want to buy a subscription should write to me. also at that address.