Small companies feel pain as China tightens
By Michael Pettis
Much of last week’s newsletter was on the data that was released on inflation, new loans, investment, etc. One thing worth noting in the current environment is that small and medium enterprises (SMEs) are hurting, and last week’s minimum reserve hike – which came out the same day as the CPI number – will undoubtedly make things worse since any withdrawal of credit is unlikely to affect the top customers (SOEs and local governments) and will fall disproportionately on marginal borrowers. In early May I wrote that the very uneven process of rebalancing in China was likely to have adverse consequences on the SME sector.
Rebalancing in the context of China means, for the most part, a significant increase in the consumption share of GDP from its astonishingly low level of 35% in 2009 (and perhaps lower last year). As regular readers know, I do not believe that China’s high savings and low consumption rates are a function of any remarkable preference on the part of Chinese households for savings over consumption.
They are simply the automatic consequence of a system in which increases in GDP growth are subsidized by transfers from the household sector, which effectively constrains the relative growth of household income and, with it, household consumption. In that case the only way for China to rebalance would be for Chinese household income to grow faster than GDP.
This requires three things above all. It requires that wages grow faster than productivity, that the currency appreciates, and that real interest rates rise. This does not seem to be happening in China, as I discussed in a blog entry last month. Wages have been rising quickly in the past year, but the currency is barely appreciating in real terms against the dollar (and is probably depreciating on a trade weighted basis), and of course real interest rates are declining sharply.
In the aggregate the impact on the growth of the household income share of GDP is probably negative, at least if the World Bank is correct in saying that consumption growth has been actually declining since early last year (and as Tuesday’s retail sales data seem to confirm). But this uneven rebalancing has another impact, I argued. If wages are rising and the cost of capital is declining, we should be seeing a shift internally in favor of capital-intensive industries, the SOE sector, for example, and away from labor-intensive industries, which includes SMEs for the most part.
I had anecdotal evidence corroborating the theory. Since last year several of my students who come from manufacturing families had told me that their parents were finding it increasingly difficult to retain workers and were getting squeezed out of business. Two months ago my central-bank-seminar student Huang Haidong, who comes form a prominent Wenzhou family (Wenzhou is often considered the cradle of the SME industry in China), sent me an email saying that according to his research a very large number of Wenzhou SMEs were facing bankruptcy.
Wages and capital
The evidence has since become more than just anecdotal. Last week I saw the following article in the current issue of Caixing:
China’s small- and medium-sized enterprises are looking at an increasingly difficult business environment amid financing difficulties and rising production and labor costs, the Ministry of Industry and Information Technology (MIIT) said in a report released on June 2.
To rein in excess liquidity, the central government has tightened its monetary policy and raised interest rates twice so far this year, but the measures also add to the difficulty for SMEs to obtain bank lending and have driven up financing costs, according to the report, which was jointly conducted by MIIT and the Chinese Academy of Social Sciences.
…The official Xinhua News Agency reported on June 2 that three “relatively big” private companies went bankrupt in Wenzhou of Zhejiang Province. Wenzhou is known as one of the largest centers of private enterprise growth in China. But a local official responded by saying they are isolated cases.
The article argued that high wages and rising borrowing costs are the key challenges facing SMEs. The latter may seem surprising given what I said about the declining real cost of capital, but remember that SMEs get little of their financing from the banking sector and have to pay very high borrowing costs for off-balance sheet and informal lending. As cheap capital has been sucked up by SOEs, real estate developers (those who can access bank lending), infrastructure investors and local and municipal governments, little has been left for less-favored entities and the result has been that their cost of capital has risen even as wages have too.
The point was picked up Friday in a very interesting article in the Financial Times. According to the article:
Small Chinese businesses are feeling the effects the government’s monetary tightening and face a cash squeeze that may be worse than during the global financial crisis in 2008, according to an official warning.
From tile manufacturers in Shanghai to shoe factories near Hong Kong, smaller businesses have driven Chinese growth over the past two decades, accounting for about 60 per cent of gross domestic product. So, a sharp slowdown in their activity would weigh heavily on the Chinese, and by extension, world economy.
The article goes on the make comparisons with the last squeeze of SMEs:
In late 2008, during the financial crisis, a collapse in global export markets and aggressive monetary tightening drove the Chinese economy to a near standstill. Beijing estimated that 20m migrant workers, many of whom were employed by the same kinds of small firms that are now short on cash, lost their jobs.
Dong Tao, a Credit Suisse economist, said China could now face trouble again, as companies delay bill payments and factory owners abscond without paying wages.
“If this continues, and these smaller companies start to fail in large numbers, then the economy will slow more than the market anticipated and the government bargained for,” he said.
Most of the focus in the press has been on the effect of rising borrowing costs for SMEs, although many of these are not heavily leveraged and in my discussions with friends who own, or who are related to owners of, SMEs, wages has been at least as much of a problem. But the attempts by the PBoC to tighten credit (relative to the enormous demand from credit from infrastructure and real estate developers and local governments) have certainly made conditions tougher for those SMEs who do rely on credit. The FT article goes on to say:
The cash crunch has come despite repeated prodding by the government to help private businesses. Chinese banks have traditionally preferred to lend to state-owned groups, judging that they pose negligible credit risk due to their government backing.
This bias is especially pronounced when the government restricts credit, as it has done over the past year. China has raised benchmark lending rates by 100 basis points to 6.31 per cent, but small businesses have seen much steeper increases.
Monthly lending rates at credit unions and informal lending institutions in the entrepreneurial cities of Wenzhou and Xiamen have reached 5 per cent in the past few weeks, up from 1.5 per cent just nine months ago, according to Credit Suisse.
Earlier this week, in an effort to boost access credit, the Chinese banking regulator said it would ease tough capital rules on bank lending to smaller businesses. In one measure, loans of less than Rmb5m ($770,000) need not be counted towards banks’ loan-to-deposit ratios.
This last point is interesting. Last week a CBRC circular went out that suggested a different way of accounting for small loans to SMEs which should, in principle, increase banking incentives to fund SMEs by reducing the capital requirement.
This seems to me a very complicated way of alleviating the cash crunch among SMEs. The proper way would be to reduce the demand for credit from real estate developers and infrastructure investors, but of course Beijing loves its administrative measures. Anyway the only efficient way of reducing demand from real estate developers and infrastructure investors would be to have them pay a fair cost for their capital – and remove the implicit credit support. This however would threaten to throw a huge number of essentially insolvent projects and companies into bankruptcy, so the preferred solution is to keep the cost of capital low and to keep their borrowings growing.
Will the attempt to on increase incentives to lend to SMEs work? I am not sure. It would depend on too many factors and squeezing out liquidity through hikes in the minimum reserve ratio will almost certainly fall disproportionately on SMEs. Of course if the regulators simply imposed a formal or informal quota on the banks (e.g. by the end of the year a minimum of X% of your loan portfolio must be in the form of small loans to SMEs), I suspect that they will comply, but then we run into the automobile loans problem of a few years ago. When the regulators demanded that Chinese banks increase the number of auto loans, they duly did, and within a very short time a remarkable share of those loans went non-performing. It turns out that it is very easy to meet aggressive loan target numbers, but it is a lot less easy to make good loans.
Moreover in China credit is not controlled through price, but rather through quantity. In other words we don’t use the cost of capital to decide the quantity of credit since the cost of capital is artificially set at extremely low levels – at which level the demand for loans is almost infinite.
Instead we use loan quotas. This means that if banks lend materially more to SMEs either they will also be forced to lend less to other borrowers or they will have to allow credit growth to exceed whatever target they would have otherwise set. Already there is some concern about the flow of capital into low-income housing projects. Last week‘s Beijing-based Global Times had this article:
At least 40 percent of public housing projects may not start as scheduled this year, with the loan bottleneck resulting from monetary tightening policy mainly to blame for the delay, analysts said on Monday. Construction of affordable housing in cities nationwide is far behind schedule, according to data cited in a Xinhua report.
In Shanghai, only 25 percent of scheduled affordable housing projects had started by the end of May, Xinhua reported. In Jiangsu Province, work on about 30 percent of a total of 450,000 affordable housing projects had started, while in Zhejiang Province, work on about 61,600 affordable housing units, 33.2 percent of the total, was underway.
The government plans to build 10 million units of affordable housing this year as alternatives for homebuyers in cities where average property prices have almost doubled during the past two years. Under the plan, construction of all these affordable housing projects must begin by the end of October.
Hui Jianqiang, a senior analyst with the E-House R&D Institute, said work on at least 40 percent of affordable housing projects will not begin this year unless the central government links local officials’ performance to public housing construction. “Housing activity is anemic partly due to stricter approval procedures but mainly because of tight lending standards under current monetary tightening policy,” Hui said.
More lending to SME is undoubtedly a good thing, but if it is achieved by diverting loans from local governments, for example, it will come at the cost of lower growth and rising financial distress in the short term. And as we have seen in the past, short-term growth concerns always trump long-term sustainability issues.
Whatever the outcome I think we need to keep a close eye on the behavior of SMEs. They are by far the most efficient part of the Chinese economy and the only part creating real value. My experience in other developing countries suggest that SME owners tend to be the most sensitive to and aware of changes in risk, and if we start to see rising bankruptcies among SMEs, coupled with disinvestment and increasing capital outflows, that is almost always a very worrying sign. When SME owners start to worry, so should we.
MIT on environmental degradation
To turn to a completely different issue, lat week a friend sent me an interesting article that came out in one of the MIT magazines. According to MIT News:
A recent study released by the MIT Joint Program on the Science and Policy of Global Change quantifies the damage to the Chinese economy caused by a lack of air-quality control measures between 1975 and 2005. Not surprisingly, the MIT researchers found that air pollutants produced a substantial socio-economic cost to China over the past three decades.
…To observe how changes in pollutants, and their associated health impacts, have historically affected the Chinese economy, the MIT researchers modeled the number of cases of health incidences caused by air pollution — such as restricted-activity days, respiratory hospital admissions and asthma attacks, to name a few examples — given a pollution level and the number of people exposed. Then the model calculated the summed costs of these incidences — i.e., payments for health services and medicine, loss of labor and productivity from time off work, loss of leisure time needed for healing — to estimate the total change in available labor supply.
Similar studies conducted by the World Bank have found that air pollution in China caused damages equal to 4-5 percent of the Chinese GDP between 1995 and 2005. However, these estimates are based on static measurements that do not measure the cumulative, long-term impacts of health damages. The MIT study found a significantly higher level of damage, equaling 6-9 percent of the Chinese GDP. The dynamic, cumulative method used in the MIT study may be particularly applicable to developing countries that are experiencing rapid growth.
I mention this because I have often argued that Chinese GDP growth has been substantially overstated during much of the past thirty years. Part of the reason for the overstatement is that the future costs of environmental degradation should in principle be included as a deduction to current growth.
After all if environmental degradation reduces future economic output because of health problems, not to mention because destroying rivers, farm land, and so on is the economic equivalent of selling assets and calling the proceeds income, then the growth in economic value it generates today should be reduced by the destruction in economic value.
So what is China’s real GDP?
This is hard to do, of course, but it eventually gets accounted for in the form of lower growth in the future. If farmers produce less tomorrow because water is polluted, then future economic value added is lower. If workers spend additional money on health care tomorrow, this money is transferred from other, more productive spending.
This happens everywhere, of course, but I would argue that in many countries, where environmental degradation has been less and has occurred over a much longer period, it is already showing up in lower GDP growth today, so it probably results in a much lower overstatement of growth. In fact in rich countries where environmental degradation has slowed sharply, or even reversed, it may be causing GDP growth to be understated.
The other source of GDP overstatement in China is misallocated investment. One way of thinking about it is that if NPLs were correctly identified, the annual accumulation of the non-collectible portion of NPLs should be deducted from current GDP growth numbers to arrive at a more accurate estimate of GDP. After all growth “created” by wasting money is not really growth, and NPLs represent the amount of money that has been wasted.
In order correctly to identify NPLs we would need to include loans that might not technically be NPLs at current interest rates, but would be if interest rates were raised (by at least 400-600 basis points) to their “correct” level. Why? Because these loans are benefitting from the implicit annual debt forgiveness granted to them by household depositors – and the fact that they can pretend to be performing with the help of massive debt forgiveness should not change the fact that they are nonetheless un-repayable.
The combination of these two sources of GDP overstatement – uncounted environmental degradation and ignored NPLs – is pretty substantial. To show how substantial, assume that GDP has been overstated by anywhere from 2 to 4 percentage points over the past ten to fifteen years. This would imply that China’s GDP today is actually about 55% to 85% of its stated size – or to put it another way, that China’s economy is anywhere from 15% to 45% smaller than we think.
This is a pretty big haircut. I have no idea what the correct deduction is (none of my numbers seem especially implausible), but even very rough ballpark numbers suggest that China’s GDP may be sharply overstated. At the very least they also suggest that all those breathless predictions about when China will have the world’s largest GDP may turn out to be as simple-minded as the same predictions made about the USSR in the 1960s or, perhaps a little more plausibly, about Japan in the 1980s. And for the same reasons: in each case we start from the assumption that the country’s real GDP, inflated as it is by misallocated environmental costs and overstated investment numbers, is much larger than it really is. Much, much larger.
By the way notice that if we discount GDP by 20-40%, the astonishingly low household consumption share of China’s GDP – 35% in 2009 – rises to 44-59% – still very low by global standards, but not quite as surreal. Could it be that much of China’s GDP really is overstated, and with it total savings too?
This is an abbreviated version of the newsletter that went out last week. Academics, journalists, and government and NGO officials who want to subscribe to the newsletter should write to me at email@example.com, 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.