The Economist tends to be better at economics than other media outlets, because it's their focus.
However, that doesn't exempt them from mistakes in trying to find a story.
They try and refute three signals of a bubble - asset prices too high, investment too high (leading to inefficient projects) and bank lending too high. Their mistakes are subtle, but they're there...
They cite the "reasonable" Shanghai A share P/E of 28. Paragraphs later, they mention that China is not very levered. The problem with this is that leverage typically inflates P/Es in the short term, meaning that a 28 P/E in an unlevered environment is actually really high. There are three reasons for this.
Firstly, if I have a company that has 100 in EBIT (earnings before interest and taxes - basically, the money a company makes before paying debtholders, equityholders and the government), and no debt (assume a 0% tax rate), then all 100 flows to the equityholders. The earnings are 100. 5% revenue growth translates to 5% earnings growth.
If, instead, you have 100 in EBIT and 90 in debt, then only 10 flows through to equityholders... but if EBIT goes to 105, then all 5 of the new EBIT flows through to equityholders, because you dont have to pay more interest on debt just cuz your revenue goes up. Thus, it's much easier to see higher equity earnings growth in a levered economy - in this example, 5% revenue growth leads to 50% equity growth! A 28 P/E for a low-leverage country means that growth expectations are off the charts. If the country misses, then the stock market is in trouble.
The second problem is that if the expected return on invested capital exceeds only the interest payments required on debt capital (it doesn't even actually have to, just needs to be expected to), then any expectation of further leverage in the future would mean very, very fast growth, and the P/Es are high. Thus, that P/E of 28 may factor in expectations of colossal further leverage. The banks, however, are seriously overexerted (see below), so that leverage may not come for a long time. That's another recipe for an equity collapse.
The final problem is that most Chinese companies are asset-heavy - manufacturing plants, natural resource mines, etc. They're also much more bureaucratic than most US companies. If this means that there are lots of fixed costs, then the companies there have operating leverage - similar to the EBIT example above, except instead of the 90 going to debtholders, its going to servicing fixed costs. That means that any slowdown at all can crush asset values.
Thus, it's pretty likely that a 28 P/E is very high. Home prices are probably the same way.
They cite that home prices relative to average incomes has fallen in the past decade, and if you look at the pool of Chinese homebuyers, they're not average - the ratio of home prices to average homebuyer income is the same as the home prices to average income in developed countries. The problem with this analysis is that you have to do the same thing to developed countries - including poor people in the developed country number and excluding poor people from China's number is not apples to apples. In this case, the developed country number probably drops well below 4 or 5, so China's number is still high. China is also continually building MORE houses, while they're not increasing the number of rich people that fast. Thus, housing prices can collapse very, very quickly. This is exacerbated by the fact that occupancy rates are already low in the biggest cities.
Finally, China's property boom being financed by saving, not bank lending, helps to not decelerate the property market so much, but the banks are still wayyyy overlevered - it's a product of deficient banking resources, not reasonable banks. If the banks there collapse, perhaps the bailout measures there will be smaller than in the US but they'll still have to be substantial. Meanwhile, China is more unequal than the US, and has less mobility, so even if a small number of people or builders start defaulting, it's unlikely that there will be buyers to pick up the slack, as there would be with an equivalent sized banking crisis here. In other words, if the price elasticity of demand is higher in China due to greater economic inequality, home prices can still collapse.
On capital investment, there are a number of problems as well. The first problem is that GDP numbers seem to be getting more and more inflated, but let's put that aside because the evidence is anecdotal, not statistical. The author looks at ICOR and TFP as measures of the efficiency of Chinese investment. This, too, has problems. ICOR doesn't acknowledge that a huge percentage of their capital investment has gone into infrastructure and buildings. Construction itself is a tremendous contributor to GDP in China, and it counts no matter how efficient it is to build the building or infrastructure, so if China is accelerating construction and construction is a big piece of GDP (as it is in China - 40+%), you expect ICOR to stay stable. It only dips when you stop accelerating construction. (disclaimer: I haven't worked with ICOR that much, so if I'm misinterpreting it, please let me know, but I don't think so...)
TFP is a) dubious in its year to year relevance, b) strongly affected by what happens in other countries, and c) far less relevant when large portions of your GDP are in commodities production and commodities have increased so much in price over the last few years. That has the ability to mask declines in incremental capital efficiency.
Finally, the author does concede that bank lending is in trouble and the gvt will have to stomach a lot of those losses. He also ignores the fact that credit to GDP is less important than credit to bank deposits (and I don't know what those have been). He cites that the credit growth:gdp growth has been less than most developed countries since 2004, ignoring that the rest of the world had a massive credit bubble since 2004.
Overall, while I understand that people who think China will go the way of Japan (crash, and never recover) are probably overstating the case, it is probably naive to look at China and pretend it's not in a bubble. Maybe I overrate the size of the bubble - that is certainly possible - but most of what we see in China right now seems very, very bubbly.
No comments:
Post a Comment