Is China a bubble?
No, at least according to The Economist:
By most measures average prices have fallen relative to incomes in the past decade (see chart 1).
The most cited evidence of a bubble””and hence of impending collapse””is the ratio of average home prices to average annual household incomes. This is almost ten in China; in most developed economies it is only four or five. However, Tao Wang, an economist at UBS, argues that this rich-world yardstick is misleading. Chinese homebuyers do not have average incomes but come largely from the richest 20-30% of the urban population. Using this group’s average income, the ratio falls to rich-world levels. In Japan the price-income ratio hit 18 in 1990, obliging some buyers to take out 100-year mortgages.
Furthermore, Chinese homes carry much less debt than Japanese properties did 20 years ago. One-quarter of Chinese buyers pay cash. The average mortgage covers only about half of a property’s value. Owner-occupiers must make a minimum deposit of 20%, investors one of 40%. Chinese households’ total debt stands at only 35% of their disposable income, compared with 130% in Japan in 1990.
China’s property boom is being financed mainly by saving, not bank lending. According to Yan Wang, an economist at BCA Research, a Canadian firm, only about one-fifth of the cost of new construction (commercial and residential) is financed by bank lending. Loans to homebuyers and property developers account for only 17% of Chinese banks’ total, against 56% for American banks. A bubble pumped up by saving is much less dangerous than one fuelled by credit.
OK, but what about investment, isn’t that a bubble?
China’s second apparent point of similarity to Japan is overinvestment. Total fixed investment jumped to an estimated 47% of GDP last year””ten points more than in Japan at its peak. Chinese investment is certainly high: in most developed countries it accounts for around 20% of GDP. But you cannot infer waste from a high investment ratio alone. It is hard to argue that China has added too much to its capital stock when, per person, it has only about 5% of what America or Japan has. China does have excess capacity in some industries, such as steel and cement. But across the economy as a whole, concerns about overinvestment tend to be exaggerated.
. . .
Even in industries which clearly do have excess capacity, China’s critics overstate their case. A recent report by the European Union Chamber of Commerce in China estimates that in early 2009 the steel industry was operating at only 72% of capacity. That was at the depth of the global downturn. Demand has picked up strongly since then. The report claims that the industry’s overcapacity is illustrated by “a startling figure”: in 2008, China’s output of steel per person was higher than America’s. So what? At China’s stage of industrialisation it should use a lot of steel. A more relevant yardstick is the America of the early 20th century. According to Ms Wang of UBS, China’s steel capacity of almost 0.5 tonnes per person is slightly lower than America’s output in 1920 (0.6 tonnes) and far below Japan’s peak of 1.1 tonnes in 1973.
. . .
Given the scale of the spending, some money is sure to have been wasted, but by and large, investment in roads, railways and the electricity grid will help China sustain its growth in the years ahead. Some analysts disagree. Pivot, for instance, argues that China’s infrastructure has already reached an advanced level. It has six of the world’s ten longest bridges and it boasts the world’s fastest train; there is little room for further productive investment. That is nonsense. A country in which two-fifths of villages lack a paved road to the nearest market town still has plenty of scope for building roads. The same goes for railways. Again, a comparison of China today with the America of a century ago is pertinent. China has roughly the same land area as America, but 13 times more people than the United States did then. Yet on current plans it will have only 110,000km of railway by 2012, compared with more than 400,000km in America in 1916. Unlike Japan, which built “bridges to nowhere” to prop up its economy, China needs better infrastructure.
OK, but aren’t the banks in trouble?
The biggest cause for worry about China is the third point of similarity to Japan: the recent tidal wave of bank lending. Total credit jumped by more than 30% last year.
. . .
However, too many commentators talk as if Chinese banks have been on a lending binge for years. Instead, the spurt in 2009, which was engineered by the government to revive the economy, followed several years in which credit grew more slowly than GDP (see chart 3). Michael Buchanan, of Goldman Sachs, estimates that since 2004 China’s excess credit (the gap between the growth rates of credit and nominal GDP) has risen by less than in most developed economies.
Even so, recent lending has been excessive; combined with overcapacity in some industries, it is likely to cause an increase in banks’ non-performing loans. Ms Wang calculates that if 20% of all new lending last year and another 10% of this year’s lending turned bad, this would create new bad loans equivalent to 5.5% of GDP by 2012, on top of 2% now. That is far from trivial, but well below the 40% of GDP that bad loans amounted to in the late 1990s.
But what about the Chinese stock market bubble?
Japan’s stockmarket and land-price bubbles in the early 1960s offer a better (and more cheerful) analogy to China than the 1980s bubble era does. Japan’s economy was poorer then, although relative to America its GDP per person was more than double China’s today, and its trend rate of growth was around 9%. According to HSBC, after the bubble burst in 1962-65, Japan’s annual growth rate dipped to just under 6%, but then quickly rebounded to 10% for much of the next decade.
South Korea and Taiwan, which experienced big stockmarket bubbles in the 1980s, are also worth examining. In the five years to 1990, Taipei’s stockmarket surged by 1,600% (in dollar terms) and Seoul’s by 700%, easily beating Tokyo’s 450% gain in the same period. After share prices slumped, annual growth in both South Korea and Taiwan slowed to around 6%, but soon regained its previous pace of 7-8%.
The higher a country’s potential growth rate, the easier it is for the economy to recover after a bubble bursts, so long as its fiscal and external finances are in reasonable shape. Rapid growth in nominal GDP means that asset prices do not need to fall so far to regain fair value, bad loans are easier to work off and excess capacity can be more quickly absorbed by rising demand. (Italics added.)
Imagine that, rapid NGDP growth makes it easier to work off credit excesses. I wonder what happens if NGDP falls during a credit crisis. (You knew I wasn’t going to write an entire post without mentioning NGDP. )
China has lots of problems: poverty, pollution, inefficient SOEs, lack of rural property rights, political repression, and worst of all (in my view) a dangerously unbalanced male/female gender ratio. Bubbles are the least of their problems (although asset prices in China will continue to be highly volatile, and at times this will make it seem as if the bubble worriers were correct.)
BTW, Nick Rowe has another excellent post on bubbles. He draws a useful distinction between the perspectives of social scientists and market participants. I entirely agree with his post, which meshes with my earlier argument that anti-EMH theories are useless. Anti-EMH theories are generally created by social scientists, who are wasting their time if they think they can outsmart the market. If someone like George Soros develops a useful anti-EMH theory, I assume he would be smart enough to keep it to himself. Once the theory is in the public domain its market value falls to zero.
Tags: bubbles
23. January 2010 at 23:42
So three-factor and four-factor models are worthless? Nassim Taleb’s contrarian long volatility strategy became worthless the day the Black Swan was published? Mean reversion doesn’t exist?
Or maybe the market isn’t as rational as you might like to think.
24. January 2010 at 00:03
[…] TheMoneyIllusion » Is China a bubble? […]
24. January 2010 at 07:22
Jimmy, I’d like to think the market is irrational so that I could beat the market. But I find it difficult to get rich, and so I have a grudging respect for market efficiency. There are tests that show the market is efficient, and other tests that show it is inefficient. But I find the efficiency tests far more convincing. The tests showing inefficiency are often a transparent product of data mining.
The EMH is not literally true, of course, but it is much more useful than anti-EMH models (at least to me.)
24. January 2010 at 10:30
The EMH models did not work out so well for Alan Greenspan.
25. January 2010 at 08:06
Jimmy, Greenspan was right that monetary policy should not react to bubbles. He may have been wrong about regulation, but that has nothing to do with the EMH, it has to do with second best policies in a world of FDIC and TBTF.
Furthermore, the housing crash did not cause this recession, tight money did.
And wasn’t Greenspan famous for saying “irrational exuberance?” Doesn’t sound like he believed in the EMH.
25. January 2010 at 15:08
“that has nothing to do with the EMH”
Of course it does. If the price of a security wonders randomly about its intrinsic value, then why on earth would you rely on anything other than market discipline to regulate the market?
The “it’s hard to beat the market” part of the EMH is totally irrelevant to the debate that people are having in the real world today. As far as I know, that part of the debate was settled in the early 70s. Can we please move on now?
As for Greenspan, he was famous for saying “irrational exuberance,” and then not doing anything about it. He has admitted that and you very well know that.
And I do agree with you that monetary policy is not a good tool for popping bubbles. That does not change what the EMH is and what it says.
26. January 2010 at 07:57
jimmy, Yes, he didn’t do anything about it, but my point was that he didn’t believe in the EMH. That is pretty obvious from his adoption of Shiller’s views on irrational exuberance. I can’t be held responsible for people misusing the EMH. If he opposed efforts to rein in Fannie and Freddie because of the EMH, then shame on him. But I’ll bet he favored increasing regulation of those institutions. A better argument is that he didn’t understand how much FDIC and TBTF had distorted decisions by private banks. But if those view were based on the EMH, then he didn’t really understand the EMH and was misusing it. The EMH says absolutely nothing about market prices being at their socially optimal elvel. Nothing at all. Does the EMH say we shouldn’t worry about pollution? Of course not.
26. January 2010 at 17:41
The EMH has nothing to do with “socially optimal level.” The banks lost huge quantities of money buying assets that were massively overvalued. According to the EMH, those asset prices were wandering about their intrinsic values in 2006. Does anyone really believe that?
Let me put it another way. The defining characteristic of any hypothesis is that it can be falsified. What evidence would you have to see in order to falsify the EMH if not the asset bubble we just saw?
27. January 2010 at 19:49
Jimmy, You don’t understand the EMH. It does not say that asset prices wander around their intrinsic values. In a world of uncertainty traders may not know the intrinsic value of a security. The asset price may move far from the intrinsic value. I don’t think the term “intrinsic value” is even clearly defined.
I would consider the EMH falsified if it was clear that mutual funds successful one year were signficantly more likely to be successful the next. But studies show that is not the case. Smart mutual fund managers aren’t able to beat the market. It is mostly luck.
I would also consider the EMH falsified if stock prices responded to Fed annoucements with a lag of 30 minutes. But again, they don’t, they respond immediately. There are all sorts of things that might falsify the EMH.
Unfortunately, many economists think that data mining falsfies the EMH, whereas it merely shows computers are good at predicting the past.
28. January 2010 at 10:49
“Jimmy, You don’t understand the EMH. It does not say that asset prices wander around their intrinsic values.”
It most certainly does.
“Now in an uncertain world the intrinsic value of a security can never be determined exactly. Thus there is always room for disagreement among market participants concerning just what the intrinsic value of an individual security is, and such disagreement will give rise to discrepancies between actual prices and intrinsic values. IN AN EFFICIENT MARKET, HOWEVER, THE ACTIONS OF THE MANY COMPETING PARTICIPANTS SHOULD CAUSE THE ACTUAL PRICE OF A SECURITY TO WANDER RANDOMLY ABOUT ITS INTRINSIC VALUE… ALTHOUGH UNCERTAINTY CONCERNING INTRINSIC VALUES WILL REMAIN, ACTUAL PRICES OF SECURITIES WILL WANDER RANDOMLY ABOUT THEIR INTRINSIC VALUES.”
Fama, Eugene F. “Random Walks in Stock Market Prices.” Financial Analysts Journal (1965): 55-59.
I’m not sure how much clearer he could have stated it. What you’ve done is you’ve changed the definition of the EMH in order to raise the standard of falsification. What you are describing and defending is NOT the EMH. Call it what you like, “the unbeatable market hypothesis,” “the instantaneous adjustment hypothesis,” “the index funds are a really good investment hypothesis.” Just don’t call it the Efficient Market Hypothesis.
29. January 2010 at 06:45
jimmy, My fault. I don’t think we disagree about defintions, rather I misunderstood how you and Fama used the term “intrinsic value.” I thought you meant the intrinsic value that is estimated ex post. As when people say “we now know that tech stocks were priced above their intrinsic value in 2000.” I say that because your statement implied that the asset price should always be close to its intrinsic value. Obviously Fama doesn’t believe that, otherwise he wouldn’t believe in the EMH. He also observed the tech bubble in 2000 and the housing bubble in 2006, and doesn’t think those examples refuted the EMH. He argues that at the time investors rationally thought the tech industry might do much better than it actually ended up doing ex post. Obviously Fama is using intrinsic value as people believed at the time, given their information set.
Suppose a perfect and cheap artifical diamond is invented tomorrow. What is the intrinsic value of real diamonds today? Based on what we know today, at least $1000 per carat. But based on tommorrow’s invention, almost nothing. But prices today will still be close to their intrinsic value today as long as people don’t know what will happen tomorrow. I am pretty sure that is what Fama would say.
You said;
“According to the EMH, those asset prices were wandering about their intrinsic values in 2006. Does anyone really believe that?”
Yes, Fama and I believe that.
29. January 2010 at 12:04
“I say that because your statement implied that the asset price should always be close to its intrinsic value. Obviously Fama doesn’t believe that, otherwise he wouldn’t believe in the EMH. He also observed the tech bubble in 2000 and the housing bubble in 2006, and doesn’t think those examples refuted the EMH. He argues that at the time investors rationally thought the tech industry might do much better than it actually ended up doing ex post. Obviously Fama is using intrinsic value as people believed at the time, given their information set.”
That’s clearly not what he believed in the 60s and 70s. “Intrinsic value” clearly doesn’t mean, “what investors THINK the actual value of a company is.” If that were the case, then a stock’s market price would always equal its intrinsic value by definition. I don’t understand where all this ambiguity is coming from.
As for the tech bubble and housing bubble, Fama still denies they ever occurred. He believes that bubbles don’t occur. The definition of a bubble is that market prices depart from intrinsic values. It is the exact opposite of what the EMH predicts.
Open any recent finance textbook. How does it define a bubble? “A situation where observed prices soar far higher than fundamentals and rational analysis would suggest… Investment bubbles eventually pop because they are not based on fundamental values”
Now compare to Fama (1965). “If security exchanges are ‘efficient’ markets, then stock prices at any point in time will represent good estimates of intrinsic or fundamental values.”
Clearly we are arguing over definitions here. If you believe that there was a tech bubble, then you don’t believe the market was efficient. At least not as market efficiency was defined in 1965.
30. January 2010 at 19:42
Jimmy, I disagree at a very basic level. You can have RE and also have “bubbles.”
Example from experimental economics:
Vernon, iirc, has done experiments where he had a virtual stock, and an option for the stock. In order to build RE into the experiment, he told the participants the price of the stock in the future. (he could control the price.) When the price of the stock was scheduled to drop to 0, people were still trading the option at a price above that and still making money off of it. The only people who lost money were the ones holding the option when the music stopped.
There are other examples from the real world:
When a large company announces bankruptcy, the price of its stock doesn’t drop to zero, even when they announce that the stockholders will get nothing. The market responds to this shock, mostly rationally and people still end up offloading their stocks to people who are most willing to take the risks doing noise trading on it.
Money and exchange are ways of conveying information and markets more or less efficiently rout the damage, to where it will do the least harm.
30. January 2010 at 19:43
Sigh, Jimmy, there is no such thing as “intrinsic value.”
31. January 2010 at 07:27
Jimmy: You said;
That’s clearly not what he believed in the 60s and 70s. “Intrinsic value” clearly doesn’t mean, “what investors THINK the actual value of a company is.” If that were the case, then a stock’s market price would always equal its intrinsic value by definition. I don’t understand where all this ambiguity is coming from.
As for the tech bubble and housing bubble, Fama still denies they ever occurred. He believes that bubbles don’t occur. The definition of a bubble is that market prices depart from intrinsic values. It is the exact opposite of what the EMH predicts.
Open any recent finance textbook. How does it define a bubble? “A situation where observed prices soar far higher than fundamentals and rational analysis would suggest… Investment bubbles eventually pop because they are not based on fundamental values”
Now compare to Fama (1965). “If security exchanges are ‘efficient’ markets, then stock prices at any point in time will represent good estimates of intrinsic or fundamental values.”
Sorry, but I strongly disagree with what you say here. I am almost certain that Fama would interpret “fundamentals and rational analysis” completely differently that you would.
In the first few decades of the biotech industry, almost all companies lost money. Only a few made profits. The P/Es for the entire industry were negative numbers. Now please tell me what “rational analysis” says these companies should have been worth in 1990 or 1995. If you use simplisitic P/E ratio formulas, then they’d be worth nothing. But if you have a forward-looking approach that considered likely growth in biotech earnings, then the companies would have some value. Ex post people will almost certainly have been too optimistic or too pessimistic about biotech (I think they were too optimistic), because it is so hard to predict new scientific discoveries. But the process of setting prices might well have been completely rational even if ex post a different figure would have seemed appropriate. I am almost certain that this is what Fama is saying about the tech bubble.
Fama may be wrong, but I think I do understand what he is claiming about the world.
31. January 2010 at 19:45
Surely your last comment about Soros was a joke. He’s talked a lot in many places about his theory of reflexivity in which prices end up influencing fundamentals which influences prices and causes a chain reaction pushing things away from equilibrium.
1. February 2010 at 07:14
Jeff, Any information he let’s out has zero market value. I don’t have to pay to know it, if it is in the public domain.
I was half-joking. If someone really did find a market anomaly, they’d be better off keeping it to themself, as releasing the information would make the anomaly go away. I doubt his theory of investing will work in the future. There are two reason why:
1. Either the model was flawed to begin with, or . . .
2. It was correct but won’t work once other investors know about it.
1. February 2010 at 11:20
1. Soros might be trying to trade some of his financial success for academic respectability. The trade might not be working, but the assumption shouldn’t necessarily be that he is giving out information that has no value.
Here are two well known examples of two respected investors providing useful information to the market: Einhorn’s Lehman speech and the Enron research that Chanos did on Enron both contained information that definitely wasn’t priced in. Listening to information from these guys who have demonstrated their ability to make money is different from listening to an average investment bank analyst, who are if anything wrong about stocks on average.
2. Do you really believe the market is as efficient as you sketched out in your response above?
There is enough evidence out there about momentum working or stocks with high short interest under performing other stocks that this view no longer seems defensible without extreme nuances.
2. February 2010 at 20:29
“Sorry, but I strongly disagree with what you say here. I am almost certain that Fama would interpret “fundamentals and rational analysis” completely differently that you would.”
You’re again injecting ambiguity where I think there is none. For me, the definition of intrinsic value is quite simple. It is the present value of all future cash flows discounted at the rate the market sets based on the probability of receiving that cash flow. Obviously if Intel’s corporate bonds start selling at a lower yield than U.S. Treasury’s, the market is not efficient.
“In the first few decades of the biotech industry, almost all companies lost money. Only a few made profits. The P/Es for the entire industry were negative numbers. Now please tell me what “rational analysis” says these companies should have been worth in 1990 or 1995. If you use simplisitic P/E ratio formulas, then they’d be worth nothing. But if you have a forward-looking approach that considered likely growth in biotech earnings, then the companies would have some value. Ex post people will almost certainly have been too optimistic or too pessimistic about biotech (I think they were too optimistic), because it is so hard to predict new scientific discoveries. But the process of setting prices might well have been completely rational even if ex post a different figure would have seemed appropriate. I am almost certain that this is what Fama is saying about the tech bubble.
Fama may be wrong, but I think I do understand what he is claiming about the world.”
Now we’re talking about 2 different problems. It is difficult for the market to price improbable events with large payoffs or losses. I agree with you. The market is less efficient at pricing biotechs than it is at pricing Intel. I’m not sure what this is supposed to prove?
Asset prices were not wandering about their intrinsic values during the housing bubble or the tech bubble. This is clearly not what Fama predicted in 1965.
3. February 2010 at 12:19
Jeff, Obviously anything is possible in the Soros case, and I certainly agree he cares more about reputation than money at this point. But if everyone believes him then his ideas won’t work, so his reputation will get trashed.
I think markets are approximately efficent. I believe that mutual funds that invest on the basis of studies that find flaws in the EMH, will not consistently outperform index funds. Does that help?
Jimmy, You said,
“You’re again injecting ambiguity where I think there is none. For me, the definition of intrinsic value is quite simple. It is the present value of all future cash flows discounted at the rate the market sets based on the probability of receiving that cash flow. Obviously if Intel’s corporate bonds start selling at a lower yield than U.S. Treasury’s, the market is not efficient.”
If you mean ex post, then how is the market supposed to know this? Obviously this is not what Fama has in mind. He would never deny that efficient markets may misprice assets relative to their ex post performance. In 1985 Microsoft was grossly mispriced relative to its ex post cash flow. Maybe 100-fold. What does that prove? If it is ex ante, then I don’t disagree.
You said;
“Now we’re talking about 2 different problems. It is difficult for the market to price improbable events with large payoffs or losses. I agree with you. The market is less efficient at pricing biotechs than it is at pricing Intel. I’m not sure what this is supposed to prove?”
I would say they are just as efficiently priced as any other asset. There is simply more uncertaintly. Inefficiency implies the price is wrong, ex ante. And I see no evidence that biotech prices were wrong ex ante.
I still think you are confusing ex ante and ex post, which is what most critics of the EMH do. They simply point to a market where the price was wrong, ex post, and say “see, markets aren’t efficient.”
4. February 2010 at 14:59
And I think that you’re distinction between ex ante and ex post is a made up one, and one that is not falsifiable. If what you claim the EMH says is actually what it says, people would be making very little fuss over it. It would be definitionally true, like claiming A=A. There would be no bubble in the world large enough to disprove the EMH. All you’re claiming is, “you don’t know when a bubble will pop, therefore markets are efficient and asset prices are wandering about their intrinsic values.” As Larry Summers pointed out in 1985, the first does not prove the second.
6. February 2010 at 07:04
Jimmy, If you view was right, the EMH would be wrong almost by definition. Obviously there are lots of unexplained movements in asset prices. Surely that doesn’t disprove the EMH. I have read some articles by Fama, and he makes the same arguments that I do about 2000, that the market thought the internet companies would do better than they did. He doesn’t see that as a problem for the EMH.