Deflating a bubble before it busts would be a strenuous business
Economists may be finally closing in on the reason for asset bubbles. How to pop them before they grow too large, however, is a much harder problem.
The study of bubbles has steadily gathered urgency during the past four decades as crashes became more spectacular and more damaging. The stock crash of 1987 was a wake-up call for those who had assumed that markets function efficiently; there was no obvious reason why rational investors would suddenly conclude that US companies were worth 23 per cent less than the day before. The tech bubble was even more troubling because many observers had warned of a bubble for years before the crash, to no avail. The same thing happened with the housing bubble, only when that one burst it took the real economy with it - as tends to happen when rapid asset-price declines are combined with high levels of debt.
Researchers have developed a large and diverse array of theories about why asset prices suddenly rise and crash. The difficulty in identifying the cause of bubbles has made it hard to design policies to prevent them. But recently, a growing number of economists are zeroing in on the idea of what they call extrapolative expectations. For whatever reason, it seems, investors sometimes decide that a recent run of good returns represents some sort of deep structural trend.
A new paper by Zhenyu Gao, Michael Sockin and Wei Xiong supports this idea. Looking at housing prices during the bubble, they compared states based on how much they changed their capital gains tax rates. The states that raised taxes more saw less of a bubble and crash.
But even more tellingly, the authors found that in states with lower capital gains taxes, purchases of investment properties that the owners didn’t plan to live in tend to increase more in response to recent price gains. The simplest explanation for this pattern is that when prices go up and there aren’t tax hikes to make people expect a reversal, they start to think that the trend will continue for the foreseeable future. And they buy accordingly.
A recent study by researchers at the Bank of Canada provided different evidence that pointed in the same direction. In surveys, they found, most people expect that house prices will rise and fall during a five-year period. But when the researchers show people data on recent prices, they start to believe that current trends will continue for the next five years.
Other papers found the same pattern. Economists Theresa Kuchler and Basit Zafar found that in areas where local housing prices increased more, people tend to expect that prices will go up more nationwide in the years to come. And researchers Zhi Da, Xing Huang and Lawrence Jin found that extrapolative expectations are in full effect on crowdsourcing websites.
Obviously, investors don’t always expect recent trends to continue forever. So what makes them start to extrapolate? It may be that when a price trend gets big enough for people to notice, human psychology tends to assume that this is the new normal. If house prices in your neighbourhood normally go up and down, but then they suddenly start going up by big amounts each year for several years, you might assume that the game has simply changed. In that case, the obvious thing to do is to buy, buy, buy.
Defenders of the idea of efficient markets might retort that if this happens, savvier investors would simply bet against the trend-followers and bring things back into line. But economic theorists have long understood that because rational investors have limited firepower to short a bubble, they often find it more worthwhile to ride the rising prices for a while. That just makes the bubble worse.
So economists are starting to get a picture of why bubbles happen. Some event - maybe an increased supply of credit to home buyers - increases demand for houses or stocks and pushes up prices for a few years. Speculators see that and decide that prices just go up now, so they buy in, increasing prices even more. Rational investors know there’s a bubble, but decide to buy in for a while and try to sell at the top. The crash only happens when speculators run out of money to keep buying.
If this unified theory of bubbles turns out to be right, the next question becomes: How can governments nip the process in the bud? Gao and his colleagues suggest that tax hikes could be an answer. If economists can figure out how many years of rising prices - say, three or five - is needed to create the impression of a permanent trend in speculators’ minds, then policy makers could adopt a rule where capital gains taxes temporarily go up whenever prices rise for that many years in a row.