Don't be early in bubble predictions
By Scott Sumner
Alex Tabarrok has an interesting new post on trends in housing prices:
In 2005, I thought housing prices were rising above the fundamentals and I said so. In 2008, as the fall in housing prices was well under way, I wrote a blog post and later a NYTimes op-ed saying that the housing price bubble was not nearly as big as people thought. I wrote:
I think that housing prices went beyond the fundamentals sometime around 2004…but 2004 levels are still well above long run trend.
…Prices will probably drop some more but personally I don’t expect to ever again see index values around 110. Do you? If we don’t see the massive drop back to “normal” levels then the run up in prices should be described as a shift to a new equilibrium…[with some overshooting, rather than as a bubble.]
To put it mildly, not everyone agreed with my argument. I certainly got the timing wrong-I didn’t think the recession would be as long or as deep as it was. Nevertheless, some people are coming round to my point of view. Karl Smith, for example, has a new post Was There Ever a Bubble in Housing Prices? which concludes more or less, as I did nearly ten years earlier, that the answer is no. What happened was greater liquidity which made housing prices gyrate more like stock prices but “the fundamental driver isn’t irrational bubble behavior. It is competition over a scarce resource.”
I have frequently argued that bubbles are a myth, or more precisely the “bubble” concept is generally not a very useful tool for understanding the world. In the comment section to Alex’s post, Alan Goldhammer made this remark:
First of all there were some prescient folks who called the ‘bubble’ very early on including Dean Baker and the Calculated Risk duo of Bill McBride and the late Doris Dungy.
At first glance it would seem that an early prediction is better than a late prediction. But in a 2010 post I argued just the opposite—the earlier predictions were less accurate than later predictions, indeed not accurate at all.
When Dean Baker warned of a housing bubble in 2002 and Robert Shiller warned of “irrational exuberance” in the stock market in 1996, they weren’t merely suggesting that bubble conditions might develop in the future, they were claiming that asset prices were already at excessive levels.
In retrospect, those claims don’t seem particularly prescient. Tabarrok’s post shows that housing prices are now above 2002 levels, even in real terms. And of course stocks are now far above 1996 levels. That doesn’t mean they were necessarily wrong; perhaps housing and stocks are in a new bubble. But it does mean that one cannot simply point to these early predictions as being obviously correct bubble calls, as it is by no means obvious that prices were excessively inflated in those two cases.
I often suggest that bubble claims will be increasingly popular during the 21st century. That’s partly because recent bubbles are widely (and wrongly) viewed as having done great damage to the economy, and also because with the “new normal” of ultra low interest rates, asset values will often seem higher than justified by many of the usual metrics.