I’m not a big fan of investment ‘gurus’ that make a habit of warning us against financial market bubbles… all the time. The equity market, the housing market, technology stocks, bond prices. Everything is always a bubble. Yet, many of these gurus are given every opportunity to express their deeply rooted worries about the next bubble, without that bubble ever bursting. Now, I understand that these doom and gloom stories make good headlines, but why is it that so rarely these doomsayers are held accountable for the fact that they have been dead wrong? In some cases for years on end!
Learning From History
Conveniently, a recently published study shows some reservation is warranted when forecasting bubbles. In his study titled ‘Bubble Investing: Learning from History’, William Goetzmann, Professor of Finance at Yale, defines a bubble as ‘a large price decline after a large price increase or, a crash after a boom.’ A boom is then defined as a market value increase of at least 100% (a doubling) within a single year and a crash as a drop of at least 50% in the following year(a halving). Note that a halving implies that you give away all gains realized in the first year.
Mr. Goetzmann’s results, after analyzing more than 3000 equity market returns in 21 stock markets starting in 1900, are pretty straightforward. Bubbles are just rare. The historical probability of a halving in the year following a doubling of the market is 4%. Hence, only one out of every 25 boom years is followed by a crash (halving) in the following year. In addition, and this is at least equally important, the historical probability that the market will rise another 100% or more in the following year is 8%. Yes, that’s right! The chance of a subsequent doubling is twice that of a subsequent halving. So, instead of forecasting booms to burst you would have done much better by forecasting booms to continue.
Now, one argument, much to the delight of doomsayers because it enables them to express their bubble worries for an extended period of time, is that bubbles take time to burst. To follow up on this notion Mr. Goetzmann also looks at a period of five years after a boom occurs. Obviously, as the horizon increases, so does the probability of a crash. But even now the historical probability of a bubble remains relatively small. Just 15% of the years in which the market rises by 100% or more is followed by a halving in the next five years. And, bubble forecaster please take notice, the probability of another doubling (which is 26%) beats the probability of a halving in this case as well.
So, I guess at least some cautiousness should be taken into account before referring to a bubble. Market bubbles are scarce and happen far less often as we might think. Obviously, bursting bubbles come with long-lasting effects and crashes of 50% do tend to stick in your memory. Not in the least because, mostly with hindsight, there is always that obvious reason for why the bubble burst, right? And then there are, of course, the examples in which waiting for the big collapse is rewarded. The most recent example, the bursting of the US housing market bubble that led to the ‘Great Recession’ actually got its own movie, ‘The Big Short’. But, on average booms rarely transform into bubbles.
The Big Short (2015): http://www.imdb.com/title/tt1596363/