Shiller’s Record at Short-Term-Timing Predictions Is Poor

Updated on

Valuation-Informed Indexing #218

by Rob Bennett

Last week’s column examined public statements in which Yale Economics Professor Robert Shiller showed that he believes in short-term timing. I wish he didn’t. Shiller’s breakthrough finding was showing the critical distinction between short-term timing and long-term timing — short-term timing really doesn’t work (as Fama showed) but long-term timing is price discipline and price discipline always works and is always 100 percent required for investors seeking to have some realistic hope of long-term investing success. When Shiller advocates short-term timing, he blurs this critical distinction in people’s mind and thereby undercuts the case for long-term timing and indeed for the value of his own research.

This week I will look at Shiller’s track record. Most interviews with Shiller focus on his short-term-timing predictions rather than on the important implications of his research re how investors should be changing their stock allocations in response to big price swings as a means of keeping their risk profiles roughly constant. The reality is that Shiller possesses about the same ability to predict where the market is headed in the short term as everyone else — his track record is poor. We all should make an effort to give up our preoccupation with what doesn’t work (short-term timing) and move on to the truly exciting finding of Shiller’s 1981 research — his finding that long-term timing (changing your stock allocation because of a big shift in valuations with an understanding that you may not see benefits for doing so for as long as 10 years) always works.

The short-term-timing prediction for which Shiller is most famous is one which he never even made. The idea that he successfully predicted the tech crash of 2000 is a myth. Shiller’s book was published in March 2000. But it was pure coincidence that the book happened to arrive at the precise time that tech stock prices were tumbling. The book described a research project that extended back two decades in time. There is no evidence that Shiller aimed to have the book come out at the time tech stocks would be crashing. Nowhere in the book or anywhere else did he say that he believed that there would be a crash in tech stocks in early 2000. It was pure coincidence that the tech crash and the publication of Shiller’s book happened to take place at the same time. Those saying that Shiller demonstrated an ability to predict short-term price changes because his book happened to come out in March 2000 are off base.

Even if Shiller had deliberately published the book in March 2000, that would not have shown an ability to predict short-term price changes unless he had specified that he foresaw only a crash in tech stock prices. The larger market did not collapse until September 2008. Shiller DID predict the crash of the larger market (he talked of a crash that would produce losses equal in size to the destruction of “all the homes in the country” that would take place “over some interval in the first decade or so of the 21st Century”). The long-term prediction was accurate and of great significance. The short-term prediction for which Shiller has become famous never happened and would have been a failed prediction if it had.

Shiller predicted a stock crash this year in interviews he gave late last year and early this year. That prediction has not yet failed as there are two months remaining in the year. But time is running out. And Shiller said in a recent interview that he does not believe that the time has yet come for investors to lower their stock allocations. Given that the crash needs to come within two months for Shiller’s prediction to turn out correct and that Shiller himself does not view a crash as imminent today, I think it would be fair to say that there is a good chance that that prediction too will be proven a failure and that, even if it is not, it will not be because Shiller possesses an ability to make sound short-term predictions. If he possessed that ability, he would be telling us today that a crash is coming soon.

Shiller made another failed short-term prediction in the early months of 2009. He said at that time that investors should not feel safe getting back into stocks until the P/E10 level dropped below 10. In the years since, we have not yet seen the P/E10 level drop below 10 and Shiller has said that he himself is now holding 50 percent stocks. I think he is right that we will see a P/E10 value below 10 before this secular bear market comes to an end. But I think he was mistaken to express an opinion in 2009 that we were likely to see that day within a few months (and I said so in a podcast that I recorded at the time). As a long-term prediction, the claim that we would be going below 10 was sound. As a short-term prediction the claim  that we would go be going below 10 was a mistake.

Shiller made an important prediction in his 1996 testimony to the Federal Reserve (Federal Reserve Chairman Alan Greenspan came up with the term “irrational exuberance” as a result of this testimony). Shiller said that investors who were highly invested in stocks in 1996 would live to regret that choice within 10 years. That’s a long-term prediction. I think that Shiller was more right than wrong re that one. As it turned out, investors who remained highly invested in stocks in 1996 despite the insanely high valuations that applied at the time came to regret their choice by the end of 2008 rather than by the end of 2006. Shiller was off by a little and it was his effort to be too precise that caused the error. Had he said that investors who were heavily invested in stocks would regret that decision by the end of the first decade of the 21st Century, he would have made both an accurate and a very important prediction.

Long-term valuations-based predictions of stock market returns always work. Shiller discovered that. It is a discovery of huge importance. Those who interview Shiller should ask questions aimed at eliciting information re how we all can make greater use of our new ability to make effective long-term predictions.

Short-term predictions never work. Not when made by ordinary investors and not when made by extremely knowledgeable investors like Shiller. When Shiller’s short-term predictions fail, he undermines the case for long-term predictions because people note the failure and conclude that Shiller is not better than anyone else at this stock-market prediction business. That’s extremely unfortunate. I wish that inerviewers would stop asking Shiller’s about his beliefs about where stock prices are headed in the short term and I wish that Shiller would stop taking the bait when such questions are posed to him.

Nobody can practice effective short-term timing. Even the guy whose research showed us all that long-term timing always works!

Rob Bennett has recorded a podcast titled Are Middle-Class Investors Too Dumb to Invest Effectively? His bio is here.

Leave a Comment