Valuation-Informed Indexing #353
By Rob Bennett
Robert Shiller’s Irrational Exuberance is the most important book ever published on stock investing, in my assessment. That said, the book is marred by one glaring weakness — it offers zero guidance on how investors who believe in its theory of how the stock market works should invest their money. Readers of the book certainly get the point that valuations matter a great deal. But how much should investors lower their stock allocations when prices reach the levels where they have remained for most of the time since the book was published? Don’t read Irrational Exuberance to find the answers. You won’t find any contained within its pages.
The first time that I heard Shiller direct himself to this all-important question was in the days following the 2008 price crash. The P/E10 value had dropped to 13 (a bit below fair-value levels) and most Buy-and-Holders were asserting that stocks were now “on sale” and that investors should be taking advantage of the great opportunity being presented to them. Shiller disagreed in an interview published at the time. He advised investors to remain out of stocks until the P/E10 level dropped below 10.
I posted a podcast at the time in which I analyzed Shiller’s comments. I had both good and bad things to say. I was happy to hear someone not repeat the conventional wisdom of the time that stocks were now “on sale.” Prices were obviously a lot better than they had been for a long time; the P/E10 level had reached 44 in January 2000 and had remained above 20 for many years. But it simply was not true that stocks had become an amazing bargain, except in relative terms. The fair-value P/E10 level is 15. So, yes, stocks were slightly on sale. But I felt that it was dangerous to encourage investors to buy too heavily. What if prices fell further still? Investors who had bought heavily out of a belief that there was a big sale going on would likely panic in response to a further drop and even further destabilize our then-troubled economic system by selling heavily.
I did say in all of the podcasts that I recorded in those days that stocks offer a strong long-term value proposition when the P/E10 value is 13. So in an important sense I agreed with my Buy-and-Hold friends. But I included caveats in my statements noting that it was entirely possible that prices would continue downward for some time; there has never in U.S. history been a secular bear market that ended before the P/E10 level fell to 8 or lower. I was happy to hear Shiller tell the other side of the story, that, while prices were good for the first time in a long time, there was a good chance that prices would continue to fall or at least not rise for some time and investors thinking of buying stocks needed to take that possibility into account before putting money on the table.
What made me uneasy with Shiller’s comments was that he didn’t make any more effort than the Buy-and-Holders did to note that there was more than one way that things could go. It was certainly true that prices could continue to fall and my personal assessment was that that was the most likely scenario. But chiller's work and the work of all of the researchers who have followed in his footsteps shows not only that long-term timing always works but also that short-term timing never does. Neither Shiller nor I nor my Buy-and-Hold friends nor anyone else could say for sure which way prices were headed. To offer truly useful advice, we needed to make that clear. I didn’t think that Shiller had done that and I said so in the podcast that I recorded at the time.
Shiller acknowledged the mistake in an interview he gave in more recent months. Shiller said in the recent interview that the P/E10 metric cannot be used for market timing, noting that he had once made the mistake of thinking that it could. I can understand why Shiller would feel that he was “burned” by P/E10 in his earlier comments and why he would be unwilling to make the same kind of mistake again. But I find his new comments highly unfortunate.
It is true that P/E10 should never be used for short-term timing. The entire historical record shows that it cannot be used effectively for that purpose. When someone uses the term “market timing” it is heard by almost all of his listeners to be a reference to short-term market timing. So in a literal sense what Shiller was saying was so -- P/E10 really cannot be used for market timing as it had been practiced for the many years before Shiller came on the scene.
But to suggest that P/E10 cannot be used effectively for long-term timing is to throw doubt on the value of Shiller’s “revolutionary” (his word) finding that valuations affect long-term returns. To say that valuations affect long-term returns is to say that valuations matter. And to say that valuations matter is to say that investors who adjust their stock allocations in response to big valuations are likely to be rewarded for doing so somewhere down the road. Those investors are practicing a form of market timing! The form that always works! It’s painful to hear the fellow who made this world-changing discovery advance words that are sure to add to the confusion that already plagues discussions of the far-reaching implications of his Nobel-prize-winning research.
I think that Shiller is like most of the rest of us. I think he senses that investors need to take valuations into consideration when making important strategic choices. But I think that, like all the rest of us, Shiller learns by talking things over with lots of other smart people and that his understanding of the implications of his own research findings has been hindered by the lack of debate that has been heard on these questions in the first 36 years since publication of his research. I think that Shiller took a positive step by offering a contrary view back in early 2009 and that, given the big jump in prices that we saw following his advice for investors to stay out of stocks, he has become reluctant to offer unconventional views again.
Making mistakes is part of the learning process. We should not be penalizing people in this field for making mistakes because, when we do penalize them, they become reluctant to stick their necks out in the future. The primary reason why the learning process in this field proceeds so slowly is that we all see it as so important to avoid mistakes that we hold back from taking the chances we need to take for the learning process to move forward at a quicker pace.
Rob’s bio is here.