Valuation-Informed Indexing #261
by Rob Bennett
The core idea behind the Valuation-Informed Indexing concept is that long-term returns are highly predictable. When valuations are low, long-term returns are amazing. When valuations are moderate, long-term returns are good. When long-term returns are high, long-term returns are poor.
Predictable returns — that’s investor heaven! So why all the controversy? Why is there so much resistance to buying into this research-backed strategy?
The problem is that short-term return sequences are not at all predictable. And investors get to see short-term market reactions, of which there are many, far more often than they get to see long-term market reactions, of which there are few. Asking investors to believe that returns are highly predictable is asking them to believe in something that they have seen disproven again and again and again.
Consider five cases in which seemingly reasonable predictions were made that were subsequently shown not to play out in the real world:
1) Many people predicted disaster following the 1987 price crash. The Buy-and-Holders advocated staying the course. Prices recovered quickly and in years to come performed many dramatic leaps upward.
2) Jack Bogle advised investor caution in the early 1990s, when prices rose to levels that seemed dubious to most informed observers. Prices continued rising for a long time and Bogle came to doubt even more than he did before his ability to see into the market’s future.
3) Robert Shiller predicted in testimony to the Federal Reserve given in the Fall of 1996 that investors who remained heavily invested in stocks would come to regret it within 10 years. Valuations remained high in 2006. Few Buy-and-Holders felt that they had made a mistake sticking with their high stock allocations.
4) Shiller advised investors in the wake of the 2008 crash that it would not be safe to re-enter the market until the P/E10 level dropped below 10. The P/E10 level shot up to 20 a few months later and has never since returned to anywhere close to the level where it stood when Shiller made his prediction.
5) I predicted a few years back that stock prices would fall by 65 percent by the close of 2016. As that deadline has come closer, my critics have demanded to know whether I will change the prediction if January 2017 comes without us having experienced a price crash. I respond that I won’t have much choice. My critics laugh at me re that one.
We cannot predict the short term. The evidence for that is strong. We want to be able to predict the short term. Anyone who could do it could get very rich and very popular very quickly. But smart people fail at this over and over again. All signs are that it is a fool’s errand to try to predict short-term return sequences.
So how can I be so confident that long-term return predictions work?
The belief is a bit counter-intuitive. But, if you work carefully through the logic chain, I think you may agree that it makes sense to believe that long-term returns are predictable for the same reason that short-term return sequences are not.
The Buy-and-Holders believe that investors act rationally to ensure that mispricings are quickly exploited for profit and thereby eliminated. The Buy-and-Hold Model hangs together. Unfortunately, it cannot be reconciled with Shiller’s finding of a strong correlation between today’s P/E10 level and the market price that applies in 10 years. If price changes were caused by unforeseen economic developments, as the Buy-and-Holders theorize, there should be no correlation between today’s P/E10 level and the price that applies 10 years out. Yet there has been a strong correlation for 145 years now. That’s as far back as we have records.
So the Buy-and-Holders are wrong.
If price changes are caused by investor emotion (with economic developments playing a secondary role because economic developments influence shifts in investor emotion), Shiller’s finding makes sense.
But why isn’t there predictability in the short term as well as in the long term?
There is a force causing the market to set prices at fair-value levels in the long term. That is what markets do. The purpose of a market is to set prices properly. Long-term timing always works because the market couldn’t continue to function if it did not.
That force does not apply in the short term. In the short term, human emotion can set prices wherever it wants to set them.
Bogle was right to see that prices had gotten a bit out of hand in the early 1990s. But he was wrong to think that the market would necessarily take action to set things right in one or two or three years.
Shiller was right to anticipate in 1996 that investors going with high stock allocations would regret them within 10 years but he was wrong to feel as certain as he did that the price crash would arrive by then. The bull that ended in 2000 was the biggest bull in U.S. history. We didn’t see a full-out price crash until September 2008, twelve years from the date of Shiller’s testimony. He got the important thing right but was a bit off re one technical point (the precise timing of the crash).
I think Shiller will eventually be proven right re his prediction that it would not be safe to re-invest in stocks until the P/E10 level dropped below 10. Again, he was right about the important matter and wrong re the technical timing point.The drop to a P/E10 below 10 is still “priced in” to the market today.
I may or may not be proven wrong re my prediction that we will see a crash by the end of 2016. But investors who focus on the timing of the next crash are focused on the wrong thing. it doesn’t matter so much whether the crash comes in 2015, 2016 or 2017. What matters is whether it comes or not. I believe (although I of course cannot be sure) that I will be proven right re that point.
We cannot predict short-term market developments. So we cannot predict with precision even long-term market developments. But we can say that stocks are more risky as prices rise. And the strategic implications of that critical insight stretch in many directions.
Rob Bennett’s bio is here.