I believe that the case for Valuation-Informed Indexing, the model for understanding how stock investing works that posits that investors must change their stock allocations in response to big valuation shifts to keep their risk profile roughly constant over time, is superior to Buy-and-Hold, the model that posits that no form of market timing is a good idea. The difference between the two models is that Buy-and-Holders believe that stock price changes are caused by economic developments and thus are rooted in something real and lasting while Valuation-Informed Indexers believe that stock price changes that cause overvaluation are caused by irrational exuberance and thus are rooted only in ever-shifting investor emotions.
How strong is the case? I see it as very strong indeed. Stock prices rose by 126 percent from 1996 through 1999. It is impossible for me to believe that the value of the underlying companies rose by that amount in that short amount of time. It is equally impossible for me to credit the sharp price drop that we saw in late 2008 as the product of a rational assessment of the economic realities. If it were that, we would not have seen the quick reversal of that price drop that we saw in the Summer of 2009. Prices came down too quickly to be attributable to a rational assessment of economic realities and then rose too quickly to be attributable to a revised assessment of economic realities. If it were the discovery of economic realities that drove price changes, price changes would take place more gradually and less erratically.
Investment strategies used by hedge funds have evolved over the years, although the biggest changes have come in the use of computers to develop portfolios. Rosetta Analytics is a woman-founded and woman-led CTA that's pioneering the use of artificial intelligence and deep reinforcement learning to build and manage alternative investment strategies for institutional and private Read More
I see the case as being so strong that I have said on a few occasions that all of the evidence available to us today supports Valuation-Informed Indexing and that none of it supports Buy-and-Hold. I still generally believe that that is so. But there is one reality that causes me to question the grand sweep of that claim. Stock prices have remained at very high levels for almost the entire time-period from 1996 through today. That’s 23 years. There is no earlier time in U.S. history when stock prices have remained that high for that long. We are travelling in uncharted waters.
That should cause those who believe in Valuation-Informed Indexing to experience at least some doubts. The idea behind the strategy is that stock gains caused by irrational exuberance will disappear in time. Thus, they cannot be counted on to finance a retirement. When prices get very high, an investor is better off moving a portion of his assets to an asset class paying a smaller return to protect himself from the devastating price drop that in the past has always followed from times when prices rose to very high levels. The risk of the strategy is that, if the price drop does not come for a long time, the investor misses out on the higher gains that he would have enjoyed by being fully invested in stocks during that time. Valuation-Informed Indexers don’t pretend to be able to identify when a price drop will come. But we believe that stock allocation changes make sense when prices get so high that even if the price drop does not come for some time the net effect will be positive for the investor who adjusts his stock allocation.
The historical return data shows that, once the P/E10 level reaches 25, the price drop always arrives within 10 years. That’s why Robert Shiller predicted in Julyb1996 that investors who elected to maintain their high stock allocations would live to regret it within 10 years. Shiller did not know when he made that prediction that the P/E10 level would rise to its highest level in history by the end of 1999. So I don’t find it at all surprising that the price drop did not arrive until 12 years after Shiller made his prediction. My assessment is that Shiller got the big thing right by predicting the economic crisis while being proven wrong about a relatively technical point (the precise year when the price drop would occur).
Had prices remained at fair-value levels or lower following the big drop experienced in late 2008, I would say that the Valuation-Informed Indexing concept had passed an important test with flying colors. But that is of course not what happened. The P/E10 level hit 13 (only a bit below fair-value price levels) for a brief time in February 2009, then headed upward. By the end of 2009, we were at high price levels again. And prices have remained high until today. We have not yet seen prices drop to the rock-bottom levels that we have seen at the end of every earlier bull/bear cycle.
I am able to offer four explanations for why the usual pattern of price changes has not applied.
One, it can take 40 years for a bull/bear cycle to complete itself. Thus, we only have seen four cycles in the time-period (from 1870 forward) for which we have good records of prices. It could be that we are living through an outlier result, that if we had good records of stock prices covering a much longer time-period, the pattern we are living through today would not seem so odd.
Two, it could be that the fact that we hit the highest P/E10 level on record (the 44 that applied in January 2000) made high prices a bit more sticky this time around and thus it is taking a bit longer for them to come down to below-fair-value levels.
Three, this is the first time in which investors who believe that high prices are not a concern have peer-reviewed research (the research done by Eugene Fama purporting to show that the market is efficient and that thus high prices are the product of rational investor choices) supporting their viewpoints. Investors who believe that high prices are the product of rationality might be slower to lose confidence in them and pull them downward.
Four, there is evidence that the Federal Reserve has taken steps to keep stock prices from falling too hard too soon out of a concern that sharp price drops might do serious damage to the economy.
I find those four explanations all added together somewhat but not entirely convincing. I find it exceedingly strange that prices have remained so high for so long. I still believe in Valuation-Informed Indexing. My confidence in the concept remains strong for many reasons. But I am no longer going to say that I believe that all of the evidence available to us supports the concept. I think that the length of time for which prices have remained very high justifies at least some doubt in the minds of reasonable people.
Rob’s bio is here.