by Rob Bennett
Wade Pfau, Associate Professor of Economics at the National Graduate Institute for Policy Studies in Tokyo, Japan, has posted preliminary research at his web site showing that Valuation-Informed Indexing provides more wealth for 102 of the 110 rolling 30-year periods in the historical record while Buy-and-Hold does better in 8 of the periods.
Pfau’s research is breakthrough stuff, in my assessment. That said, I believe that his research greatly understates the benefits of Valuation-Informed Indexing (changing your stock allocation in response to big swings in valuations). There are six reasons.
1) Pfau’s research assumes an unlikely scenario -- that Buy-and-Holders will hold to their high stock allocations in the face of huge losses.
We have seen three secular bull market in the historical record. The average price drop that followed (from the high point to the low point) was 68 percent. The average 20-year return following the bull top was 0.7 percent (including dividends). It is exceedingly unlikely that more than a tiny percentage of Buy-and-Holders stuck with their high stock allocations through so many years of financial pain and yet any investors who sold obtained worse results than the results than caused Buy-and-Hold to come up short in 102 of 110 of the 30-year periods in the historical record.
Valuation-Informed Indexers do not experience the same difficulty sticking to their planned stock allocations because Valuation-Informed Indexers go with far lower stock allocations at times when stocks are priced to crash (times of high valuations). So failing to consider the reality that most Buy-and-Holders are forced to sell when prices crash slants things in favor of Buy-and-Hold.
2) The compounding returns phenomenon will continue to enhance the edge enjoyed by Valuation-Informed Indexers beyond the 30-year period examined by Pfau.
I have run hundreds of comparisons of Valuation-Informed Indexing and Buy-and-Hold using a portfolio allocation calculator (“The Investor’s Scenario Surfer”) that generates realistic 30-year returns sequences and permits the user to shift his stock allocations on a year-by-year basis in response to the changes in P/E10 levels that happen to turn up. By watching Valuation-Informed Indexing prevail in roughly 90 percent of the tests, I have come to understand the two drivers of its long-term superiority: (1) taking valuations into consideration virtually assures that the Valuation-Informed Indexer will sooner or later go ahead; and (2) once the Valuation-Informed Indexer goes ahead, it becomes almost impossible for the Buy-and-Holder to catch up because both types of investors are heavily invested in stocks at times when the long-term value proposition of owning stocks is good (time of low and moderate valuations).
An investor who adopts a Valuation-Informed Indexing strategy at age 25 will indeed almost certainly be ahead at age 55, as Pfau’s research indicates. But the financial edge he will possess at that time (which can be huge -- I have seen returns sequences where the Valuation-Informed Indexer has a portfolio double the size of the Buy-and-Hold portfolio at the end of 30 years) will just keeping growing beyond the 30-year period examined in this research. An investor who becomes a Valuation-Informed Indexer at age 25 and dies at age 85 might enjoy 60 years of compounding on the advantage he obtains by taking valuations into account. In fact, an investor who begins at an early age taking valuations into consideration when setting his stock allocation might be able to gain a second edge by living through a second bull-bear cycle and then enjoy years of compounding on the differential obtained via the second edge.
3) The higher returns obtained by the Valuation-Informed Indexers are obtained at greatly reduced risk.
It’s not entirely fair to compare the results obtained with a low-risk strategy with those obtained from a high-risk strategy. Take the 8 cases in which Buy-and-Hold generated slightly better numbers. Is it right to say that Buy-and-Hold ”worked” in those eight cases? I say “no.” There is no way to know in advance when one of the rare cases in which Buy-and-Hold generates good numbers is going to turn up. With the odds that any given time-period will be one of those cases being less than 1 in 10, the investor is taking a long odds bet. It is proper to say that Buy-and-Hold on rare occasions produces better numbers. But I think it needs to be added that it does so at greater risk. On a risk-adjusted basis, Valuation-Informed Indexing is always superior.
4) Valuation-Informed Indexers are able to obtain better returns on their non-stock investments.
Pfau assumes that Valuation-Informed Indexers and Buy-and-Holders will obtain equal returns on their non-stock investments (he uses the returns earned on Treasury bills for the non-stock money and intends also to test what would happen if the money were instead placed in bonds). Treasury Inflation-Protected Securities were paying 4 percent real in early 2000. Valuation-Informed Indexers were able to lock in that return because they were at the time looking for alternatives to high-priced stocks. But Buy-and-Holders disdained the 4 percent real return on TIPS as not being comparable to the 6.5 percent real average return for stocks (which Buy-and-Holders believe to be the most likely return starting from any valuation levels).