By Rob Bennett
Valuation-Informed Indexing #379
Buy-and-Hold would be the ideal investing strategy if stock returns played out in the form of a random walk. They do not. The last 36 years of peer-reviewed research in this field shows that returns play out in a hill-and-valley pattern that usually lasts roughly 35 years and in which the first 20 years of returns show gradually increasing valuation levels and the last 15 years of returns show generally low valuation levels. The most likely explanation for the hill-and-valley pattern is that price changes are caused by shifts in investor emotion that push prices upward for about 20 years and that then cause low prices when investors' natural irrational exuberance is popped by the economic realities and transformed into an irrational depression which must be exhausted by the passage of time before a new price climb can begin.
Valuation-Informed Indexing makes use of the 1981 research findings of Robert Shiller showing that today's P/E10 level can be used to predict the return that will apply 10 years out, 15 years out and 20 years out (but not one or two or three years out - all of the evidence available to us today indicates that the Buy-and-Holders are right in their claim that short-term timing doesn't work). If it is true that valuations affect long-term returns, then stock investing risk is a variable thing rather than a static thing. Investors seeking to maintain the same risk profile at all times must be willing to adjust their stock allocations upward at times of low valuations and downward at times of high valuations. Valuation-Informed Indexers believe that investors should aim to keep their risk profiles constant, not their stock allocations.
The current hill-and-valley pattern is the longest of the four that exist in the historical record (which dates back to 1870). Its bull portion of the cycle began in 1982, when the P/E10 level stood at 8, half of the fair-value P/E10 level of 15. It ends in 2000, when the P/E10 level hit 44, by far the highest P/E10 level on record (the P/E10 level just prior to the onset of the Great Depression was 33). The P/E10 level has been gradually dropping since 2000 but still stands today in the low 30s. So we would need to see a price drop of 50 percent for stocks to return to fair-value price levels. If prices were to drop to one-half of fair value, as they have before the earlier three bull/bear cycles resolved themselves, we would of course see a price drop of a good bit more than 50 percent from today's levels before steady upward movement in the P/E10 level could be achieved again.
So we are now 36 years into the current cycle. Since the downward movement in P/E10 levels has not completed itself, we cannot make final statements as to the results achieved by investors who began investing at different points in the cycle. But we can draw tentative conclusions, either being satisfied with noting how investors have fared thus far into the cycle or speculating as to how they will end up doing, presuming that a 50 percent price drop will be seen in the next year or two or three.
If you invested a lump sum in stocks at the beginning of the bull/bear cycle (1982), you earned an annualized return of 15 percent real by the time the market hit the top of hill portion of the hill-and-valley pattern in 2000. That result supports the Valuation-Informed Indexing principle that investors should go with higher stock allocations when prices are insanely low. When the P/E10 level is 15, the market never produces a poor long-term return; the only possibilities are good returns, great returns and amazing returns. I would call a 15 percent annualized return "amazing." Earning that return for 18 years running would have turned a $100,000 lump sum into nearly $1.2 million. Investors putting their money into stocks at those sorts of time periods enjoy lots of early retirements!
Buy-and-Holders would argue that there is no way to know when those magical time periods have come to an end and that investors who are stuck with stocks from 1982 until today have done just fine. The annualized return for those 36 years is 8.9 percent real. That is indeed a super return. But I cannot help pointing out that most of that super return came from the first half of the 36-year time-period, the half that began with a P/E10 level of one-half of the fair-value P/E10 level. The annualized return from January 2000 forward was only 3.3 percent real. And, in the event that we see a 50 percent price drop in the near future, that would bring the portfolio amount that started at $100,000 a bit below the $1.2 million figure where it stood at the top of the bubble. Stocks can provide good returns even starting from high price levels. But the downside of the investment choice - which virtually disappears at times of super low prices - should become a serious concern at times of sky-high price levels.
There was much talk in the wake of the 2008 crash as to whether middle-class investors should remain in stocks. Shiller advised at the time that investors resume buying stocks only after the P/E10 level dropped below 10 (which it never did - prices started moving up again soon after we touched a P/E10 level of 13). Buy-and-Holders point with a sense of vindication to the numbers showing that investors who bought stocks in March 2009, the turning point of the crash, have earned an annualized return of 16 percent real in the years since. That would have turned $100,000 invested in 2009 into $360,000.
Even a 50 percent crash taking place tomorrow wouldn't turn that into a bad investment decision. The $360,000 would be reduced to $180,000, which is an 80 percent increase on the original investment amount. That's an annual return of about 9 percent per year over the nine-year time-period. Not bad at all! But, again, valuation levels were promising in March 2009. Investors who joined the party at later and more expensive entry points would not fare nearly as well in the event that we see a drop to fair-value price levels in the near future.