By Rob Bennett
The Investor’s Scenario Surfer is a portfolio allocation calculator that permits investors to compare the results of three rebalancing strategies (80 percent stocks, 50 percent stocks, and 20 percent stocks) with a Valuation-Informed Indexing strategy (changing your stock allocation in response to big valuation shifts) over the course of a randomly chosen 30-year return sequence consistent with those we have seen in the historical record. Valuation-Informed Indexing strategies show better results in roughly nine out of ten of the tests performed. This result challenges the conventional investing wisdom of recent decades that “timing doesn’t work” and that rebalancing is an effective way for an investor to “Stay the Course” for the long term.
At issue is the fundamental question of whether valuations affect long-term returns. Today’s dominant model for understanding how stock investing works, the Buy-and-Hold Model, is rooted in a belief in the Efficient Market Theory, which posits that stocks are always priced more or less properly. If this is so, both overvaluation and undervaluation are meaningless concepts and investors should indeed avoid both short-term timing (changing your stock allocation with the expectation of seeing a benefit for doing so within 12 months or so) and long-term timing (changing your stock allocation in response to big valuation shifts with the understanding that doing so may not yield benefits for as long as 10 years). Yale Professor Robert Shiller’s research showing that valuations affect long-term returns discredits the Efficient Market Theory and the Buy-and-Hold Model and suggests that long-term timing should produce higher risk-adjusted returns than rebalancing. The Investor’s Scenario Surfer allows investors to view hundreds of hypothetical but realistic returns sequences to see how investing pursuant to Shiller’s findings is likely to play out in the real world.
The first step to using the calculator is to click the “Clear All Entries” button. The user then needs to enter three inputs to have the calculator start generating annual returns: (1) A starting-point portfolio value (I choose “100,000” in the test run described below); (2) the rate of real return that he wants to assume will apply for whatever percentage of the portfolio is not invested in stocks (I choose “2.0” in the test run described below); and (3) a starting-point valuation level identified through use of the P/E10 (P/E10 is the price of the S&P index over the average of the last 10 years of its earnings) valuation metric (I choose “20 Bear Market” in the test run described below — this is the valuation level that applies today). At this point, the user needs to click the “Enter” button.
The calculator will year by year generate a return that is part of a realistic 30-year return sequence of returns. The return sequences are in part random but not entirely so. A filter is applied to insure that the return sequences generated are consistent with the return sequences we have seen in the historical record.
That is, when returns get too high, a Reversion to the Mean factor will tend to pull them down. And, when returns get too low, a Reversion to the Mean factor will pull them up. However, the calculator reveals only one year of the return sequence at a time, at which point the user needs to enter the stock allocation he chooses to adopt for the next year. So the user never knows when the Reversion to the Mean phenomenon will kick in. Returns may continue climbing higher even from very high levels and returns may continue dropping lower even from very low levels, just as they do in real life. The filters were developed by John Walter Russell, former owner of the www.Early-Retirement-Planning-Insights.com site and are based on a regression analysis of the historical stock-return data.
When the “Enter” button is pushed for the first time (after the three inputs noted above are entered — there is a fourth, optional input to indicate amounts added to or withdrawn from the portfolio each year; I entered “0” in that input box in the test run described below), the Surfer reveals to the user the portfolio value, stock allocation percentage, and P/E10 value that applies for the current year of the 30-year return sequence being developed. In the test run that I am running to illustrate how the calculator works, the calculator tells me after I hit the “Enter” button that the Current Balance is 100,000 (the amount I chose), the P/E10 value is “20” (the value I chose) and that the stock allocation is “30 percent” (the percentage I chose). It is now left to me to elect either to stick with the same stock allocation for the first year of the 30-year return sequence (done by clicking the “Enter” button once again) or to change the stock allocation (done by choosing a new stock allocation from the drop-down box provided and then clicking the “Enter” button once again).
For guidance on how to know when to change stock allocations, please review the article at this site entitled “The Stock-Return Predictor.” The Predictor employs a regression analysis of the historical data to reveal to investors the likely long-term return on stocks starting from any of the various valuation levels. The Predictor indicates that a 30 percent stock allocation makes sense at today’s valuation levels (there is some individual judgment involved here — an investor must always take into consideration his financial circumstances, his life goals and his risk tolerance as well as the effect of valuations on long-term returns when choosing his stock allocation). So I will choose a 30 percent stock allocation for the first year of my hypothetical but realistic 30-year return sequence.
After I hit “Enter” the Scenario Surfer reveals the return I earned in Year One. It was a bad year for stocks. The P/E10 level dropped to 15.3 (fair value). My portfolio balance dropped to $95,870 (of course the drop was minimized by the low stock allocation elected). The calculator lists the portfolio balances that would apply for those following an 80 percent stocks rebalancing strategy, a 50 percent stocks rebalancing strategy, and a 20 percent stocks rebalancing strategy. The numbers are: (1) the value of the 80 percent rebalancing portfolio is $85,652 at the end of Year One; (2) the value of the 50 percent rebalancing portfolio is at $91,783 at the end of Year One; and ((3) the value of the 20 percent rebalancing portfolio is at $97,913 at the end of Year One.
Now that stocks are priced at fair value, I will increase my stock allocation to 60 percent; even at fair-value prices, I am reluctant to go with a stock allocation higher than that because stock prices usually fall to price levels far below fair value in the wake of out-of-control bull markets. After choosing “60 percent” from the pulldown menu for my stock allocation for Year Two, I hit “Enter” and the Surfer once again works its magic.
This time. I see an increase in the P/E10 level to 16.6. and an increase in my portfolio value to $106,043. For Year Three, I stick with my 60 percent allocation and my portfolio value drops a tiny amount to $104,566. At this point, the Valuation-Informed Indexing portfolio value is higher than any of the three rebalancing portfolios. For Year Four, I again stick with the 60 percent allocation and my portfolio value drops again, this time to $93, 599.
After four years of investing, I have a smaller portfolio than I started with. Only the 20 percent stock portfolio is up. This does not signal that the decisions to go with stock allocations of higher than 20 percent were ill-informed ones. Long-term stock returns are highly predictable. But they are not precisely predictable. There is always a range of possibilities that may turn up. The allocation choices made were reasonable ones, given the valuation levels that applied when the choices were made. In this particular case, the 20 percent stock allocation choice was rewarded. But the odds are that stock allocations higher than that will produce better results over the long term. The key to effective long-term investing is always seeking to have the probabilities on your side and not getting caught up in the emotion that can follow from seeing a strategy fail to immediately generate satisfactory results.
The P/E10 level at the end of Year Four is 11.4, a low P/E10 level. This means that the odds of big price jumps are much enhanced and the odds of big price drops are much diminished. So I am going to shift to a stock allocation of 80 percent. There may well be further price drops. But my study of the historical data tells me that the odds are at this point on the side of the investor going with a high stock allocation. I choose an 80 percent stock allocation, click “Enter,” and the Surfer reports —
— A small but not insignificant increase in portfolio value. I am now at $108, 594. Positive territory! And I am beating all three rebalancing portfolios. Here is a graphic showing the choices made and the results obtained for each of the first five years of the test run:
Note: The results for the remaining 25 years of the test run will be shown and discussed in Part Two of this article.