Valuation-Informed Indexing #289
by Rob Bennett
The last five columns examined a recent article by Michael Kitces (Should Equity Return Assumptions in Retirement Projections Be Reduced for Today’s High Shiller CAPE Valuation?) that advances the highly counter-intuitive and yet entirely accurate claim that: “The ideal way to adjust return assumptions…[may be] to do projections with a ‘regime-based approach to return assumptions. This would entail projecting a period of much lower returns, followed by a subsequent period of higher returns.”
This changes everything that we once thought we knew about how the stock market works. The old (and still dominant) belief was that stock prices fall in the pattern of a random walk because price changes are caused by economic developments. If what Kitces is saying is so (I strongly believe that it is), prices do not fall in a random walk at all. They play out according to a highly predictable long-term pattern. For about 20 years, valuations rise (with short-term drops mixed in). Then, for about 15 years, valuations drop (with short-term rises mixed in). It is investor emotion that is the primary determinant of stock price changes. Investors can reduce risk dramatically while also increasing return dramatically by adjusting their stock allocations in response to big valuation shifts and thereby keeping their risk profile roughly constant as one “regime” is replaced with another.
The implications are very far-reaching indeed. Consider what Kitces’ claim suggests re retirement planning. If there are long time-periods in which valuations are headed upward and other long time-periods when valuations are headed downward, the safe withdrawal rate (the percentage of a portfolio that a retiree can take out to cover each year’s living expenses) cannot possibly be a constant number. If Kitces is right about the need to accept that stock returns are regime-based rather than random-walk based, the safe withdrawal rate is a variable. The magic number is not always 4 percent, as we once believed. In regimes in which valuations are headed upward, the SWR is a number bigger than 4 percent, sometimes a lot bigger. In regimes in which valuations are headed downward, the SWR is a number smaller than 4 percent, sometimes a lot smaller.
Kitces’ article shows that doubts about the Buy-and-Hold concept have grown strong enough since the 2008 crash that we are almost where we need to be for there to be widespread acceptance that the SWR is a variable and that we should be telling aspiring retirees to identify the SWR that applies for their particular retirement by taking into consideration the valuation level that applies on the day the retirement begins. There have been numerous articles in big-name publications in recent years pointing out the dangers of the once-highly-popular 4 percent rule. So we are indeed close to a breakthrough on our understanding of how to construct effective retirement plans.
But we have not quite moved the football over the goal line.
Kitces was asked in the comments section for his article how the regime-based concept affects retirement planning. He didn’t say what I believe, that it is not possible to calculate the SWR accurately without including a valuations adjustment. He said: “The ‘good’ news is that this doesn’t really change SWR much, as the reality is that the SWR results all come from high-valuation environments in the first place. Actually, the key distinction is that if you retire and valuations are NOT high, the SWR is more like 5%-6%, not 4%!”
I am not able to make sense of that claim. It is certainly true that the cases in which no withdrawal rate higher than 4 percent permitted the retiree’s portfolio to survive were all cases in which high valuations applied. What Kitces is missing is that the return sequences that caused withdrawal rates higher than 4 percent to fail were not worst-case return sequences. The 4 percent rule worked in that tiny number of cases. But if you applied worst-case return sequences (return sequences that we have seen in the historical record but that did not pop up in any of the tiny number of cases in which the 4 percent rule worked despite insanely high valuations), the 4 percent rule failed badly. The historical data shows that, when valuations are what they were in 2000, a retirement plan employing a 4 percent withdrawal rate has only a one-in-three chance of surviving 30 years. Not safe! Not close!
Kitces does not agree with me. And the investing advice world in general does not agree with me. But I do think we are getting closer all the time. I believe that we are one stock crash away from seeing the biggest change in our understanding of how stock investing works ever seen in the history of investing analysis. Even Kitces is not as sure of himself as he suggests he is in the words of his that I quoted above.
I engaged in extensive e-mail correspondence with him re this topic a number of years back. He took the same position then that he takes in his recent article — he believes that the SWR can climb higher than 4 percent at times of low valuations but that it can never drop below 4 percent no matter how insanely high valuations go. But his mind is not entirely closed. He said: “If Russell’s [the reference is to John Walter Russell, my partner in development of The Retirement Risk Evaluator, the first New School SWR calculator, one that adjusts for the valuation level that applies on the day the retirement begins] projections of a sub-2% withdrawal rate prove to be true, or are anywhere close, those who retired back in 2000 may ultimately find that the ‘4% safe withdrawal rate’ was still far too aggressive, making the point once again about how critical it is to incorporate market valuation into retirement projections!” Kites told me that: “I’m quite encouraged with John Russell’s results” and that “I’m not saying that the New School is wrong, I’m just saying that I don’t think it’s appropriate to state that the New School has proven the Old School wrong to the point that the Old School should be utterly and completely dismissed.”
I believe that following the next price crash Michael will be an even stronger advocate of Valuation-Informed Indexing than he is today. I believe that we all will be more open to exploring the far-reaching implications of Shiller’s “revolutionary” (his word) findings of 1981 at that time. I believe that we stand today on the threshold of discovering some amazing things about how stock investing works in the real world and that Michael’s article about regime-based return assumptions will then be appreciated as an article that played a significant role in taking us all to a better understanding of a subject that affects the working of our economic system and even of our political system in important ways.
Rob Bennett’s bio is here.