Valuation-Informed Indexing #307
by Rob Bennett
Wade Pfau recently published an article in the Wall Street Journal (What Is the Sustainable Spending Rate for Retirees in 2016?) that offers a new take on the question of how to properly calculate the safe withdrawal rate. Wade modifies the concept a bit and offers new terminology to refer to his modified version of the concept — he calculates the “sustainable spending rate” rather than the “safe withdrawal rate.”
Wade does something that I find interesting and worthy of further study but also a tiny bit troubling.
I have been arguing for 14 years now (as a lone voice in the days before people like Wade came on the scene) that the Buy-and-Holders made a terrible mistake in not including a valuations adjustment in their safe withdrawal rate calculations. Shiller showed in his “revolutionary” (his word) research of 1981 that valuations affect long-term returns. So everyone who works in this field has known on at least one level of consciousness for 35 years now that stock investing risk is variable rather than constant and that it is thus not possible to calculate the safe withdrawal rate accurately without including an adjustment for the valuation level that applies on the day the retirement begins.
It obviously makes me happy to see an article appear in the Wall Street Journal that reports on the safe withdrawal rate concept and that includes a valuations adjustment, especially given that it was written by a fellow who is a respected figure in the field and who is someone that I consider a personal friend (Wade and I worked together on a daily basis for 16 months on issues relating to the safe withdrawal rate concept and he assured me in the strongest possible terms that Valuation-Informed Indexing really is a huge advance over Buy-and-Hold in every possible way). Few of us get to experience vindication (I have been banned at over 20 investing sites as a result of my “controversial” claims) in such a clear way in our journey through this Valley of Tears. For all this I am of course grateful. Thanks, Wade!
You can tell there’s at least a small “but” coming, right?
Wade does something that might make perfect sense but that leaves me queasy. I don’t at all want to suggest that you should go by what I say on the issue in question; Wade is a well-informed and skilled researcher — it is entirely possible that he is on the right track and that I am on the wrong track. But I feel that I owe it both to Wade and to his and my readers to say why I am not so sure, why I have doubts as to whether the way in which he is proceeding will end up being the last word on the subject of how to plan effectively for retirement.
Wade includes not only valuations but also interest rates in his calculations. He explains that: “An alternative way to view the historical data is to go beyond merely considering past withdrawal rate outcomes by seeing how they relate to retirement date values of the underlying sources of returns…. Using a statistical method called ‘multivariate regression analysis,’ we can analyze the relationship between the maximum sustainable withdrawal rate over a thirty-year retirement with a 50/50 asset allocation to stocks and bonds, and the values of Shiller’s PE10 and the ten-year Treasury rate at the start of those retirements.”
I have personal experience with Wade’s talent at working with sophisticated statistical tools. So I am highly confident that he is using this tool in a technically proper way and that it is a tool capable of generating important insights. So part of me applauds the advance. This new concept is certainly worthy of further investigation.
My reluctance to offering a full endorsement is rooted in a concern that this new approach may move us away from appreciation of a powerful insight brought to us by the Buy-and-Holders that in my assessment has stood the test of time (in contrast to the Buy-and-Hold idea that valuation adjustments are not required).
I feel 100 percent comfortable including valuation adjustments because I see the resistance to consideration of investor emotions as a blind spot for the Buy-and-Holders. They see investing as a rational enterprise; mispricing can only be the result of irrationality and so it makes Buy-and-Holders uncomfortable to consider valuations. They don’t have a good reason for ignoring valuations. They do so because they are constitutionally not able to accept the reality that investors are human and humans are highly emotional creatures.
The same cannot be said of interest rates. The reason why Buy-and-Holders do not consider interest rates in their safe withdrawal rate calculations is that they believe that investors price in interest rates when setting the price of stocks. There is no need under the model to consider the effect of interest rates separately as this effect is baked into the cake by being reflected in prices (which of course ARE considered in retirement calculations done by the Buy-and-Holders).
I don’t feel comfortable saying that it is a bad idea to include both of the factors ignored by the Buy-and-Holders. But I also don’t feel entirely comfortable endorsing the idea of considering both factors. I feel confident that valuations need to be considered and worry that considering interest rates as well might detract from the clarity of findings that reflected only a valuations adjustment.
All that said, it is entirely possible that Wade is onto something important. Dogmatism is a big mistake this early in the proceedings. In general the motto in investing analysis should be: Let a thousands flowers bloom! Wade is generating new ways of looking at old questions and we all should be grateful for his explorations.
Rob Bennett’s bio is here.
