Yesterday’s blog entry set forth the text of a an e-mail that I sent to academic researcher Wade Pfau on December 16, 2010. Wade responded the next day.
Wade said in his December 17 e-mail that: “In thinking about it further, I owe you further apology as my comment was based on a misunderstanding about your past claims.” He explained that: ” I understood that you meant this as the safe
withdrawal rate, but somehow I still confused myself into thinking that this number represented your prediction for the actual withdrawal rate. If it was a prediction, then naturally you would need to indicate the uncertainty surround it. But it is not a prediction. It is just a lower bound that should be safe. So you don’t have to be so worried about its precise value. ” He speculated that the person who had suggested in an earlier thread at the Vanguard Diehards board that I had been dogmatic in my posting “made a similar mistake” and that this mistake “confused my ability to understand your response.”
He added that “actually I am a neophyte to the whole withdrawal rate debate. I didn’t even know about Bengen or the Trinity study until July” but “now I am quite interested in this topic.”
Wade expressed confusion over “why some Bogleheads are so threatened by using valuations.” He noted that John Bogle has in many cases said things “not all that different from what you said.”
He concluded the e-mail with a kind compliment. He said: “I should also just say that you are a very good writer….Your writing really commands attention.”
The text of my response is set forth below:
That all sounds good.
Please enjoy your week away. I expect to send you an e-mail next week that will provide a link re your question about the timing of John’s work and perhaps address a few other substantive points. But I of course understand that you will not see it until the following week.
Bernstein’s discussion of SWRs is on Page 234, if I recall correctly. I have cited it so many times over the years that I have the page number memorized. The claim that I often make is that the Old School SWR studies are “analytically invalid” and I often cite Bernstein’s words in support of this claim. One of my fellow community members who does not like me using that phrase sent Bernstein an e-mail asking him if he agrees that the Old School studies are “analytically invalid.” Bernstein said that “of course” they are analytically valid. But he followed that up by saying that anyone giving thought to using one of them to plan a retirement would be well-advised to “FuhGedDaBouDit!” That’s the point that I mean to convey with the claim that they are “analytically invalid”! The purpose of the studies is to help people plan retirements. They are not designed in such a way as to be able to do that effectively. So in my eyes they are analytically invalid. They do not the job that they were set up to do.
It may make you feel better to know that John Walter Russell also made the mistake of confusing the number that is at the lower bound of the confidence interval (the SWR) with the number most likely to turn up. John and I exchanged about 10 e-mails on this point a long time ago. I agree with you that there’s a good chance that the other poster in the thread you read was making a similar mistake. This is not intellectually difficult stuff (at least not the parts that I understand — I have no background with the statistical tools). But there is an undercurrent here that is EXTREMELY counter-intuitive. I have seen it throw many smart and good people off track. I of course would like to be able to figure out how to communicate the points in a way that avoids the confusion that enters into just about every discussion of these matters. I have picked up some clues as to how to do that over time. But it is the hardest job that I have ever tackled in my life. The way to spin this in a positive way is to observe that, if the confusion is today very deep, the prospect of making a giant leap forward in our understanding of how stock investing works once we overcome the confusion is also great.
Your words about John Bogle are 100 percent right on! It was by reading Bogle’s “Common Sense on Mutual Funds” in the mid-1990s that I got on the track that I am now on. I am the biggest Boglehead in the world. The investing strategy that I recommend is called Valuation-Informed Indexing. It is a mix of Bogle’s best ideas and Shiller’s best ideas. I say that Bogle and Shiller go together like chocolate and peanut butter. The thing that I say that some view as anti-Bogle is that Bogle made the biggest mistake in the history of personal finance when he said that it is possible to “Stay the Course” without being willing to change your stock allocation in response to big valuation shifts. But I don’t see it as being such a big deal that Bogle made a mistake. We obviously all make mistakes and Bogle’s ideas have led to many breakthroughs (including Valuation-Informed Indexing — there clearly would not be any VII today without the insights contributed by Bogle).
It may be that some of the anger felt by some of the Bogleheads is BECAUSE my ideas are so influenced by Bogle’s. The angry ones could dismiss my ideas easily if they differed in many ways from Bogle’s. Because they are so closely related,
it is hard for them to dismiss them. Yet the strategy recommendations are very different. In January 2000, a Valuation-Informed Indexer would probably have been going with a stock allocation of about 10 percent. A Buy-and-Holder would
probably have been going with 70 percent. That’s a big difference! My only difference with Bogle is over the valuations question, but valuations are so important that we often end up in very different places.
As you note, it’s not that Bogle rejects the idea that valuations affect long-term returns. I learned this from him! It’s that Bogle does not IMPLEMENT the insight. He SAYS that valuations matter. But his allocation recommendations do not take valuations into account. That’s the entire deal. That’s the only real question in dispute in the eight-year-long debate.
Bogle said in an interview that he thinks VII can work:
But when I sent him an e-mail asking for his help in dealing with the abusive posting, he did not respond:
Please have a great week away from all this and perhaps we will be able to talk over some ideas for further research when you get back. There are all sorts of possibilities. I can assure you that I had zero idea what I was getting into when I put up that first post back on the morning of May 13, 2002. This is a deep well!
One last point. Over the years I have had a number of people ask me about international SWRs. It sounds like your recent paper will go a long way to answering their questions. I will definitely be checking it out and linking to it in future days.