Recent blog entries have reported on correspondence that I have had with www.IndexUniverse.com in an effort to get an article relating to Valuation-Informed Indexing published at the site. Set forth below is the text of an e-mail that I sent to Matt Hougan on February 6, 2008.
Thanks so much for your frankness in your response. That’s a huge help.
I of course agree that the insight that drives the Valuation-Informed Indexing approach is nothing more than common sense. It IS controversial today, however. I have seen thousands of people respond very well to these ideas. I have also seen a larger number get upset by seeing them. It’s not the ideas themselves that are controversial. It is the implications that readers can quickly see follow from the ideas that cause upset. Seeing these ideas appear on a computer screen cause some people to become alarmed that their investing strategies are flawed. My strong hunch is that that is going to continue to be the case for so long as prices remain as high as they are today.
You’re right about what happens in the ultra-long-term. The Return Predictor shows that the most likely 10-year annualized real return today is 1.55 percent. That’s obviously not too good. The most likely 30-year annualized real return is 5.43 percent. That’s very good indeed. Investors who follow Valuation-Informed Indexing will not “miss out” on the 30-year return by lowering their stock allocations for a time. The idea is to lower your stock allocation a bit at times of high prices and then increase it a bit when prices return to more reasonable levels. The investor following this approach suffers smaller losses during the time when prices are falling (on a long-term basis) and thus is able to invest more in stocks during the time when prices are rising (on a long-term basis).
You can see this clearly by using The Investor’s Scenario Surfer, another calculator available free at my site. This calculator permits the investor to compare what happens over a 30-year basis for those following rebalancing strategies and for those following valuation-informed strategies. Valuation-informed strategies come out ahead 90 percent of the time. That’s of course just what common sense says should happen. Yet is is indeed a controversial finding today. Most indexers today view rebalancing as a sound strategy.
You’re also right about dividends comprising the bulk of long-term returns enjoyed by stock investors. That’s a key reason why long-term returns drop so low at times of high valuations. The S&P dividend payout percentage today is pathetic (largely because prices are so high). That’s largely why the likely long-term return is so low. As prices return to more reasonable levels, the dividend payout percentage will of course increase. The expected long-term return will of course also increase. Stocks simply do not provide the same long-term return at different valuation levels. The return varies DRAMATICALLY with changes in valuations.
Your question about “opportunity cost” is a good one. There is indeed a potential opportunity cost. At today’s valuation levels, the likely rewards of lowering one’s stock allocation are so great as to overwhelm any potential opportunity cost. But that will not continue to be the case as valuation levels continue downward (over the long term, not necessarily the short term). It is fair to say that the general rule is that one should be highly invested in stocks, largely because of the potential opportunity cost of not being highly invested in stocks. Today’s prices are exceptional and the general rules does not apply.
My partner is John Walter Russell, owner of the Early-Retirement-Planning-Insights.com site. John has been doing research on these questions on a full-time basis for close to six years now. He can respond effectively to any statistical question you have (or I can ask him the question and report on the response to you). It’s not possible for me to respond to every possible statistical question in a single e-mail. We have examined thousands of questions.
I can assure you that there is no genuine dispute over what the numbers say. Every researcher is working from the same data. The reason why different researchers come up with different findings is that some include adjustments for valuations in their methodologies and some do not. If you take a bird’s eye view of how stocks perform in the long run, you get very positive numbers. There is no dispute on this point. If you compare how stocks did during the time-periods in which the P/E10 level was high with those in which it was moderate or low, you find that there has never been poor long-term performance starting from a moderate or low P/E10 level and that there has never been anything better than bone-crushing losses starting from the P/E10 level that applies today. There is not a single exception to this rule in the historical record.
The findings of the studies depend 100 percent on how the studies are set up. My common sense tells me that valuations must have an effect, so I place more confidence in the studies that are set up to reveal the difference. I certainly have no objection to pointing out in an article that there are studies that report very different findings and noting why. I want people to know about the other studies and to consider them. But I do not consider it a compelling argument to say that, if you ignore the effect of valuations, your findings will suggest that valuations do not have an effect. This “finding” is an inevitable result of any methodology that ignores this critical factor.
I am happy to go with whatever publishing approach you think best. My personal view is that it would be best for the first article to introduce the concept, and for follow-ups to go into greater detail on what the historical data says re particular aspects of the question.
Again, I can respond effectively to any question that you have or that any interested reader has. But I cannot respond to every question at one time. This is an idea that can be fully understood and fully accepted only in stages (because it is so at odds with today’s conventional thought, even though the idea is indeed rooted in a common-sense understanding of how things must work).
Please let me know if there is a particular statistical question that you need answered to begin to feel more comfortable. I am happy to do anything that can be done to help you feel better that these ideas are backed by the historical data. But I don’t want to try your patience by seeking to respond to too many questions in this one e-mail.
I’d like to make one last point re the controversy matter. The number of people who have responded in an extremely positive way to these ideas is in the thousands. Some of the posts that I put to the Motley Fool site on these questions are among the most highly recommended (by other community members) in the history of the site. You are right that there are some who have responded negatively. The full reality, though, is that there are many who are having thoughts along these lines and who would be excited to see someone spell out why their thoughts are not out of line.
It is my belief that you would do your readers a great service by giving them an opportunity to hear the other side of the story in a way in that would help them make sense of it. Your concern about accuracy is of course a good one. I think the key is to have standards (which I am confident I can meet, given how many years of work I have put into developing these ideas and given the number of experts that I have referenced and contacted and the amount of statistical work that has been done supporting the concept) but also to understand that there is no amount of data that can definitively prove things one way or the other, that ultimately the best way to proceed is to permit readers to hear both sides and to decide for themselves. Many will not be convinced. I am 100 percent confident that a good number will be. My confidence comes from six years of experience in putting forward these ideas and seeing the reactions that they generate among indexers.