Last week's column argued that, while short-term timing (trying to guess where prices are headed in the next year) really does not work and is to be avoided, long-term timing (changing your stock allocation in response to big price swings with the understanding that you may not see benefits for doing so for as long as 10 years) always works and is to be celebrated. Does it really matter so much?
I believe that it matters a lot. I believe that taking valuations into consideration is the key to successful long-term investing. But I have learned over the past eight years that many investors feel an intense emotional resistance to the idea that long-term timing works. I can make the strongest intellectual case in the world for Valuation-Informed Indexing and it won't mean a thing if people do not let the words sink in. So we need to examine a bit why it is that many investors are so cold to the idea.
You've heard about banks that are too big to fail, right? The idea that timing works is the idea too big to accept as possible. If we learn that timing works, everything we know about stock investing changes.
The changes are wonderful stuff. The thing that stock investors most want to be able to do is to avoid losing their money in a stock crash. If it turns out that Yale Professor Robert Shiller is right that valuations affect long-term returns, guess what follows? If valuations affect long-term returns and long-term returns can therefore be known in advance (not precisely but to a considerable extent) by those making reference to the valuation level that applies on the day they buy, we no longer need to worry about losing our money in crashes. We can know when crashes are coming and step aside.
It sounds too good to be true. But we have an historical record that we can use to check whether the idea stands up to scrutiny.
From 1900 forward, there have been four stock crashes that remained in effect for a significant period of time. The first came after a time when we permitted the P/E10 level to rise above 24. The second came after a time when we permitted the P/E10 level to rise above 24. The third came after a time when we permitted the P/E10 level to rise above 24. The fourth came after a time when we permitted the P/E10 level to rise above 24.
There has never been a time when we permitted the P/E10 level to go above 24 in which we did not see a price crash. There has never been a price crash at a time in which we did not permit the P/E10 level to go above 24. I am beginning to detect a pattern.
Stock crashes are optional.
That's good news. No?
It's good news. And if the only thing we learned when we learned that valuations affect long-term returns was how to avoid stock crashes, I am confident that we would have used Shiller's findings to develop the Valuation-Inforned Indexing model back in 1981 and we would all be following it today. What's holding us back are the other implications that follow from Shiller's findings.
We don't take the concepts of overvaluation and undervaluation seriously today. Pretty much everyone acknowledges that overvaluation and undervaluation exist. But we pay these ideas lip service. If we took the concepts of overvaluation and overvaluation seriously, there are a number of things we would do very differently.
Web sites and radio shows and newspapers report the Dow and S&P numbers every day. How often do you hear the numbers reported adjusted for the effect of overvaluation? At the time when stocks were priced at three times fair value, the Dow was at about 12,000. Shouldn't the news reports have been saying that the Dow was temporarily priced at 12,000 but that the true value was 4,000? Wouldn't that have been the more accurate statement given the level of overvaluation?
People who at that time were buying index funds to finance their retirements were not really directing most of their savings to the purchase of stocks, were they? Take a person who invested $1,000 per month in stocks. Wasn't it only about $350 of his monthly 401(k) contribution that was going to stocks while $650 was going to the purchase of cotton-candy nothingness fated to be blown away in the wind over the course of the next 10 years or so? Isn't that the right way to think about it given that stocks were priced at three times their fair value at the time? Doesn't the word "overvalued" mean "mispriced"? The money you turn over to cover the cost added through a mispricing does not benefit you in any way, does it?
We would give people different sorts of advice re how to plan their retirements if we took the concept of overvaluation seriously. Today's retirement studies tell retirees that the safe withdrawal rate is always 4 percent. If overvaluation is a meaningful concept, doesn't it follow that the safe withdrawal rate varies with changes in valuation levels? Stocks are sometimes priced at one-half fair value and at other times are priced at three times fair value. Shouldn't the safe withdrawal rate be six times higher when the true value of a portfolio is double its nominal value than what it is when the true value is one-third of nominal value?
If Shiller is right and Fama and Malkiel and Bogle are wrong, 90 percent of the textbooks on stock investing get it wrong. If Shiller is right and Fama and Malkiel and Bogle are wrong, 90 percent of the investment calculators on the internet need to be corrected. If Shiller is right and Fama and Malkiel and Bogle are wrong, 90 percent of the risk management advice and asset allocation advice and retirement planning advice we have heard over the past 30 years is wrong.
What if everything you thought you knew about stock investing turned out to be wrong? That's the question that we all should be turning over in our minds today.
I see two possibilities. One is that Shiller is wrong and that the market really is efficient and that both overvaluation and undervaluation are meaningless concepts. The other is that Shiller is right and we need to revisit every strategic issue that we thought had been settled during the Buy-and-Hold Era. Either valuations matter or they don't. If they matter, every number used in investment analysis needs to be adjusted for the extent of the overvaluation or undervaluation that applies at the time. There is no in-between possibility that makes sense.
My take is that Shiller was right. My take is that Shiller's finding was the most important finding in the history of investment research. My take is that we need to begin exploring the implications of Shiller's finding that valuations matter in a serious way and taking our understanding of how stock investing works to places it has never been before.
No comments:
Post a Comment