I suspect Buffett's article will be characterized as very negative, but he is actually not as bearish as one might think. His projection of 4% real returns after frictional costs translates into 5-6% real returns before these "horrendous" costs, which he estimates at $130 billion per year, an amount equal to more than a third of the Fortune 500's total profits in 1998. (Buffett's diatribe about these costs are well summarized in last Thursday's Fool on the Hill.) Five to six percent real returns are only slightly below the remarkably stable 6.6-7.5% real returns that the stock market has delivered in every major 50-plus-year period since 1802, according to Jeremy Siegel's analysis in his outstanding book, Stocks for the Long Run. Thus, for those investors who are able to minimize frictional costs and at least match the overall market's rate of return, Buffett's projected returns will be disappointing only to those who irrationally expect a continuation of the unprecedented returns of the past 17 years.
Let's examine the assumptions underlying Buffett's projections. He argues that the return on stocks in the future will likely mirror the growth of profitability of American businesses plus dividends. He writes, "The Tinker Bell approach -- clap if you believe -- just won't cut it.... Investors as a whole cannot get anything out of their businesses except what the businesses earn." He assumes 3% real GDP growth, implies 4-5% growth in corporate profitability, adds 2% for inflation, and subtracts 1-2% for frictional costs, which yields approximately 6% nominal returns or 4% real returns, assuming investors continue to value earnings at the same levels that they do today.
Could stocks do better? Yes, Buffett says, citing two possibilities: corporate profitability in relation to GDP continues to rise and/or interest rates fall even further, thereby increasing the amount that investors would be willing to pay for each dollar of earnings. However, given that these factors are already at historical extremes, it is clear that Buffett expects these trends to halt and even regress toward the mean.
I agree with Buffett regarding the importance of these two factors and the implicit assumption underlying them: that the market's valuation level in the future will continue to be a function of two factors, corporate earnings and the P/E investors are willing to assign to those earnings. However, I think there are a number of other possibilities that Buffett does not address that could either cause corporate profits to rise more quickly than he projects or valuation ratios to rise even further -- or both.
Corporate Profits Rising Faster Than Expected
Regarding corporate profits, I believe it is quite possible (though I won't go so far as to say likely) that many factors, especially businesses' adoption of the enormous advances in technology over the past couple of decades, are just beginning to trigger a boom in growth and productivity. The data is starting to show this. According to recently revised historical GDP data released by the Commerce Department's Bureau of Economic Analysis on October 28, and Business Week's analysis of it, U.S. economic and productivity growth has been significantly higher than earlier data showed, and is accelerating.
That there would be a substantial lag time between the investments in technology and the results is not surprising. A century or so ago, it took decades for the advances made possible by electricity and the electric motor to affect national productivity levels. When I look at how some companies today -- still remarkably few -- have adopted technology, especially the Internet, to streamline operations and communications, I can't help but be bullish on the impact of more and more companies doing the same.
Other factors could also contribute to higher-than-expected growth of corporate profits, such as improved corporate governance, more insider ownership (a topic covered in one of my previous Boring Port columns), greater focus by managers on improving returns on capital rather than expanding empires, an increase in the value of companies' brands, and research and other intellectual property that aren't always reflected in GAAP earnings.
Expanding Valuation Multiples
Regarding the multiples that investors place on corporate earnings, I believe it is possible that, even if Buffett's assumption of steady interest rates is correct, average P/Es and other valuation metrics could be higher many years from now than they are today, for three reasons. First, Stocks for the Long Run shows conclusively that, for the long-term investor, stocks have always been undervalued. For example, even if you had been unfortunate enough to invest $100 in the market just before the 1929 crash, 30 years later you would have had $565 after inflation, versus $141 and $79 had you invested in bonds and T-bills, respectively. Similar results are true of every other market crash this century.
Also, stock returns, unlike those of other asset classes, have been remarkably stable over time. Over rolling 20-year periods back to 1802, stocks have never failed to beat inflation by less than 1% annually, whereas bonds and T-bills have declined in value by more than 3% annually over their worst 20-year periods. As Siegel's data and other data like it become more widely known, the risk premium investors assign to stocks could decline, resulting in an increase in their valuation. (This argument is the thesis of the new book, Dow 36,000, which has merit, but I think takes the argument to an extreme; click here to read a summary of the book that appeared in Atlantic Monthly.)
Second, the environment for investors could continue to improve. A high degree of corporate disclosure, transparent, real-time capital markets, and easy access to information on the Internet all increase the accessibility of stocks and reduce the risks of owning them.
Third, the economic characteristics of American businesses could continue to improve. I believe today's leading American businesses have superior economic characteristics to similar businesses in earlier periods, and that this trend could continue. For example, let's compare a group of leading companies in 1972, known as the Nifty Fifty, with a similar group of today's stocks. (Click here to see a list of the stocks and the data about them that I have compiled.) You can see that the average return on equity of the 20 companies that are in both Nifty Fifties is 18.7% in 1972 versus 28.7% today. While return on equity is not a perfect measure by any means, this data provides evidence that the same companies are generating higher returns on equity today than they were during a bull market period of an earlier era.
I agree with Buffett's main points, but am slightly more optimistic than he is. While I believe that most of the bullish theories I've noted above will turn out to be wishful thinking and self-delusion (after all, five centuries before Christ, Demosthenes noted that "What a man wishes, he will believe"), I think that at least one or two of these scenarios are likely to materialize, resulting in annual stock market returns that will be slightly (say, 1-3%) better than Buffett projects.
W. Tilson appreciates your feedback at Tilsonfunds@aol.com. To read his previous guest columns in the Boring Port and other writings, click here.
Make a Living Foolin' Around.
|Recent Boring Portfolio Headlines|
|10/30/00||American Power Conversion's Ugly Earnings|
|10/23/00||Cisco's Formidable Challenge|
|10/16/00||Cisco, Apple, and Probabilities|
|10/09/00||Perils and Prospects in Tech|
|10/02/00||Learn From Mistakes|
|Boring Portfolio Archives »|