August 3, 2010
Shiller’s P/E model has been accurate in predicting future returns over long time horizons (at least 10 years); it is notoriously inaccurate at predicting short-term (e.g., one-year) returns.
In January, I spent 90 minutes with Shiller over lunch, having fortuitously bumped into him at the Atlanta airport. At that time, he made the comment that even when stocks were at 20-times average ten-year earnings, investors could still do respectably in the following period.
Today, he is more pessimistic. He’s concerned that eroding confidence by American consumers and businesses could lead to a downward spiral of reduced spending, which in and of itself could trigger a double-dip recession.
And he concluded our conversation by saying that while he expects returns in stocks to be positive, he doesn’t think they’ll be stellar – he’s uncertain whether investors will be better off in stocks or in bonds over the next ten years.
The argument for stocks as cheap
Jeremy Siegel uses a different method to value stocks and reaches a different conclusion – his analysis suggests that, compared to long-term averages, stocks are undervalued by 25% to 30%.
The biggest difference between his approach and Shiller’s is that his is forward-looking, focusing on consensus earnings forecasts for this year and next. Among his criticisms of Shiller’s methodology is that mega-writeoffs, such as the $80 billion writedown by AIG, will distort the earnings base from which backward-looking calculations are conducted for years to come.
Siegel has looked at U.S. stock market valuations over a 200-year period. During that time, the average stock multiple of earnings has been 15-times – that compares with a multiple of consensus earnings forecasts of 13-times for this year and 11-times for next year.
The impact of low interest rates
Siegel’s average of 15-times earnings includes periods of double-digit inflation, when multiples are typically depressed. Excluding those periods of double-digit inflation, the average multiple that the market paid for earnings was 17-times.
If earnings forecasts for next year are accurate, then returning to that long-term average of 15-times earnings would result in stocks increasing by 35%; rising to the historical low-inflation valuation norm of 17-times would see stocks rise by 50%.
Would you like to send this article to a friend?Remember, if you have a question or comment, send it to .