P/E Ratios as Predictive of Future Returns
I personally don't believe that P/E ratios of themselves are a determinant of stock prices, rather it is discounted cash flows that matter. But a good case can be made that P/E's and their trends are a proxy for causes which determine future stock prices (and thus P/E's can be used predictively). John Mauldin presents a couple of the supporting arguments in his April 8th and April 22nd newsletters.
The first comes from his review of Ed Easterling's book Unexpected Returns: Understanding Secular Stocks Market Cycles. The second is from work done by Jeremy Grantham. You can read the links for more details, but here's a quick synopsis of each:
"Easterling demystifies secular cycles and explains that they are not an unexplainable pattern or coincidence; rather secular cycles are caused by the trend in the P/E ratio. Corporate earnings tend to grow over time, whether in secular bull or bear periods. During the periods of secular bulls, earnings growth is magnified by a rising P/E ratio. For example, when earnings grow by 6% annually over a 12- year period, they double. Compounding has a powerful effect. If the P/E ratio remains the same over that period, the stock market would double based upon the increase in earnings. During a secular bull cycle, P/Es have often started near 10 and ended over 20, a doubling of the P/E ratio. When combined together, however, the effect is a four-fold increase in the market! A raging bull that provides stellar returns....
In a secular bear cycle, P/E ratios fall from high to low. In the prior example, had the P/E started at 20 and ended at 10, the change in the P/E would have offset the rise in earnings to produce a flat result. Many secular bear markets have been long, volatile periods with little to show at the end. And some of them start with P/Es well over 20 and end well under 10--those periods deliver negative returns. Even when the market is flat over a long period, Easterling points out that the effects of inflation cause investors to have a substantial loss in purchasing power.
His fourth section in Unexpected Returns explains the reasons that P/Es tend to rise and fall over many years. In a step-by-step process, he builds the components that tie factors in the economy to the stock market. The result is a clear understanding of the impact of inflation on P/E ratios. With inflation recently at low levels, the likely direction for the stock market is down, because rising inflation is never good for the stock market, nor is deflation. To expect inflation to always be low and stable simply ignores history."
"Here's a study done by Jeremy Grantham, where he breaks up the years from 1925-2001 by looking at the average price to earnings level for the year. He then groups the years based on this valuation into 5 different buckets. The highest price to earnings years was labeled the "most expensive 20% of history"; the lowest price to earnings years was labeled the "cheapest 20% of history." What he found is that over the next 10 years the cheapest or second cheapest quintiles had an average compound return of 11%. However, if you invested in the most expensive quintile in history, the average compound return over the next 10 years was zero."