Excerpt from the Hussman Funds' Weekly Market Comment (7/19/10):
Investors who allow Wall Street to convince them that stocks are generationally cheap at current levels are like trout - biting down on the enticing but illusory bait of operating earnings, unaware of the hook buried inside.
I continue to urge investors to have wide skepticism for valuation metrics built on forward operating earnings and other measures that implicitly require U.S. profit margins to sustain levels about 50% above their historical norms indefinitely. Forward operating earnings are Wall Street's estimates of next year's earnings, omitting a whole range of actual charges such as loan losses, bad investments, restructuring charges, and the like. The ratio of forward operating earnings to S&P 500 revenues is now higher than it has ever been. Based on historical data (see August 20, 2007 Long Term Evidence on the Fed Model and Forward Operating P/E Ratios), the profit margin assumptions built into forward operating earnings are well beyond two standard deviations above the long-run norm. This is largely because, as Bill Hester noted in his research article last week, forward operating earnings are heavily determined by extrapolating the most recent year-over-year growth rate for earnings. In the current instance, this is likely to overshoot reality, and in any event, has little to do with the long-term cash flows that investors can actually expect to receive over time.
I can't emphasize enough that when you hear an analyst say "stocks are cheap based on forward operating earnings" it would be best to replace that phrase in your head with "stocks are cheap based on Wall Street's extrapolative estimates of a misleading number."
More sober and historically reliable measures of market valuation create a much more challenging picture. Apart from our own measures, which indicate continued overvaluation, there are several good indicators of market valuation that are not overly sensitive to year-to-year fluctuations in profit margins. One is based on the 10-year average of actual net (not operating) earnings, which is advocated by economist Robert Shiller, and another is Tobin's "q" ratio which is based on comparing market value to replacement cost, and is advocated by Andrew Smithers. Both of these measures largely agree with our own measures, both presently and on a historical basis. Based on last week's valuations, both suggest that the S&P 500 is substantially overvalued.
[Geek's Note: The chart above is based on the ratio of the CAPE and q to their respective historical averages. Note that the axis is logarithmic, so the level of 0.4 corresponds to a valuation ratio of exp(0.4). This is about 1.5, or 50% overvalued. In contrast, major secular buying opportunities as we observed about 1950 and again in 1974 and 1982 occurred at values of about -0.6, which corresponds to a ratio of log(-0.6). This is about 0.55, implying that the market at those points was about 45% below historical norms.]