Stocks have jumped about 70% from the March lows. As we’ve noted, they have also blasted past most estimates of fair value, which are generally around 900 on the S&P 500 on a cyclically-adjusted price-earnings ratio (see professor Robert Shiller’s chart below),
London economist Andrew Smithers says valuations are even more extreme. In a recent interview with Kate Welling of Weeden & Co., he put the overvaluation at 40%:
Would you mind running through how you arrive at that 40% overpriced valuation?
Certainly. The EPS on the S&P 500 for the 12 months to June 20, 2009 were $7.51, making the P/E with the index at 1073 (when I ran the numbers on Oct. 13) 143, which was 10 times the long-term average P/E, using data which start in 1871.
Which superficially sounds like a lot more than 40% overvalued —
Yes, but that doesn’t mean the market is horribly expensive, because profits have recently been
quite depressed. Equally of course, though this point is often ignored, claims that the market is selling at some low future multiple would not show that the market today is cheap.
In order to assess value, it is necessary either to calculate the level at which the EPS would be if profits were neither depressed nor elevated, or to use a metric of value which does not depend on profits. The cyclically adjusted P/E (CAPE) normalizes EPS by averaging them over 10 years [See chart above]. It thus follows the first of those two possible methods. Using even longer time periods has advantages, particularly as EPS have been exceptionally volatile in recent years —and using longer time periods raises the current measured degree of overvaluation.
The other methodology we use measures stock market value without reference to profits: the q ratio. It
compares the market capitalisation of companies with their net worth, also adjusted to current
prices. The validity of both of these approaches can be tested and is robust under testing —and they produce results that agree. Currently, both q and CAPE are saying that the U.S. stock market is
about 40% overvalued.
Of course, today’s overvaluation doesn’t tell you much about what stocks will do next week, next year, or even the next 5-10 years. As the chart above shows, before the 2007 market crash, stocks were overvalued for the better part of 20 years–and observing that didn’t help you make money. On the contrary, it usually got you fired.
What today’s valuation does suggest is that stocks are priced to return a bit less than average over the next decade, perhaps 3% real per year (inflation adjusted), as compared to the 6%-7% average.
Today’s valuations also suggest that stocks may have gotten way ahead of themselves, especially in light of the structural problems that will continue to bog down the economy.
As the chart above illustrates, every one of the prior mega-busts in the past century has been followed by a “trough” in which the cyclically adjusted PE ratio hit the high single-digits. We didn’t quite make it there in March (the P/E bottomed around 12X), although we did get close.
This, combined with what is likely to be a decade of deleveraging, consumer retrenchment, and sluggish growth as we work off our debt binge, suggests that we still yet might hit that single-digit low before we take off on another secular bull market again. This could be achieved either through another market crash, or a prolonged period of backing and filling as earnings growth gradually reduces the long-term PE ratio (this is what happened in the 1970s).
On the other hand, it is possible that that enormous stimulus and zero interest rates over the past two years will produce that “v-shaped” recovery. At this point, given the extent of the recent rally, it would presumably have to be one heck of a “V” to send stocks soaring from here. But the last eight months have already made idiots out of almost everyone.