"Valuations are on the high side," warned Janet Yellen last week... "but not outside historical norms." It appears the range of her norms needs to be adjusted... As The Treasury's Office of Financial Research warns [8], stocks appear more costly than P/E ratios and other basic indicators would suggest.
As David Stockman exclaimed yesterday, [9] forward P/Es are useless in judging valuation as Wall Street plays the EPS game. As OFR explains, other fundamental valuation metrics tell a different story than the forward PE. This brief focuses on a few — the CAPE ratio, the Q-ratio, and the Buffett Indicator — that are approaching two-standard deviation (two-sigma) thresholds.
Why is two-sigma relevant? Valuations approached or surpassed two-sigma in each major stock market bubble of the past century. And the bursting of asset bubbles has at times had important implications for financial stability. The two-sigma threshold is useful for identifying these extreme valuation outliers. Assuming a normal distribution in a time series, two-sigma events should occur once every 40-plus years; in equity markets, they occur more frequently due to fat-tail distributions.
click image for large legible version
[10]
CAPE Ratio. If one-year earnings assumptions based on peak profit margins are potentially misleading, then it seems logical to consider valuation metrics based on normalized (long-run average) profit margins. In 1934, Graham and Dodd argued average earnings should cover a period of at least 5 years, and preferably 7 to 10 years, on the basis that current earnings rarely reflect a company’s sustainable earnings capacity. They noted that longer periods are “useful for ironing out the frequent ups and downs of the business cycle” and provide a better measure of a company’s earnings power than a single year. Shiller enhanced this concept with CAPE, which is the ratio of the S&P 500 index to trailing 10-year average earnings (earnings are based on generally accepted accounting principles, or GAAP, and are inflation-adjusted). Although CAPE’s 10-year timeframe is somewhat arbitrary, it captures earnings over one or two business cycles rather than over a single year, better reflecting sustainable earnings.
The historical CAPE average based on a 133-year data series is approximately 17 times, and its two-standard-deviation upper band is 30 times. The highest market peaks (1929, 1999, and 2007) either surpassed or approached this two-sigma level (1999 exceeded four sigma). Each of these peaks was followed by a sharp decline in stock prices and adverse consequences for the real economy. At the end of 2014, the CAPE ratio (27 times) was in the 94th percentile of historical observations and was approaching its two-sigma threshold.
Q-Ratio. The Q-ratio, defined here as the market value of nonfinancial corporate equities outstanding divided by net worth, suggests a similar message of equity valuations approaching critical levels. Instead of using a traditional accounting-based (historical cost) measure of net worth, the Q-ratio incorporates market value and replacement cost estimates. The Q-ratio also includes a much broader universe of nonfinancial companies (private and public) than CAPE.
Buffett Indicator. The ratio of corporate market value to gross national product (GNP) is at its highest level since 2000 and approaching the two-sigma threshold. This indicator is informally referred to as the Buffett Indicator, because it is reportedly Berkshire Hathaway Chairman Warren Buffett’s preferred measure to assess overall market valuation. Historically, this indicator’s message is consistent with CAPE, particularly in identifying periods of extreme valuation before the Great Recession and the 1990s technology stock bubble.
* * *
Of course, what really explains why Janet can't/won't see this ugly reality...
[11]
* * *
As OFR concludes,
These readings illustrate the potential for “quicksilver markets,” in which prices shift rapidly and unpredictably, the Washington-based analyst wrote. All three gauges are “nearing extreme highs.”