Profit Margins – Is the Ladder Starting to Snap?
John P. Hussman, Ph.D.
Since mid-2014, the broad market as measured by the NYSE Composite has been in a broad sideways distribution pattern, with an increasing tendency in recent months for advances to occur on weaker volume and declines to follow on a pickup in volume. While capitalization-weighted indices have done somewhat better since mid-2014, the S&P 500 Index is unchanged since late-December. On the basis of broad market action across individual stocks, industries, sectors and security types, as well as a general widening of credit spreads and other risk-sensitive measures, we continue to infer a shift toward increasing risk-aversion among investors.
On the valuation front, we should begin by emphasizing that the most historically reliable valuation measures – those most closely related to actual subsequent market returns over full-cycle horizons of about 7 years or more – are those that adjust for the cyclical variation of profit margins (see Margins, Multiples and the Iron Law of Valuation). Indeed, on a 10-year horizon, these measures are generally about 90% correlated with actual subsequent total returns in the S&P 500, and this relationship has persisted even in recent market cycles.
Though it’s not widely recognized, measures such as the ratio of market capitalization/nominal GDP and the S&P 500 price/revenue ratio are actually better correlated with actual subsequent total market returns than price/operating earnings ratios, the Fed model, and even the raw Shiller P/E (though the Shiller P/E does quite well once one adjusts for the embedded profit margin). To fully understand the present valuation extreme, recognize that the market cap/GDP ratio is currently about 1.29 versus a pre-bubble norm of just 0.55, with “secular” lows such as 1982 taking the ratio to about 0.33. To fully understand the present valuation extreme, recognize that the S&P 500 price/revenue ratio is currently about 1.80, versus a pre-bubble norm of just 0.8, with “secular” lows taking the ratio to about 0.45.
To put these figures in perspective, if we assume that nominal GDP and corporate revenues will grow perpetually at a 6% nominal rate (which is much faster than we actually observe here), and the market does not experience another secular market low until 2040 – a quarter century from now, the S&P 500 Index would still be approximatelyunchanged from current levels at that secular low 25 years from today. The arithmetic here is relatively simple. For market cap/GDP the annual percentage change of the S&P 500 Index over that 25 year period would be (1.06)*(0.33/1.29)^(1/25)-1 = 0.37%. For the price/revenue ratio, the calculation would be (1.06)*(0.45/1.8)^(1/25)-1 = 0.28%. In both cases, dividend income would result in a somewhat higher total return over that quarter-century horizon.
One obtains less extreme conclusions, though still uncomfortable, if one assumes that these multiples simply touch their pre-bubble norms a decade from now. In that case, the annual percentage change in the S&P 500 Index over the coming decade would be -2.67% and -2.26%, respectively. If all of this seems preposterous, keep in mind that these revolting long-term prospective returns in stocks are simply a less recognized variant of what we observe more clearly in the bond market, where long-term interest rates are now negative in about one third of the developed world. Investors are literally paying their governments for the privilege of lending to them on a 5-10 year horizon. Investors are collectively out of their minds if they believe that current equity prices don’t quietly reflect the same absurd state of affairs.
We’re not talking about dusty, arcane valuation levels that investors can assume will never be seen again. Indeed, market valuations fell below pre-bubble norms even during the most recent market cycle (see Why Warren Buffet is Right and Why Nobody Cares). Deviations in valuation from historical norms have reliably been associated with similar deviations in long-term returns from their historical norms – which is why the S&P 500 has posted a total return of less than 4% annually over the past 15 years – and even then only because valuations have again been pushed to bubble extremes. Understand also that these valuation measures haven’t missed a beat in a century of market cycles.
Our objective isn’t to convince anyone to abandon their own disciplines – only to clearly lay out the current situation for those who value our work. I’ve been very open about the awkward transition that resulted from my 2009 insistence on stress-testing our methods against Depression-era data (see A Better Lesson than “This Time is Different” for a discussion of how we addressed those challenges). But recognize that our valuation methods haven’t changed. The same methods accurately identified extreme overvaluation in 2000, improved enough to encourage a constructive shift in early 2003 as a new bull market was emerging, accurately identified extreme overvaluation in 2007, improved enough to recognize undervaluation in 2008-2009, and now identify the third valuation bubble in 15 years – one that we expect will conclude no better than the first two.
As I frequently emphasize, a resumption of favorable market internals and a retreat in credit spreads would suggest a fresh shift toward risk-seeking investor preferences that could defer the immediacy of our downside concerns. Unfortunately, that would not improve the dismal long-term returns that are already baked in the cake of current valuations. Presently, we observe obscene valuations coupled with evidence of a shift toward increasing risk-aversion among investors. For that reason, I continue to believe that the present moment will likely be remembered among a small handful of the very worst points in history to invest in equities from the standpoint of prospective return and risk.
Profit margins – is the ladder starting to snap?
One of the central features of popular valuation measures such as price to forward earnings is that they take profit margins at face value. Year-ahead earnings estimates have the additional feature that analysts can (and nearly always do) base their estimates on the assumption that margins will improve in the future. We know from a century of evidence that correcting for the variation in profit margins produces valuation measures that are far better correlated with actual subsequent market returns. But when profit margins are elevated, the temptation to take them at face value (or extrapolate them to even greater extremes) seems too much for Wall Street to resist.
Though it’s typically not appreciated by investors, the deficits of one sector of the economy are (and must be) offset by surpluses in other sectors. As detailed in An Open Letter to the FOMC: Recognizing the Valuation Bubble in Equities, this relationship is captured by what economists know as the “savings-investment identity,” which can also be expressed as the Kalecki profits equation:
Corporate Profits = (Investment – Foreign Savings) – Household Savings – Government Savings + Dividends
In this equation, investment means gross domestic investment – real things like capital equipment and housing. Foreign savings are essentially the inverse of the U.S. current account balance. The reason I place parentheses around (Investment – Foreign Savings) is that in U.S. data, increases in gross domestic investment are typically accompanied by deterioration in the current account balance, while an increasing