Behind the Potemkin Village

There’s an apocryphal story that in 1787, during the journey of Empress Catherine II to Crimea, Prince Grigory Potemkin, the governor of the region, erected fabricated villages along the Dnieper River, which would be disassembled after she passed by, and rebuilt again downstream overnight.

When one examines the collapses of the tech bubble and the housing bubble, it’s evident that one of the central elements of those collapses was the gradual recognition by investors that the overvalued pieces of paper they were holding were actually little Potemkin Villages; temporarily glorious and impressive on the surface, but backed by much less than investors had imagined was there. What sort of “catalyst” is needed for a Potemkin Village or a Ponzi scheme to disappoint? Only the gradual or sudden discovery of the reality behind it: the recognition that there is no “there” there.

Investors presently appear to be taking past investment returns and economic growth at face value, without considering their underlying drivers at all. My impression is that while the U.S. may very well encounter credit strains or other economic dislocations in the coming years, nothing is actually required for yet another equity market collapse except the gradual recognition by investors of a reality that already exists. Strip away the effects of short-term cyclical factors that have now been largely exhausted, and it becomes clear that the financial markets have again become a Potemkin Village.

I’ve previously detailed the argument that if interest rates are low because growth is also low, no increase in market valuation is “justified” at all by the lower interest rates, and yet future returns on stocks will be commensurately lower anyway (see the Geek’s Note in Valuations, Sufficient Statistics, and Breathtaking Risks to understand this from the standpoint of discounted cash flows). In that situation, elevating market valuations results in a reduction of prospective stock market returns by a far greater amount than interest rates have been reduced.

Market returns don’t just emerge from nowhere. They are driven by the sum of three factors: growth in fundamentals, income from cash distributions, and changes in valuations (the ratio of prices to fundamentals). Since 1960, for example, the S&P 500 has enjoyed an average rate of return of about 10% annually, which derives from three main components: growth in earnings (and the overall economy) averaging about 6.3% annually, dividend income averaging about 3.0% annually, and a gradual increase in price/earnings multiples that has contributed about 0.7% annually to total returns. One can also make a similar attribution using other fundamentals. For example, the 10% annual total return of the S&P 500 since 1960 also derives from growth in S&P 500 revenues averaging 5.7% annually since the 2000 peak, dividend income averaging about 3.0% annually, and a much steeper increase in the S&P 500 price/revenue ratio contributing 1.3% annually (taking the current price/revenue multiple to the same level observed at the 2000 market peak).

Consider these drivers today. Combining depressed growth prospects with an S&P 500 dividend yield of just 2.0%, the likelihood is that over the coming 10-12 years, even a run-of-the-mill reversion of valuations will wipe out the entire contribution of growth and dividend income, resulting in zero or negative total returns in the S&P 500 Index on that horizon, with an estimated interim market loss on the order of -60%.

Here are the facts: over the past several decades, due to a combination of demographic factors and persistently slowing productivity growth, the core drivers of real U.S. GDP growth have declined toward just 1% annually, with a likely decline below that level in the coming 10-12 years. Indeed, in the absence of any recession, U.S. nonfarm productivity growth has averaged just 0.8% annually since 2010 and 0.6% over the past 5 years, while the U.S. Bureau of Labor Statistics estimates labor force growth of just 0.3% annually in the coming years (which would be matched by similar growth in employment only if the unemployment rate does not rise from the current level of 4.3%). Add 0.6% to 0.3%, and the baseline expectation for real GDP growth is just 0.9%. Nominal growth is likely to be similarly weak.

While S&P 500 earnings growth has slightly outpaced revenue growth over the past two decades because of rising profit margins, recent record profit margins have now stagnated and have begun to retreat, resulting in the likelihood that earnings growth will match (at best) or even lag, overall economic growth in the years ahead. At the same time, the valuation measures we find most reliably correlated with actual subsequent S&P 500 total returns now average between 150-170% above historical norms that they have approached or breached by the completion of every market cycle in history. For a review of the historical reliability of these measures and popular alternatives, see the table in Exhaustion Gaps and the Fear of Missing Out.

What if inflation bursts durably higher? Wouldn’t one expect higher future nominal returns over the long-term? The answer is yes, but the present valuation premium would still not be justified. Rather, as we observed in the mid-1970’s, the likely first response of the market would be for valuation multiples to surrender their premium, and possibly move to a substantial discount, after which higher long-term returns would emerge. So inflation or no inflation, I don’t expect that any plausible economic outcome will prevent a wholesale market collapse over the completion of the current cycle.

Keep in mind that even if the most reliable valuation measures we identify were to move from 170% above their historical norms to still 35% above those norms, the move would imply a -50% market loss (as we observed in both the 2000-2002 and 2007-2009 cycle completions). The longer-term effect of that normalization would be to subtract about -6.7% from 10-year S&P 500 total returns, and about -5.6% from 12-year total returns (subtracting -9.5% and -7.9%, respectively, if valuations were actually to visit their historical norms 10-12 years from now, even if they never move below those historical norms again). Add in about 2% of dividend income, and a few percent in nominal growth, and the market would still be struggling to claw out 10-12 year total returns of zero.