"Strictly speaking, there can be no such thing as an ‘investment issue’ in the absolute sense, i.e., implying that it remains an investment regardless of price… Had the same attitude been taken by the purchaser of common stocks in 1928-1929, the term ‘investment’ would not have been the tragic misnomer that it was. But in proudly applying the designation ‘blue chips’ to the high-priced issues chiefly favored, the public unconsciously revealed the gambling motive at the heart of its supposed investment selections."

– Benjamin Graham & David Dodd, Security Analysis


At the market open of Friday, January 24, our estimate of likely 12-year nominal total returns for a conventional passive investment portfolio (60% S&P 500, 30% Treasury bonds, 10% Treasury bills) fell to just 0.04% annually, below even the previous record of 0.34% set in August 1929. This extreme reflects the combination of record equity market valuations and depressed interest rates. That’s not an “equilibrium” situation. It’s a combination that joins insult with injury, creating weak prospects for the future returns of passive, diversified buy-and-hold strategies, across the board.

Understand this. The more glorious this bubble becomes in hindsight, the more dismal future investment returns become in foresight. The higher the price investors pay for a set of future cash flows, the lower the return they will enjoy over time. Whatever they’re doing, it’s not “investment.”

The chart below shows our estimate of 12-year total returns on a conventional passive investment mix (blue) along with actual subsequent returns. As I’ve often noted, we use a 12-year horizon because that is the point where the “autocorrelation” of reliable valuation measures typically hits zero, meaning that overvaluation or undervaluation at one point becomes uncorrelated with later valuations. So “mean reversion” is most likely on a 12-year horizon. That said, valuation extremes like 1929 and 2000, along with lesser extremes like 2007, have generally been followed by profound market losses (and associated spikes in expected future returns) over a much shorter time frame.

I’ve regularly observed that at the peak of every valuation bubble, recent returns have invariably been greater than one might have projected several years earlier, because the advance to hypervaluation produces temporary returns that are later erased. That’s why, for example, actual returns were higher during the 12-year period from 1988-2000 than one would have projected in 1988, and it’s why actual returns have been higher during the recent 12-year period than one would have projected 12 years ago. The problem, as investors should recall from the 2000-2002 and 2007-2009 collapses, is that the extra returns from those “errors” are invariably erased.

Still, it’s useful to understand how far below-average the returns of the S&P 500 are likely to be in the coming years even if valuations don’t normalize at all, and instead remain at their present extremes forever. As I’ve detailed before (and review below), the U.S. economy is presently running at a structural real GDP growth rate of only about 1.6%, reflecting the combination of demographic labor force growth and trend productivity. That’s the real economic growth that we would observe if the rate of unemployment was simply held constant at current lows indefinitely.