Wrecking Ball

As of last week, the valuation measures that we find most strongly correlated with actual subsequent 10-12 year S&P 500 total returns in market cycles across history now range between 140-170% above their pre-bubble norms. While valuations at those pre-bubble norms are closely associated with prospective S&P 500 total returns in the range of 10% annually, valuations at current levels are associated with expected returns of about -2.5% annually on a 10-year horizon, and roughly zero on a 12-year horizon. Nearly every market cycle in history has brought these valuation measures back toward their historical norms, and no cycle (even those associated with quite low interest rates) has failed to bring them within about 25% of those norms. Assuming that valuations do not breach historical norms (as they did even in the most recent market cycle), the associated downside expectation for the S&P 500 over the completion of the current market cycle now runs between -48% and -63%.

While it’s reasonable to believe that low interest rates justify higher than normal valuations, and lower-than-normal expected returns for stocks, we doubt that investors are comfortable with zero or negative returns being their definition of “justified.” What’s really happening is what always happens late in a speculative cycle, which is that the extrapolation of the recent half-cycle advance encourages investors to believe that the stock market simply “pays” high returns as if they were interest payments. It generally does not enter the speculator’s mind how strongly the ratio of market capitalization to GDP is correlated with actual subsequentmarket returns (though we prefer to use corporate gross value added including estimated foreign revenues, which creates a proper apples-to-apples measure). Sustained capital gains in the broad stock market, materially in excess of nominal economic growth, are typically borrowed from the future and repaid over the completion of the market cycle.

What the “low interest rates justify high valuations” mantra has really done is to ensure that all asset classes are now priced at levels that are likely to generate dismal returns in the coming years. The chart below shows our best estimate of 12-year prospective nominal total returns on a conventional portfolio mix invested 60% in the S&P 500 Index, 30% in Treasury bonds, and 10% in Treasury bills. The red line shows actual subsequent 12-year returns on this mix. The current estimate is only about 1% annually, with the Treasury bond and T-bill components responsible for virtually all of that return, as the expected return on the S&P 500 component is close to zero.

It’s useful to keep in mind that while continued low nominal economic growth may support a continued period of interest rates well-below their historical norms, the return projections above would be largely unaffected in that event. That is, low nominal growth and low interest rates would offset each other almost entirely. This isn’t a random outcome but an implication of both asset pricing arithmetic and historical data. See Rarified Air: Valuations and Subsequent Market Returns for the data and arithmetic on this.

What’s often missed in the “low interest rates justify higher valuations” argument is that this proposition assumes that future cash flows and growth rates are held constant. If instead low interest rates emerge as a consequence of low expected nominal growth, valuation multiples should not be affected at all, yet prospective returns will still be lower anyway (you can demonstrate this to yourself by reducing both r and g equally in a basic dividend discount model). To elevate valuation multiples in this situation is to drive prospective returns down twice; a mistaken form of double-counting.

Worse, recall that since earnings-based measures are heavily affected by cyclical variations in profit margins, earnings rarely represent useful “sufficient statistics” for the very long-term stream of cash flows that investors are actually buying (see Exhaustion Gaps and the Fear of Missing Out for a comparison of the relative reliability of various market valuation measures). Responding to low interest rates by elevating P/E multiples, without accounting for depressed long-term nominal economic growth prospects, and to base those P/E multiples on earnings that embed elevated profit margins (driven by a shortfall between wage growth and productivity that has already begun to reverse as the unemployment rate pushes 4.4%) is to drive prospective returns down three times; a mistaken form of triple-counting that is precisely why expected 10-12 year market returns are presently so weak.

The pendulum of speculation has swung far and has grown increasingly grotesque in order to reach this point. It will be a wrecking ball when it returns to earth.

It came in like a wrecking ball!
Did not expect that margin call!
(apologies to Miley Cyrus)