“Observation over many years has taught us that the chief losses to investors come from the purchase of low-quality securities at times of good business conditions. The purchasers view the good current earnings as equivalent to ‘earning power’ and assume that prosperity is equivalent to safety.”
“I got wiped out personally in 1968, which was the last really crazy, silly stock market before the Internet era… I became a great reader of history books. I was shocked and horrified to discover that I had just learned a lesson that was freely available all the way back to the South Sea Bubble.”
“Cahm viss me eef you vahn to live.”
Arnold Schwarzenegger, Terminator 2: Judgment Day
We’ve recently emphasized that our estimates for probable S&P 500 nominal total returns have now declined below zero on every horizon of 7 years and shorter. At longer horizons, the 6.3% growth rate that we’ve assumed for nominal GDP over the coming years will begin to bail investors out given enough time, and as a result, our projection for 10-year S&P 500 nominal total returns peeks its head up above zero, at about 2.4% annually from current levels. Looking out 15 years, the expected 15-year total return approaches 4.4% annually, and at that horizon, investors are unlikely to lose money even if actual returns are a standard deviation below our expectations. To the extent that 6.3% growth in nominal GDP seems too high (and there are certainly reasons to think so), just reduce those annual return projections accordingly.
The key point is this – everything that investors can expect to obtain from selling stocks 7 years from now is already on the table today. Valuations might move higher over the very short run, but at present valuations, investors would require a positive surprise more than one standard deviation above expectations just to pull the likely 2-year return out of negative territory. The chart below provides a quick summary of our return expectations for the S&P 500 – from current price levels – over a variety of investment horizons. I emphasize the phrase “from current price levels,” as a significant retreat in valuations is likely to dramatically shift this profile, as it has over the completion of every market cycle in history.
I was grateful last week for some very kind words from Jeremy Grantham (Trades, Portfolio) at GMO, who is certainly one of my mentors despite having never actually met. Back in the late-1990’s, one of the ways that Grantham operationalized the word “bubble” was to ask if valuations were at least two standard deviations above their norm. On a bell-curve, about 95% of your observations will be within two standard deviations one way or another, so only about 2.5% of your observations will be beyond “two sigma” on the upside.
Grantham observes that Robert Shiller’s “cyclically adjusted P/E” or CAPE is presently less than two sigmas above its historical norm, allowing for the potential for a bubble – under that definition – to carry the S&P 500 to 2250 if we allow an ugly CAPE to turn really ugly. Not that Grantham is actually bullish about this prospect, noting instead that it can be useful for prudent investors to imagine future pain “so that they can more easily process it and be less likely to do something foolish.” Still, for reasons noted below, we believe that stocks are already more overpriced than meets the eye.
We very much agree with Grantham’s sentiment in that investors can only be expected to adhere to a given discipline if they fully understand the potential risks that might emerge over the course of the market cycle. The greatest discomfort for us is that in speculative, overvalued, overbought, overbullish markets, we’ll often look like idiots, lose credibility, and later recover a flood of followers who were bloodied in a completely predictable collapse that makes us look like evil geniuses. Years ago, Grantham noted the same tendency to “arrive at the winning post with good long-term results and less absolute volatility than most, but not necessarily the same clients that we started out with.” Having addressed our earlier stress-testing concerns, we certainly expect a much easier time in future cycles even if Fed-induced bubbles become the rule.