By Mike Piper
MGM/Courtesy Everett Collection
A reader writes in, asking:
“The stock market’s tumble over the last week combined with the fact that stock valuations are still so high makes me wonder about the appropriate way to respond. Time to take some money off the table? I suspect I know what you’ll say, but I’m interested to hear anyway.”
The total U.S. stock market (as measured by the Vanguard Total Stock Market Index Fund /zigman2/quotes/202876707/realtime VTSMX +3.59% ) fell by less than 10% last week. If that made you super nervous, that’s a good indication that your stock allocation is too high. A 10% decline should not be a big deal — especially when it comes after a nine-year bull market during which the value of U.S. stocks rose by roughly 400%.
If a decline of less than 10% makes you nervous at all, imagine how you’ll feel about a 30%, 40%, or 50% decline. The goal of asset allocation is to craft a portfolio with which you’d be able “sit tight” (or possibly even rebalance into stocks) during a full-blown bear market.
Making use of market valuations
It’s true that the stock market is still highly valued relative to historical norms. (This should not be a surprise, given the huge returns over the last nine years.)
But how useful is that information for the purpose of predicting returns going forward?
The following chart shows the correlation between the S&P 500’s /zigman2/quotes/210599714/realtime SPX +1.45% valuation (as measured by PE10) and inflation-adjusted returns for periods of various lengths from 1926-2017. As you would expect, the correlation is always negative, which means that the higher the market’s valuation at any time, the lower we should expect returns to be going forward.
But the correlation between PE10 and ensuing short-term returns has been pretty weak. For instance, the correlation coefficient between PE10 and 1-year returns is just -0.22. The correlation is quite a bit stronger if we look at 10-year real returns (-0.63 correlation) or 20-year real returns (-0.75 correlation).
In other words, valuation levels are not very good at predicting short-term market returns. They are much better at predicting longer-term returns.
But even if we have good reason to suspect poor returns over the next, say, 10 years, a 10-year period of poor returns could come in a lot of forms. The market could be roughly stagnant, with inflation taking a toll. Alternatively, we might see another seven years of gangbuster returns, followed by a super bad bear market for three years. Or we might see a two-year bear market, followed by four years of good returns, then another four-year bear market. And so on. (Or the next 10 years could be a period for which valuation isn’t even predictive in the first place! A negative 63% correlation is still far from perfect.)
Point being, we never know what’s about to happen in the near term. So valuations aren’t very useful for trying to “dodge” a bear market, so to speak.
But because they do have decent predictive power over the long term, valuations are useful for questions such as, “how much should I be saving per year?” And, “how much can I afford to spend per year in retirement?”
And with today’s high valuations, we should expect pretty modest returns — suggesting that high savings rates (for those in their accumulation years) and low spending rates (for those in their retirement years) are probably prudent. This was true a year ago, and it’s still true today.
Mike Piper is the author of the “Oblivious Investor” blog, where this was first published — “Worrying about Market Declines and High Valuations.”