Are Stocks a Screaming Buy? It Depends on Your Time Horizon

Time to buy stocks? If you think long-term, it’s almost always time.

Buyers offering their cash
By SHARE

When investing, it's good to beware of following the herd, because the herd, we're supposed to have learned, can go wildly astray. When it's not chasing past outperformance, it's fleeing past underperformance—hence the reported aversion to stocks in recent years.

But has the herd, frightened into bonds for much of this year, fled so far into the hills that it's now underinvested in stocks?

By some measures, the answer is yes. The most stock-loving investors have pulled back on their positions. The Investment Company Institute reports that equity aversion rose across all age groups above 30 in the decade through 2011. The proportion of each cohort that has at least 80 percent of its portfolio invested in stocks fell during that period—by a bigger proportion the older the investor. That's not the same thing as being under underinvested in stocks, though. Vanguard reports that equity (domestic and foreign) accounted for 50.2 percent of assets under management by U.S. open-end funds and ETFs as of July 31—about the average for the past 20 years.

[Read: Are Individual Investors Destined to Fail?]

The question is, should equity allocations be even bigger? Fidelity argues in a recent report that stocks are cheap by several measures. One is the S&P 500 index's "current" price-earnings ratio—based on the trailing 12 months' earnings—which was 13.8 at the end of the second quarter, below the long-term average of around 16. If you rank all the quarterly P/E ratios since 1926, the second quarter's is in the second-cheapest quintile, says Fidelity. Historically, buying at these P/Es has produced real returns of 8.1 percent over the subsequent five years and 10.1 percent over the subsequent decade, well above the long-term averages of 6.5 percent and 6.9 percent, respectively, says the study.

True, the cyclically adjusted, or CAPE, ratio, which measures prices against the previous five or 10 years' earnings, makes stocks look more expensive than does the current method. On the other hand, some would caution, a five-year measure would include the unusually deep recession of 2007-09.

"That's not representative of what corporate America can earn," says John Fox, co-manager of Fenimore Asset Management's $740 million FAM Value fund (symbol: FAMVX).

Fidelity notes other measures of market value that hint of bargains to be had. The "equity risk premium" (ERP)—the difference between the earnings yield on equities and the yield on the "risk-free" 10-year Treasury—is historically quite high, indicating inexpensive stocks and above-average returns in coming years. Though there is some subjectivity built into ERP calculations, the idea is that there's room for stock prices to rise without eliminating a reward for the additional risk that stocks entail.

[Read: How a Dogged Focus Helps Investors.]

To be sure, investors can find plenty of reasons to shun stocks in the short term. Uncertainty about whether the United States will get its fiscal house in order; wariness about Europe; prolonged sluggishness in both the United States and Europe. But Fidelity notes similar periods in the past: the deflationary 1930s and the years 1942-51, when the Fed kept rates unusually low, P/Es were down and ERPs were high. Both periods signaled higher subsequent five- and 10-year equity returns, Fidelity says.

So lots of signals are blinking "buy." The question then becomes: How reliable are the signals? The short answer: not very. Christopher Philips, senior analyst at Vanguard's Investment Strategy Group, says the CAPE measure, for example, explains about 35 percent of the variation in real returns over a 10-year period. That leaves 65 percent unexplained.

"The challenge is that there's so much noise in the data that even when you look at long-term relationships, there's more uncertainty than certainty with respect to predictability," says Philips. "Everyone's crystal ball seems to be cracked."

Also, a lot of that 35 percent is driven by what analysts call "tail" events—historical anomalies that skew the results. Strip out the high P/Es of the dot-come bubble and the low P/Es of the recessed early '80s, says Philips, and the 35 percent falls to 15 or 20 percent. Reduce the period from 10 years to one, and it disappears altogether.