Singapore Markets closed

When Does Indicator-Watching Become Voodoo Investing?

Chris Gay

The stock market will plunge next year, take it from the National Oceanic and Atmospheric Administration. How do we know? Because New York City received below-average rainfall this summer, and even Hurricane Sandy couldn't make up the deficit. True fact: There is a surprisingly strong correlation between rainfall in New York and the performance of the stock market, according to NOAA data assembled by Lakewood, Wash.-based analyst Tom McClellan.

If rain trends don't nail you, the solar system just might. The website Money Morning ("Only the News You Can Profit From") quoted market analyst Arch Crawford in July as saying his "astro indicators" show a market implosion sometime before March. As of this writing, we're about halfway through the "Mars/Uranus Crash Cycle," so there's still time to bail if you hurry. (Unless, of course, the Mayan Calendar Gods pull the celestial power plug on December 21. Let me get back to you.)

On the credibility spectrum, these theories fall somewhere between pseudoscience and straight-up flapdoodle. But they raise an interesting question: Are there stock-market indicators reliable enough to hang your portfolio on?

[Read: Fiscal Cliff Notes: Investing Ideas for the Coming Months.]

In the short term, no. In the longer term, also no, though it's worth noting that some long-term indicators are better than others, as the folks at Vanguard recently showed.

Vanguard studied 15 commonly used stock-market indicators (plus one random metric--rainfall--just as a "reality check") and ranked them by their power to predict stock prices. The most reliable, you'll be pleased to know, is not solar radiation levels or who wins the Super Bowl. It's plain old earnings.

Looking at the broad U.S. stock market over the period 1926 to 2012, Vanguard finds that price-to-earnings (P/E) ratios at any point in time explain about 40 percent of the variation in stock prices over the subsequent 10 years. In particular, the cyclically adjusted P/E ratio (CAPE)--which measures stock prices against the previous 10 years' earnings--explains 43 percent of subsequent variation, while the conventional P/E (trailing 12 months' earnings) explains 38 percent.

Note, however, than even the best indicators explain less than half of what moves stock prices. As Vanguard is quick to note, "the fact that even PEs--the strongest of the indicators we examined--leave a large portion of returns unexplained underscores our belief that expected stock returns are best stated in a probabilistic framework, not as a 'point forecast,' and should not be forecast over short horizons."

[Read: Are Second Terms Good for Stocks?]

What's more, notes Ryan Detrick, a senior technical strategist at Schaeffer's Investment Research, "the trend of P/E ratios can be more important than the absolute level. Just because something is low doesn't mean it can't go lower."

And the other indicators? The ratio of government debt to GDP explains 23 percent of variation. Dividend yields: 19 percent. The so-called Fed model (which compares stock yields to those on long-term government bonds): 16 percent. Corporate profit margins, trailing one-year stock returns, and consensus GDP growth: zero, zero, and zero. Even rainfall pulls 6 percent.

Of course, it's just possible that Vanguard hasn't considered the full range of possibilities. There's no mention of what Warren Buffett has called the single best measure of valuation: total stock-market capitalization as a percentage of GDP. At the end of the third quarter it was 94, which amounts to flashing yellow.

Other strange omissions: The Hemline Model (which says shortening skirts evince rising optimism), the Boston Snow Theory (a white Christmas means rising prices), or the Billboard Hypothesis (something about "beat variance" in top-100 tunes).

Should the average investor even pay attention to this sort of thing? Except as a loose reference point, probably not. It's good to have a sense of where we are on the great stock/bond/cash ellipse, and on that basis we should be looping around to equities right about now. "We've had one of the poorest decades [for stocks] in the past hundred years," says Schaeffer's Detrick of the lamentable aughties. "Historically, when you've had a bad decade, the next one or two decades are pretty good."

[In Pictures: The S&P's 10 Worst Trading Days]

Unless you know a good reason to defy the eternal laws of mean reversion, this might be a good time to stop digging yourself into a fixed-income hole. "Bonds are a very crowded trade," says Detrick. "Yields are at generational lows. That's not the best time to put money into bonds."

But stocks? What if the rain theorists are on to something? Lakewood's McClellan says you can't be too sure. "What's happening is that the Federal Reserve and quantitative easing are stomping all over natural fluctuations," he says. "Whether that is good or bad is a different question."

More From US News & World Report