Take a few moments to flip through your local TV news channels this evening and you will surely notice that all the weather forecasts are remarkably similar. It is extraordinarily rare to see one meteorologist forecast hot and humid weather while another urges you to dust off your snow boots for the same zip code. Yet tomorrow morning at the watercooler, you can expect to hear how wrong the local weather forecasters always are, with one channel predicting a high of 52°F and mostly sunny conditions and another predicting a high of 53°F and only partly sunny conditions! Yes, you knew what jacket to wear, but that’s beside the point. If only those pesky weather forecasters could get their stories straight!

Now imagine if we turned the same scrutiny on today’s stock market prognosticators. Flipping through the business channels, you’d see forecasters earnestly professing that the sky is falling. Yet tomorrow they may be predicting that the market is going to the moon. Just yesterday they proclaimed that the market was a historic buy, but today forecasters are counseling investors to change course and run for the exits.

We treat these financial prophets with great respect, as if they were hard at work in their corner offices, splitting the proverbial atom. The reality, unfortunately, is much murkier.

In this age of “big data,” it is statistically possible to prove just about any hypothesis; as with anything, the frame of reference is key. If I want to say that stocks are overvalued today, I can look at the Shiller CAPE metric (the cyclically adjusted price/earnings ratio), which divides today’s price by the last ten years of earnings, and show that an investor is paying more per dollar of earnings than in all but 15% of the ten-year periods going back to 1881. Likewise, I could look at the trailing 12-month P/E ratio starting in 1995 and show that stocks are at all-time low valuations, as a result, of course, of the technology bubble in the late 1990s. How do you know which number is more accurate? Or whether either is even relevant in today’s environment?

The Vanguard white paper Forecasting stock returns: What signals matter, and what do they say now? (Davis, Aliaga-Díaz, and Thomas, 2012) does a great job of framing the importance and shortcomings of various methods of forecasting. Most important, the authors conclude that stock returns are essentially unpredictable at short horizons and that many commonly used metrics—such as dividend yields, economic growth, and the so-called “Fed model”—turn out to be very poor predictors of future returns.

The P/E ratio, on the other hand, has proved to show the most promise, which is why analysts make such a big deal over what today’s valuations tell us. However, even with the P/E metric, the “noise” often overwhelms an assumed relationship. For example, the S&P 500 Index’s 12-month P/E ratio of 16.25 would place U.S. stocks at just under the long-term average (since 1926) of 16.72. Does this mean that we should therefore expect average stock returns over the coming years? Historically speaking, a P/E ratio at this level has led to a very wide range of outcomes, with real returns ranging from 0% to 15% over the following decade. And the Vanguard paper makes it clear that, when at an “average” level, the forecasting power of the P/E ratio erodes dramatically. In one sense, the returns investors receive over the ensuing years will likely be driven, first, by “animal spirits,” leading to P/E expansion or contraction, and, second, by the fundamentals of the stocks that make up the index.

Understanding the forecasting power and shortcomings of the P/E ratio is even more important when looking at comparisons between different countries and indexes. Many of today’s popular indexes are fairly new, making historical comparisons difficult.

For example, P/E data for the MSCI Emerging Markets Index is available starting in 1995, yet many forecasters compare the historical average of this index with data for the S&P 500 Index, which has been around in some form for almost 100 years. Using this logic, it is easy to say that the Emerging Markets Index’s P/E of 12.50 is “cheaper” than the S&P 500’s P/E of 16.25. However, when looking at a common start date of 1995, both indexes are valued in the lowest 20% of observations. The case could be made that the S&P 500 is actually much cheaper based on the fact that it is trading at a nearly 60% discount to its average P/E over the shorter period (1995–2012), whereas the Emerging Markets Index is trading at only a 20% discount. But we showed earlier that, on a longer-term basis (1926–2012), the S&P 500’s P/E looks average.

So what, if anything, can or should we say about stock markets over shorter time periods? And what does this all mean?

First, maybe we should be a little more forgiving of our weather forecasters who, after all, usually don’t send us out of the house in shorts during a blizzard. Perhaps we should save a little skepticism for our investment forecasters instead.

Second, most investors are better served by tuning out the noise and maintaining a balanced, diversified portfolio because, at the end of the day, we just do not know what the future will hold.

Charley Ellis, investing pioneer and former Vanguard board member, said it best: “Like the weather, the average long-term experience in investing is never surprising, but the short-term experience is constantly surprising” (Winning the loser’s game: Timeless strategies for successful investing. New York: McGraw Hill, 2010. p. 68). Words to invest by.

I would like to thank my colleagues Michael DiJoseph and Chris Philips for their contributions to this blog post.

 

Notes:

  • All investing is subject to risk, including possible loss of principal.
  • Diversification does not ensure a profit or protect against a loss.