That includes a chapter in the forthcoming book Volatility and Time Series Econometrics (published by Oxford University Press), which explains how economic cycle volatility drives stock market volatility rather than the other way around. Previous research may have considered the association between the equity markets and economic activity, but this piece (co-authored by professors Francis X. Diebold and Kamil Yilmaz) lays out the relationship on a deeper level.
The authors have found among other things that countries with relatively high gross domestic product (GDP) growth volatility tend to have more volatile stock markets. Another insight-and a surprising one-is that while the stock market is generally pricing securities based on expectations about the future, it's usually the reverse with volatility, which is to say that Mr. Market seems to react to volatility. Part of the reason is that every country's GDP volatility is the main source of its stock market volatility: The former drives the latter, and so the market seems to digest the results after the fact.
Recognizing that volatility flows from the economy to the equity market is hardly a shortcut to fast profits. Then again, no lone piece of information about market behavior tells us everything. The practical value of studying macroeconomics and asset pricing behavior comes from synthesizing intelligence dispensed from the two halves. This isn't easy, but the lesson is clear: The biggest obstacle for investors is failing to see the forest for the trees.
Financial economists have spent quite a bit of the last 30 years identifying individual factors that help explain returns in the capital markets. But analyzing the individual components in connection with one another is still quite new. The early results are encouraging, and recent studies suggest that we're just beginning to understand how these factors, when taken together, can deliver better estimates of the equity risk premium.
For a glimpse into the new world order, take a look at this forthcoming study in the Review of Financial Studies: "Out-of-Sample Equity Premium Prediction: Combination Forecasts and Links to the Real Economy," by professor David Rapach and two co-authors. The paper shows that analysts' forecasting ability improves when factors are used collectively. It's essential to combine them since any one measure's ability to offer relevant insight about the future waxes and wanes over time.
Consider the dividend yield. Over time, this measure has proved constructive for analysts estimating stock market returns for medium-to-long-term horizons, as we discussed in the March 2009 issue of Financial Advisor ("We're All Value Investors Now"). In essence, yield tends to correspond with expected return. When the yield falls, expected return declines too, and vice versa. The problem is that return forecasts based on dividend yield alone aren't often timely. That has prompted some analysts to dismiss yield's relevancy as a factor when forecasting the equity market's performance.
On first glance, this criticism looks reasonable, especially when you consider the dramatic examples in the 1990s. The yield on broad market indices such as the S&P 500 declined for much of the decade-yet stock returns were unaffected. In fact, the second half of the 1990s was nothing short of spectacular.
No wonder, then, that some investors are unimpressed with dividend yield. It's true that dividend yield can't be trusted as a short-term indicator of equity return. What's more, simple statistical tests that look for obvious connections between the market's current yield and subsequent returns are unimpressive in the aggregate. But that ignores the powerful economic logic that binds dividends and prices over longer horizons, as discussed at some length by professor John H. Cochrane in "The Dog That Did Not Bark: A Defense of Return Predictability," in the Review of Financial Studies, July 2008.
Meanwhile, dividend yield is more valuable when used alongside other market signals, such as the Treasury market's yield curve. Consider again the dividend yield in the 1990s, but this time with the yield curve, which tends to invert (putting short rates above long rates) ahead of recessions. Although the dividend yield fell for much of the 1990s, the curve remained upward sloping during most of the decade, suggesting that economic growth was likely. Then in April 2000-just after U.S. stocks reached an all-time high-the curve inverted for the first time in more than ten years. A second, longer round of inversion dominated much of the year's second half, based on daily yields for the ten-year Treasury less three-month T-bills. The message: Economic turmoil was coming. As fate would have it, a bear market in stocks had arrived in 2000 and would last through 2002. Unsurprisingly, the U.S. also endured a brief recession in 2001.
In other words, the inversion of the yield curve in mid-2000 appeared to confirm the timeliness of the warning embedded in the long-running decline in dividend yield. Why not simply ignore yield and use the Treasury curve for projecting the equity risk premium? Because we're unsure that the curve will remain a timely warning of things to come, even though it has been over the past 40 years.