Paper on Market Forecasting Wins AQR Award

(May 29, 2012) — Research on predicting market returns more accurately over one-month and one-year time horizons has earned AQR Capital Management’s first annual Insight Award.

The award recognizes authors Bryan Kelly, Ph.D., of the University of Chicago Booth School of Business and Seth Pruitt, Ph.D., of the Federal Reserve Board of Governors, for their paper titled “Market Expectations in the Cross Section of Present Values”. The thrust of the paper: market discount rates are more volatile and less persistent than previously believed. Disaggregated valuation ratios — such as price-dividend ratio or book-to-market ratio at the individual stock level, rather than at the level of the entire US market — can be used to predict market returns more accurately over one-month and one-year time horizons.

Kelly and Pruitt’s research addresses the real-world challenges investors face, such as asset allocation, security selection, portfolio implementation, risk management and was chosen as the best of several hundred submissions from 24 countries, AQR notes. 

Kelly tells aiCIO that while aggregate measures such as the market’s overall book-to-market ratio can contain predictive information, significantly more predictability can be gained when using price ratios of individual stocks. “We stress that people have thought a lot on whether there’s predictability at all, and the evidence has been mixed. We wanted to address that question in a more comprehensive way than they looked at the data,” Kelly says.

The goal, according to Kelly, is to show that there’s a vast, rich set of data used for forecasting investment returns. The authors propose a simple way to conduct forecasts with a variety of data, uncovering market predictability with evidence.

The paper sheds light on the way assets are valued, the authors say. “If returns are predictable, the way people discount assets, due to changes in perceived risks or changes in the way they feel about risk, are also predictable. We document that investors change their valuations in a way that induces much stronger return predictability than has been previously documented,” Kelly asserts.

While previous measures to predict investment returns have relied on aggregate valuation ratios, such as price dividend ratios of the US market, valuation ratios at the individual asset level are more powerful, the paper asserts. “Explaining market price behavior of the US capital stock is among the most fundamental challenges facing economists…Our results are robust across a variety of cross sections, out-of-sample procedures and hold in both US and international data,” the paper says. “The cross section of valuation ratios, as present value identities imply, hold a wealth of information about investor expectations. A deeper understanding of economic fundamentals that drive these valuation ratios is a promising avenue for future research.”

Commenting on the award, Cliff Asness, Managing Principal at the firm, and a Chicago Booth MBA and Ph.D., says: “We created this award both to give back to the academic community, and to make sure we stay absolutely current with the latest and greatest research. Bryan’s work easily passes that last test. It is a novel methodology that has the potential to improve our ability to, among other things, make market forecasts.”