Industry Momentum: The Role Of Time-Varying Factor Exposures And Market Conditions
Friedrich-Alexander-Universität (FAU) Erlangen-Nürnberg
October 6, 2015
This paper focuses on momentum strategies based on recent and intermediate past returns of U.S. industry portfolios. Our empirical analysis shows that strategies based on intermediate past returns yield higher mean returns. Moreover, strategies involving both return specifications exhibit time-varying factor exposures, especially when applying the Fama and French (2015) five-factor model. After risk-adjusting for these dynamic exposures, the profitability of industry momentum strategies diminishes and becomes insignificant for strategies based on recent past returns. However, most strategies built on intermediate past returns remain profitable and highly significant. Further analyses reveal that industry momentum strategies are disrupted by periods of strong negative risk-adjusted returns. These so-called momentum crashes seem to be driven by specific market conditions. We find profits of industry momentum strategies to be related to market states and to the business cycle. However, there is no clear evidence that industry momentum can be linked to market volatility or sentiment.
Industry Momentum: The Role Of Time-Varying Factor Exposures And Market Conditions – Introduction
The momentum phenomenon has attracted the attention of researchers and practitioners in recent decades. Jegadeesh and Titman (1993, 2001) first documented the profitability of the so-called common price momentum strategy in the U.S, which involves buying recent winner stocks and selling recent loser stocks. Rouwenhorst (1998), Chan et al. (2000) and Asness et al. (2013), among others, show that momentum based on individual stocks can be found in countries worldwide. Asness et al. (2013) find evidence of momentum in different asset classes, such as currencies, government bonds and commodities. Moskowitz and Grinblatt (1999) reveal that the performance of common price momentum can be largely explained by industries. Capaul (1999), Scowcroft and Sefton (2005), Giannikos and Ji (2007), Chen et al. (2012) confirm that momentum in individual stocks is related to industries not only in the U.S. but in countries worldwide. On the contrary, Nijman et al. (2004) document that industries are not the predominant driver of momentum strategies in Europe. Swinkels (2002) reports that a momentum strategy that buys recent winner industry indices and sells recent loser industry indices is profitable in the U.S. and Europe. Andreu et al. (2013) corroborate these findings, focusing on exchange traded funds (ETFs). Finally, O’Neal (2000) shows that momentum also exists in sector mutual funds.
All of the above-mentioned studies point to momentum being linked to trend continuation, meaning that past winner assets continue to rise and past loser assets continue to fall. However, Novy-Marx (2012) finds that the profitability of momentum strategies in the U.S. stock market is not driven by recent past returns from the period six to two months prior to portfolio formation but rather by intermediate past returns from twelve to seven months before. Based on these findings, he concludes that momentum is not due to trend continuation but to an “echo” in stock returns. These results also apply when implementing momentum strategies based on industry portfolios, style portfolios, country indices, commodities and currencies. Goyal and Wahal (2015) and Gong et al. (2015) examine this hypothesis of momentum as an “echo” in international stock markets. However, their results do not support the findings of Novy-Marx (2012). They show that the underperformance of strategies based on recent past returns is due to a negative serial correlation between the contemporary stock return and the past stock return two months ago. Moreover, strategies based on intermediate past returns outperform due to positive serial correlation between the contemporary stock return and the past stock return twelve months ago. The authors find that strategies based on recent and intermediate past returns perform similarly when excluding these two past returns from the construction of momentum portfolios.
Our paper contributes to this debate by analyzing in detail momentum strategies based on recent and intermediate past returns in U.S. industry portfolios. Amplifying the findings of Novy-Marx (2012), we define intermediate past returns as returns during various formation periods ranging between the twelve and three months prior to portfolio formation. Our empirical results indicate that momentum strategies based on industry portfolios are profitable within the U.S. Moreover, in line with Novy-Marx (2012), strategies based on intermediate past returns from months twelve to seven exhibit higher returns than strategies based on recent past returns from months six to two prior to portfolio formation. However, when using alternative formation periods for intermediate past returns, our results do not support the hypothesis of momentum being an “echo” in returns.
See full PDF below.