Momentum Investing: The Multi-Asset Evidence

Casper Hammerich

Student thesis: Master thesis

Abstract

The tendency of investments to exhibit persistence in their relative performance was initially postulated by Jegadeesh and Titman in their 1993 Journal of Finance article. They postulated that investments, which had performed relatively well would continue to perform relatively well. Those investments that have performed relatively bad would continue to perform relatively bad. They named such a tendency momentum. According to modern nance such a tendency cannot persistently be observed and exploited as any market anomaly would be eroded by the ubiquitous market e ciency. Given that the markets are fully e cient, investors should not expect arbitrage strategies as momentum to be pro table neither in the short nor in the long run. Any excess return relative to the market would not be feasible. However, consecutive publications, following the 1993 article, have proven that momentum is more than a tendency. Long-short strategies in particularly equities and subsequently bonds, commodities and currencies have proven to be highly pro table due to underlying momentum dynamics. In con- sequence of such strategies, this thesis will illustrate that a parallel multi-asset momentum exists and that investors are able to exploit this using a cross-sectional multi-asset data sample. Adjusting the conventional momentum methodology, using a 12-year sample period and data sample of 50 indices, this thesis will illustrate that investors are able to invest in multi-asset portfolios based on the asset classes past relative performance. Shorting the past years worst performing assets to fund long po- sitions in the past years best performing assets, and holding such trading portfolio for subsequently three months, investors would earn a signi cant excess return of 9.19 percent. Additionally, this thesis will illustrate how positive market movements of equities and commodities as well as perception of low volatility tend to accelerate the multi-asset momentum. It will further illus- trate how the multi-asset momentum is driven by the worst performing asset classes and illustrate how bull markets will accelerate the multi-asset momentum and how bear markets would erode multi-asset momentum, respectively. Finally, this thesis will illustrate that capital markets should not to be perceived as fully e cient, hence investors should expect a plausible excess return to the market. Momentum should no longer be perceived as a random market anomaly, but rather as a ruling market factor.

EducationsMSc in Finance and Accounting, (Graduate Programme) Final Thesis
LanguageEnglish
Publication date2012
Number of pages99