Rolling Momentum Strategy: An Empirical Analysis

¹ Berlin Institute of Finance, Innovation and Digitalization at the Berlin School of Economics and Law, Badensche Straße 50-51, 10825 Berlin, Germany
² FOM Hochschule für Oekonomie & Management gemeinnützige Gesellschaft mbH, Leimkugelstraße 6, 45141 Essen, Germany

^* Corresponding author: martin.uzik@bifid.org

Abstract

Research background: The focus of the momentum strategy, as a procyclical investment strategy, lies in the hypothesis that the winning shares of the past will most likely develop in the same direction in the near future. The same is assumed for the performance of the loser shares. The technical trading rules of relative strength according to Levy provide the basis for this approach (Levy, 1967). The momentum strategy can thus offer investors an opportunity to outperform the market. The creation of portfolios under the momentum strategy follows simple rules: On the basis of past prices, equities are selected within a formation period according to return criteria. The stocks with the highest and lowest returns on equities in the formation period are combined into winning and losing portfolios, each with the same number. The final step is the acquisition of the winning portfolio, which is held over the specified investment period, with the loser portfolio being sold short at the same time. The empirical analysis presented in this paper focuses on the success of the momentum strategy for the STOXX Europe 600 market over a formation and investment period of six months.

Purpose of the article: The objective of this paper is to empirically test the above statements and assumptions. Portfolios are built up on a rolling basis over a period of six months and then observed with respect to their performance over a period from 1995 to 2000. The achieved returns are compared with a buy-and-hold strategy and empirically tested for return differences. Especially the years 2001, 2008, and 2020 as the crisis years of the dot-com bubble, the financial crisis, and the COVID-19 pandemic are focused on and discussed.

Methods: The data of the period are examined for performance development in a database in the form of winner and loser portfolios. The returns are calculated as AR to a reference portfolio DAX. The returns are statistically tested for significant differences to a zero return using a t test.

Findings & Value added: The results show the performance of the momentum strategy in the period from 1995 to 2000 for the stocks of the STOXX Europe 600. The strong fluctuations in the crisis years are notable.

With few exceptions, the reference returns could only provide statistically non-significant results.