Authors: Golam Sarwar; Cesario Mateus; Natasa Todorovic
Addresses: Business School, Old Royal Naval College, University of Greenwich, 30 Park Row, SE10 9LS, London, UK ' Business School, Old Royal Naval College, University of Greenwich, 30 Park Row, SE10 9LS, London, UK ' Cass Business School, City, University of London, 106 Bunhill Row, EC1Y 8TZ, London, UK
Abstract: In this study we estimate the survival time of momentum in six UK style portfolio returns from October 1980 to June 2014. We utilise the Kaplan-Meier estimator, a non-parametric method that measures the probability that momentum will persist beyond the present month. This probability enables us to compute the average momentum survival time for each of the six style portfolios. Discrepancies between these empirical mean survival times and those implied by theoretical models [Random Walk and ARMA (1, 1)] show that there is scope for profiting from momentum trading. We illustrate this by forming long-only, short-only and long-short trading strategies that exploit positive and negative momentum and their average survival time. These trading strategies yield considerably higher Sharpe ratios than the comparative buy-and-hold strategies at a feasible level of transaction costs. This result is most pronounced for the long/short strategies. Our findings remain robust during the 2007/2008 financial crisis and the aftermath, suggesting that Kaplan-Meier estimator is a powerful tool for designing a profitable momentum strategy.
Keywords: momentum survival; style portfolios; Kaplan-Meier estimator; trading strategies; UK.
International Journal of Banking, Accounting and Finance, 2018 Vol.9 No.2, pp.192 - 224
Received: 25 Nov 2016
Accepted: 28 Aug 2017
Published online: 08 Mar 2018 *