A guide to survival of momentum in UK style portfolios

This study contributes to the equity momentum literature by measuring the survival time of momentum in six UK style portfolios returns in the period October 1980 –June 2014.

In the context of the study, momentum is defined as a technical trading rule according to which a buy (sell) signal is triggered if a positive (negative) return is realised over two consecutive periods. Any returns of the same sign following immediately are then added to the duration of this momentum signal and a (positive or negative) trend is formed. We consider the trend to be broken if the return changes to the opposite sign.

While a significant body of research finds that trading strategies based on predictive firm characteristics (such as size, book to market ratio, leverage etc.) weaken in performance over time, some US studies find momentum investment strategy to be robust.

Momentum profits are not restricted to the US market either; studies have found evidence of momentum effect in the UK stock market, and in 12 European countries.

Previous literature agrees that momentum is present across various markets and across style portfolios, and that momentum profits diminish over time. However, there is a clear gap in the literature as to how long momentum returns persist, particularly in style portfolios, i.e. portfolios where stocks are selected based on firm characteristics that resemble a particular ‘investment style’.

This study redresses this gap by measuring the survival time of momentum in six UK style portfolios returns in the period October 1980 –June 2014.

Survival analysis studies the time of survival or failure of an event/object (in this case - the momentum) and the probability of survival or failure in a given time period. Here, the Kaplan-Meier estimator - a non-parametric method that measures the probability that momentum will persist beyond the present month - was used to estimate the average momentum survival time for each of the six style portfolios.

Consistent with previous findings, the study finds that positive momentum survives longer than negative momentum. The estimated mean survival time of positive momentum is 4 months across the six portfolios, whereas the mean survival of negative momentum varies from 2 months for large cap portfolios to 3 months for small cap portfolios.

The researchers applied four simple trading strategies to exploit mean momentum survival times based on these empirical findings: long only positive momentum in each style portfolio; short only negative momentum in each style portfolio; long positive/short negative momentum in each style portfolio; and long winner/short loser portfolio –a style rotation strategy based on investing in portfolios with highest positive or lowest negative momentum.

With the exception of the short only negative momentum, all strategies perform better than their buy and hold benchmarks and a feasible level of transaction costs. This is not surprising as the theoretical survival curves are almost identical with the empirical ones of the negative momentum a cross our style portfolios, indicating more market efficiency and less scope for profitable trading. Strongest performance, in terms of incremental Sharpe ratios relative to buy-and-hold is detected in long winner/short loser strategy, contributing to the evidence that style rotation strategies are profitable in the UK.

This paper also looks at differences in survival times and performance of strategies across economic regimes. It documents that momentum trading based on survival times works well in both expansions and recessions, including even the recent global financial crisis.

The paper's findings have useful implications for both traders and portfolio managers in that

1) momentum in style portfolios in the UK is present;  

2) momentum survival time can be successfully exploited in each style portfolio separately (relevant for consistent style investors);

3) momentum survival time leads to profitable trading in style rotation strategy (relevant for hedge fund managers), and

4) momentum trading based on mean survival times is feasible when accounting for costs.

To the best of the authors' knowledge, this study is the first to address survival time of the momentum in the style portfolio returns.

The accepted version of the paper A guide to survival of momentum in UK style portfolios is available for download at the link below. The paper has been published in International Journal of Banking, Accounting and Finance.


{A guide to survival of momentum in UK style portfolios}{https://www.bayes.city.ac.uk/__data/assets/pdf_file/0003/475491/momentum-survival-UK.pdf}