Loading...

Time-series and cross-sectional momentum investment strategies: International evidence

Time-series and cross-sectional momentum investment strategies: International evidence

##### Abstract

Numerous studies have found that profits that can be realised from following a momentum-based investment strategy of buying recent outperforming stocks (winners) and selling recent underperforming stocks (losers) (Jegadeesh & Titman, 1993, 2001). Momentum strategies have proved to be robust across time, countries and asset classes, leading Fama (1998) to observe that momentum remains the “premier unexplained anomaly”. The existence of the momentum abnormal returns continues to challenge the market efficiency theory.
The majority of momentum studies have investigated cross-sectional momentum strategies in which stocks are selected on the basis of their relative performances over some prior period. In a recent study, Moskowitz, Ooi, and Pedersen (2012) introduce a time-series momentum strategy which provides an alternative approach to security selection where stocks are chosen on the basis of their absolute performance over some prior period.
Although previous literature has evaluated momentum strategies in numerous markets settings, by far the bulk of these studies have concentrated on equity markets. Therefore, it is somewhat surprising that we are yet to see a comprehensive study that compares the two types of momentum strategies in this arena. The main objective of this study is to evaluate and compare the performances of the two momentum strategies in the security markets in order to examine validation of the market efficiency theory across international stock markets.
The study addresses the objective in three stages: before-transaction costs performance (raw returns), after-transaction costs performance (net returns) and after-transaction costsperformance adjusted for risks (risk-adjusted net returns). A further by-product of our research is that we utilise a large number of implementation alternatives for both momentum strategies and so provides an insight into the optimal way to implement both time-series and cross-sectional momentum strategies.
Before transaction cost (raw) return:
The first thing that has been found is that the time-series and cross-sectional momentum returns reduce as we increase the cut-offs used when choosing both winning and losing stocks and so increase the number of stocks in the momentum portfolios (i.e. including 32%, 60% and all stocks in either the winner or loser portfolios). The extension of the cut-offs from 32% to 100% results in a reduction in the returns on the momentum portfolios by approximately 50% on average based on the pooled data for the 24 markets. Having established this, we then use the 32% cut-offs over the remainder of our analysis.
The study finds that both time-series and cross-sectional momentum strategies produce significant positive outcomes under numerous implementations in the majority of developed stock markets with the major exceptions being Greece, Israel, Japan, Hong Kong, Portugal, Spain and the US. The time-series momentum strategy outperforms the cross-sectional momentum strategy under optimal implementations conditions in all markets and is statistically significant in half of these markets.
After transaction costs (net) returns, and risk-adjustment net returns:
We find that the transaction costs and standard risk explain most profitability of the time-series and cross-sectional momentum strategies. In terms of the Fama-French alpha determined using after-transaction costs return, about 6% on average of the implementations evaluated produce significant positive risk-adjusted net returns. There are absolutely no implementations that yield significant positive returns in Austria, France, Germany, Greece, Hong Kong, Israel, Japan, Norway, Portugal, Singapore, Spain and the US., while less than 5%of the implementations in Australia, Canada, and Ireland. The findings support that the market efficiency hypothesis still holds across the most markets in our sample and the existence of exploitable investment opportunities is rare.
The study particularly concentrates on the optimal implementations of both the time-series and cross-sectional momentum strategies across the 24 markets. Common characteristics of these optimal implementations for the risk-adjusted net returns are that they combine a formation and a holding period of between 15 and 18 months, a buy-and-hold portfolio construction policy and the use of either a market or inversed-volatility portfolio weighting scheme.
Based on the optimal implementation approach for each market, the overall performance of the two momentum strategies is eroded from 2.09% (raw return) to 1.34% (net return) and to 0.9% (Fama-French alpha) for the time-series momentum strategy, and from 1.43% (raw return) to 0.87% (net return) and finally to 0.51% (Fama-French alpha) for the cross-sectional momentum strategy. At each of the three steps along the way, this study finds that the time-series momentum strategy continues to outperform the cross-sectional momentum strategy; however the magnitude of the superior performance is diluted with an average difference from 0.66% (raw return) to 0.47% (net return) and to 0.39% (Fama-French alpha). In addition, the superior performance of the time-series momentum strategy relative to the cross-sectional momentum strategy comes during periods when the markets have been performing poorly but that this advantage also erodes as we proceed from raw returns to net returns to risk-adjusted net returns.
One possible explanation for the superiority of the time-series momentum strategy is that it forms portfolios of slightly smaller capitalization stocks with a greater spread in past performance between the winner and loser stocks. Both of these features suggest that thetime-series momentum strategy will outperform the cross-sectional momentum strategy. On the other hand, the transaction costs and risk from the time-series momentum strategy are higher than the costs from the cross-sectional momentum strategy which is largely a consequence of time-series momentum strategy selecting smaller and growth stocks, and generating a higher turnover over a market cycle, so it is not surprising that the outperformance of the time-series momentum strategy becomes smaller after adjusting for risk on an after-transaction costs basis.

##### Type

Thesis

##### Type of thesis

##### Series

##### Citation

Gao, X. (2012). Time-series and cross-sectional momentum investment strategies: International evidence (Thesis, Doctor of Philosophy (PhD)). University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/9343

##### Date

2012

##### Publisher

University of Waikato

##### Supervisors

##### Rights

All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.