Similar to some better-known factors, like size and value, time-series momentum historically has demonstrated above-average excess returns.
Time-series momentum, also called trend-following or trend-momentum, is measured by a portfolio of long assets that have recent positive returns and short assets that have recent negative returns. Compare this to the traditional (cross-sectional) momentum factor, which considers assets’ recent performance only relative to other assets.
In our latest book, “Your Complete Guide to Factor-Based Investing,” my co-author Andrew Berkin and I present the evidence showing that time-series momentum has provided a premium that has been persistent across long time periods and economic regimes, pervasive around the globe and robust to various definitions.
The academic evidence suggests including a strategy that targets time-series momentum in a portfolio improves the portfolio’s risk-adjusted returns. Strategies that attempt to capture the return premium offered by times-series momentum are often called “managed futures,” because they take long and short positions in assets via futures markets—ideally in a multitude of futures markets around the globe. (For those interested, see my in-depth review of the academic research of the diversification benefits of time-series momentum.)
In general, an asset that has low correlation with broad stocks and bonds provides good diversification benefits. Low or near-zero correlation between two assets means there is no relationship in their performance: Above-average, or below-average, performance by Asset A does not tell us anything about Asset B’s expected performance relative to its average.
Adding a low-correlation asset to a portfolio will—depending on the specific return and volatility properties of that asset—improve its risk-adjusted return by improving the portfolio’s return, reducing the portfolio’s volatility, or both.
With this in mind, I will review the historical evidence on the correlation of time-series momentum in U.S. stocks and how it performs when equity returns are negative. In the process, I’ll examine the question of whether the correlation turns negative when diversification benefits are most needed.
The monthly correlation of time-series momentum in U.S. equities to the U.S. market beta factor, the Russell 3000 Index and the MSCI World Index over the period January 1985 through February 2017 was zero or virtually zero in all three cases. These results show there are diversification benefits to including an allocation to times-series momentum.
However, importantly, when we look only at periods in which U.S. equities produced negative returns, we see the correlations of times-series momentum turn negative. Specifically, its correlation with U.S. market beta, the Russell 3000 and the MSCI World Index were -0.11, -0.12 and -0.15, respectively. In other words, time-series momentum tended to produce above-average returns just when the diversification benefits were needed most.