Cross-sectional momentum buys the assets that have outperformed their peers and avoids those that have lagged, while time-series momentum buys an asset when its own past return is positive and sells or avoids it when negative. Both use past returns, but they ask different questions.
The two ideas
| Cross-sectional | Time-series | |
|---|---|---|
| Compares | An asset with other assets | An asset with its own past |
| Rule | Hold the top-ranked, avoid the bottom | Hold if own past return is above zero (or the risk-free rate) |
| Market exposure | Can be low if long-short | Varies with how many assets are in an uptrend |
| Key research | Jegadeesh and Titman (1993) | Moskowitz, Ooi and Pedersen (2012) |
An illustrative example
Take four stocks with 12-month returns of +30%, +12%, +5% and -8%.
- Cross-sectional: rank them and hold the top one or two, whatever the market is doing. Even in a falling market, the least-bad stocks are held.
- Time-series: hold the three with positive returns and skip the fourth. In a market where all four are down, the rule holds nothing and moves to cash.
The numbers are made up.
Why the difference matters
- Cross-sectional rules are usually applied to stocks and aim to pick relative winners, as in momentum investing.
- Time-series rules are the idea behind trend following and can reduce losses in long bear markets.
- Both can suffer sharp reversals when trends turn, known as momentum crashes.
- They can be combined, for example by ranking stocks and holding them only when the market trend is positive.
Frequently asked questions
What is cross-sectional momentum?
It is a strategy that ranks assets by past return and holds the strongest relative to the others, regardless of whether the market is rising or falling.
What is time-series momentum?
It is a strategy that goes long an asset when its own past return is positive and avoids or shorts it when negative.
Which is better, cross-sectional or time-series momentum?
Neither is better in every case. They behave differently across market conditions, and many strategies combine them.
What is trend following?
It is the practical application of time-series momentum: following the direction of an asset's recent price trend.