Learning From AQR's "Trading on Trends"

The paper listed various time-series momentum signals which basically go long when prices move up and short when prices move down.

In this article, Absolute Momentum will be used interchangeably with trend following or time series momentum. Relative Momentum will be used interchangeably with cross-sectional momentum or relative strength.


The paper listed various time-series momentum signals which basically go long when prices move up and short when prices move down.

These include return over some time period, with 12 months being the most common. The return can be computed as the ratio of prices or of a total return index. It can also be computed as the difference of (log) prices.

Returns over several different time horizons could be used. Different time horizons can be weighted.

Returns can be smoothed to reduce random noise in the data. A common way is using Moving Average.

In addition, various Moving Average crossovers, HP filter and Kalman filter methods are listed.


Research Paper: Trading on Trends: The Equivalence of TS Momentum, Moving-Average Crossover, and other Filters

Author: Lasse H. Pedersen

Company: AQR Capital Management

No Comments

Post A Comment