Direct Answer
Momentum computed from the part of a stock's returns not explained by common factors such as the market, size and value. Residuals come from a regression, and the score is their accumulated size, often scaled by their volatility. A high score means the stock has beaten what its factor exposures would predict, while a low score means it has lagged.
Residual Momentum: Formula, Meaning, Signals, Examples and How to Use It
How Is Residual Momentum Calculated?
Momentum computed from the part of a stock's returns not explained by common factors such as the market, size and value. Residuals come from a regression, and the score is their accumulated size, often scaled by their volatility.
Formula: Score = Sum(e_t over the formation window) / StdDev(e_t), where e_t = r_t - (alpha + beta1*F1_t + beta2*F2_t + ...)
Inputs:
- Monthly stock returns
- Factor returns such as market, size and value
- A regression window (often 36 months)
- A formation window of about 12 months
Worked example. If residuals add up to 0.09 over the window with a residual standard deviation of 0.04, the residual momentum score is 2.25. The inputs are illustrative numbers, not market data.
Reading it. A high score means the stock has beaten what its factor exposures would predict, while a low score means it has lagged. Ranking on residuals aims to reduce dependence on the market's direction.
Limits. The regression introduces estimation error and requires choices about the factor model and windows. It needs more data than raw momentum and is harder to replicate exactly.
What is Residual Momentum?
Momentum computed from the part of a stock's returns not explained by common factors such as the market, size and value. Residuals come from a regression, and the score is their accumulated size, often scaled by their volatility.
How to use Residual Momentum
A high score means the stock has beaten what its factor exposures would predict, while a low score means it has lagged. Ranking on residuals aims to reduce dependence on the market's direction. A reading alone is never a buy or sell signal; it needs context from price structure, volume, volatility and the broader market regime.
Strengths
- Stripping out market, size and value effects first means the score reflects stock-specific return rather than exposure to common factors.
- Dividing the sum of residuals by their standard deviation scales each stock by its own idiosyncratic volatility, so scores are comparable across names.
- Reproducible when the formula and data convention are fixed.
Weaknesses and limitations
- Factor betas are re-estimated over a rolling window, so a stock's residual history is revised whenever the window or factor set changes.
- It works on monthly returns and needs factor return series as well, so it cannot be computed from a stock's price history alone.
- Correlated indicators can create false confidence by repeating the same underlying information.
Editorial and risk note
This page is educational content, not individualized investment advice. It should not imply guaranteed prediction, accuracy or outperformance.
Frequently Asked Questions
Is Residual Momentum a buy or sell signal?
No. A high score means the stock has beaten what its factor exposures would predict, while a low score means it has lagged. The regression introduces estimation error and requires choices about the factor model and windows. A trade built on it still needs a trigger, an invalidation rule, a position size and realistic execution assumptions.
What is the best setting for Residual Momentum?
Residual Momentum takes these inputs: monthly stock returns; factor returns such as market, size and value; a regression window (often 36 months). There is no universal best setting for the tunable ones, so test them across instruments, regimes and out-of-sample periods and favor stable parameter regions over one historical winner.
Can Residual Momentum be used by itself?
The regression introduces estimation error and requires choices about the factor model and windows. It needs more data than raw momentum and is harder to replicate exactly. For that reason it is read next to measures of direction, participation, valuation, liquidity and risk, not alone.
Does Residual Momentum work on every timeframe?
The calculation (Score = Sum(e_t over the formation window) / StdDev(e_t), where e_t = r_t - (alpha + beta1*F1_t + beta2*F2_t + ...)) can be run on any bar size, but behavior is not portable: a reading on a five-minute chart describes a very different market window from the same reading on a daily chart. Validate the exact timeframe before relying on it.
Why does Residual Momentum give false signals?
Factor betas are re-estimated over a rolling window, so a stock's residual history is revised whenever the window or factor set changes. It works on monthly returns and needs factor return series as well, so it cannot be computed from a stock's price history alone. Lag, noise and regime changes add to this, because market participants react to new information after the reading is calculated.
How should Residual Momentum be backtested?
Rebuild Residual Momentum from monthly stock returns using point-in-time data, write the rules before testing, include delisted securities where relevant, model realistic fills and costs, reserve a validation sample and compare with a simpler baseline.