Momentum investing explained

Momentum investing is an approach that uses past price movement to rank or classify assets. A relative momentum rule compares assets with one another and favours those with stronger measured performance. An absolute momentum rule compares an asset with its own past performance or a reference level. Both rely on historical observations; neither guarantees that recent strength will continue.

The label covers a family of methods. A complete rule needs to state the lookback period, ranking method, rebalance frequency, asset universe, treatment of ties and what happens to assets with weak readings. A different choice can produce a different portfolio and a different historical path. The name alone is not a full specification.

Momentum can overlap with trend following. Trend following often describes responding to the direction of one asset’s own price path. Momentum investing often describes comparing recent performance across several assets or using a past-return score. The terms are sometimes used more broadly, so a reader should inspect the actual rule rather than infer its mechanics from the label.

A worked example

Imagine three fictional funds with scores calculated from a made-up lookback: Fund A scores 12, Fund B scores 7 and Fund C scores minus 2. A relative rule ranks A first, B second and C third. An absolute filter might additionally require a score above zero, leaving C out. These round scores are invented and do not represent actual returns or a recommended allocation.

This simple ranking hides design choices. Are scores price changes or risk-adjusted returns? Are distributions included? Are currencies converted? How often are ranks updated? Does the strategy hold the top one, top several, or all qualifying funds? Each answer changes what the test represents. Define them before comparing results.

Why momentum can reverse

Recent winners can stop leading. A sudden change in market conditions can move leadership quickly, and a portfolio that rebalances periodically may respond only after the change. If a rule trades frequently, turnover can also make costs important. A past ranking is a record of the selected window, not a forecast about the next one.

Momentum rules can also be sensitive to the period used to measure performance. A short window may respond quickly but react to noise. A longer window may change more slowly and retain an asset after its trend has weakened. Those are trade-offs to measure, not universal settings to assume. Searching for the best lookback after seeing results can overfit the sample.

How to test a momentum idea

Write the eligible universe as it existed at each date where possible. A list assembled from today’s surviving funds or companies can leave out delisted names. StrategyLens tests on delisted prices where it can, while data coverage remains incomplete. Missing historical constituents should be disclosed because selection and survivorship can shape a result.

Separate the data used to choose the rule from the data used to evaluate it. A held-out period can show whether choices made on earlier observations carry forward to a later segment that was not used for tuning. The rule must remain fixed during that evaluation. Repeatedly changing it after looking at the holdout turns that period into more training data.

Compare the strategy with a clear baseline over the same dates, and show the path as well as a summary. A single annualised return hides the size and length of declines. Maximum drawdown helps describe the worst peak-to-trough fall, though it does not say how long recovery took. Trade count and turnover help readers understand how active the rule was.

Costs and the StrategyLens method

StrategyLens published simulations use the zero-cost basis set out in ADR 0033. The runs charge no commission or slippage; the buy-and-hold comparison is charged the same. Real trading costs money, so actual results can be lower, especially for a rule that rebalances often. The stated cost basis is important context and should accompany performance figures.

No backtest establishes that momentum works in a future market. Historical data can help identify how a defined rule behaved in a defined sample, and an out-of-sample check can reduce one form of overfitting. It cannot remove all uncertainty, compensate for missing data, or guarantee that the next period resembles the last.

When reading a momentum result, ask what was ranked, when the ranking was known, how the holdings changed, what the comparison was, and whether the evaluation period was kept separate. Ask how failed rules and costs were handled. These questions make the result easier to interpret without turning a historical pattern into a prediction.

Momentum is therefore best understood as a measurement and portfolio rule built from past price movement. It can be studied, compared and criticised like any other explicit method. The useful evidence is the full method and its limits, not the appealing simplicity of a list of past winners.

This is research, not financial advice. Past results, real or simulated, do not tell you what will happen next.

Rebalancing also means the backtest needs a realistic calendar. A score calculated with a month-end close cannot be acted on before that close is known. A test that trades at the same closing price used to calculate the signal can grant the rule information it did not have when placing an order. Clear signal and execution timing is essential for any price-based ranking method.

A ranking also depends on the available history for each asset. If one fund launched later than another, the strategy may have fewer observations for it. State how missing history affects eligibility so the comparison does not silently favour assets with longer records.

What is momentum investing?

It is an approach that ranks assets by past price movement and favours those with stronger measured momentum, according to a defined rule.

What is the difference between relative and absolute momentum?

Relative momentum compares assets with one another. Absolute momentum compares an asset's own recent performance with a reference such as zero or a trend threshold.

Can momentum reverse suddenly?

Yes. A rapid change in leadership or a sharp reversal can leave a momentum rule holding assets whose recent strength has ended. Past rankings do not predict future ones.