MOMENTUM - THE CRASHES DRAWN, NOT SMOOTHED
Momentum backtest: test a momentum strategy or a dual momentum rule
Type the momentum rule the way you would say it out loud, such as "hold QQQ while its 12-month return is positive, otherwise sit in cash, checked monthly", and the bench tests it on 20+ years of split- and dividend-adjusted daily history with every assumption printed, then compares it against simply buying and holding.
- 1 THESIS
- 2 EVIDENCE
- 3 BACKTEST
- 4 RISK
- 5 VERDICT
02 EVIDENCE · FUNDAMENTALS
04 RISK · IN PLAIN ENGLISH
Past performance does not guarantee future results. Educational analysis only, not financial advice.
In short
A momentum backtest applies a trend or relative-strength rule to historical prices so you can see how the rule actually behaved: when it got you into the move, when it whipsawed you, how deep the drawdown still went, and whether the turnover was worth it against buy-and-hold. In Agenttrading you describe the rule as a sentence, such as "buy NVDA when the 12-month return is positive and exit when it turns negative, rebalanced monthly" or "hold SPY while it trades above its 200-day moving average and move to cash below it", confirm the restated rule, and the test runs on 20+ years of split- and dividend-adjusted daily history with the assumptions strip printed on every result: lookback window, rebalance frequency, whether the most recent month is skipped, and the 0.1% cost per trade assumed by default. The academic record is worth knowing before you test. Moskowitz, Ooi and Pedersen (2012) documented time-series momentum across 58 liquid futures markets from 1985 to 2009 using a 12-month lookback and a one-month holding period, and the standard formation window in the cross-sectional literature is the prior 12 months excluding the most recent month, because that last month tends to reverse. Momentum also crashes: from March to May 2009 the past-loser decile rose about 163% while past winners gained roughly 8%, which is the single most useful thing a long backtest will show you that a short one hides. What this is not: a multi-asset factor-rotation optimizer. If you want to rank a whole universe of ETFs against each other every month and reallocate a portfolio automatically, Portfolio Visualizer or Allocate Smartly are built for that job. Here the point is testing one rule you can state in a sentence, on one ticker or fund, with the turnover and the bad stretches on screen instead of averaged away. Momentum backtesting is included on the Analyst plan at $49 per month. No trades are executed, nothing is recommended, and the output is educational analysis only: past performance does not guarantee future results.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
WHAT YOU GET - MOMENTUM BACKTEST
Momentum backtest, run on the bench
The whipsaws stay on the chart
Trend rules pay for their good years with choppy ones. Every false signal, every month you sat in cash while the market rallied, and every round trip that cost you a fee is drawn, not averaged into a single tidy number.
Turnover counted, not ignored
Momentum is a high-turnover idea, and costs are what usually kill it in practice. Trades are charged at 0.1% each by default and the trade count appears on the result, so you can see whether the edge survives the friction.
Lookback and skip-month made explicit
A 12-month lookback and a 6-month lookback are different strategies, and skipping the most recent month changes the answer again. Both choices are printed on every run rather than buried, so two tests are actually comparable.
Benchmarked against buy-and-hold
The comparison that settles a momentum rule is the same ticker held through the whole window. HELD UP, MIXED, and UNDERPERFORMED get stamped honestly, because a momentum test that cannot disappoint you is not evidence.
HOW IT WORKS - 4 STEPS
From a sentence to a stamped verdict
Describe the momentum rule
One sentence with the parameters that matter: the ticker or fund, the lookback window, what happens when the signal turns negative, and how often you check.
Confirm the restated rule
Lookback length, rebalance schedule, skip-month handling, and where the money sits when you are out of the market are shown as an explicit card before anything runs. Adjust until it matches what you meant.
Run it across 20+ years
The rule tests on split- and dividend-adjusted daily history with costs charged per trade, plotted against buy-and-hold on the same ticker, with the worst drawdown window shaded.
Read the turnover and the verdict
How many trades it took, where the rule whipsawed, how it behaved in 2008 and in the 2009 reversal, and whether the sample is large enough to mean anything. In plain English, never as a recommendation.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
On the same bench
A momentum rule is one family among many in the trading strategy builder, and it sits alongside the general engine described on backtesting software. The adjusted 20-year record behind each run is documented on historical stock data, and the drawdown a trend rule still hands you is explained on investment risk analysis. Momentum rules applied to funds rather than single names belong on ETF backtesting, and the opposite premise is tested on mean reversion trading. The most common momentum entry of all is covered step by step on the moving average crossover strategy. For the full method, read how to backtest a momentum strategy; plans start at $19 per month.
QUESTIONS - ASKED AND ANSWERED
Momentum backtest: the common questions
How do you backtest a momentum strategy?
Define the ticker, the lookback window used to measure momentum, the rebalance frequency, and what you hold when the signal is negative, then run it on long adjusted history with a realistic cost per trade and compare it against buying and holding the same ticker. In Agenttrading you type that rule as a sentence, confirm the restated rule, and read the result with the lookback, rebalance schedule, and cost assumption printed on it.
Does momentum investing work?
The premium is one of the most heavily documented in finance, and it is also one of the least comfortable to hold. Momentum has been found across asset classes and across decades, including time-series momentum in 58 futures markets from 1985 to 2009, but it crashes hard at sharp market turns: past losers gained roughly 163% from March to May 2009 while past winners gained about 8%. It works on average and fails at the worst possible moments, which is why the drawdown matters more than the average return.
Does dual momentum work?
Dual momentum, popularized by Gary Antonacci, combines relative momentum (pick the strongest of two or more assets) with absolute momentum (only stay invested if that asset is beating cash or Treasury bills). The absolute filter is what historically cut the deep equity drawdowns, and that is the part most worth testing on your own window. The honest caveat is that dual momentum results are sensitive to the exact lookback and the exact asset pair, so a rule that looks excellent on one 20-year window can look ordinary on another. Test the variants rather than trusting one published table.
What is the best momentum indicator?
There is no single best one, and the answer depends on how often you are willing to trade. Plain price return over a lookback window (typically 3 to 12 months) is the version the academic literature uses and the hardest to overfit. A 200-day moving average filter is the simplest trend switch and produces fewer signals. RSI and MACD react faster and whipsaw more. Because each choice implies a different turnover, the useful move is to test two or three on the same ticker and window and read the spread rather than picking on reputation.
What lookback period should a momentum backtest use?
The prior 12 months is the most studied formation window, usually excluding the most recent month because short-horizon returns tend to reverse. Shorter windows of 3 to 6 months react faster and trade more; longer windows are steadier and slower to exit. Since the lookback is the single parameter most likely to be curve-fitted, test the neighboring values too. If the result only works at exactly 11 months, you found an artifact rather than a strategy.
Do momentum ETFs work as well as a momentum backtest suggests?
Usually not exactly, and the gap is instructive. A momentum ETF such as a large-cap momentum factor fund rebalances on a fixed published schedule, holds a wide basket, and charges an expense ratio, so it delivers a diluted and delayed version of the raw factor. Your own backtest of a concentrated rule will typically show larger swings in both directions. Comparing your rule against the fund is a fair reality check on whether the extra effort and turnover buy anything.
Is momentum trading the same as trend following?
They overlap but they are not identical. Cross-sectional momentum ranks assets against each other and buys the relative winners, so it is always invested in something. Time-series momentum, which is what most people mean by trend following, compares an asset to its own past and goes to cash when the trend turns negative. The second one is the version that reduces drawdowns, and it is also the version that leaves you sitting out sharp rebounds. Both can be typed as a sentence and tested here.
Is a momentum backtest accurate?
It shows what a rule would have done in history under stated assumptions, not what it will do. Momentum results are especially sensitive to turnover and costs, so a test that assumes free trading can turn a losing rule into a winner on paper. Agenttrading charges a default 0.1% per trade, prints the assumptions strip, and flags thin samples where too few signals fired to mean anything. Past performance does not guarantee future results.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.
Your next idea deserves a verdict, not a hunch.
Bring a thesis or a ticker. Agenttrading restates the rule, shows the evidence, runs 20+ years of history, and stamps an honest verdict. You decide.
Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.