Momentum indicator trading: how to test ROC, RSI and MACD on futures

Contents

1. How do you trade with a momentum indicator?
2. What do momentum indicators actually measure?
3. ROC, momentum and RSI: what is the difference?
4. Is MACD a trend or momentum indicator?
5. Which entry and exit rules should you test for momentum?
6. What does a momentum rule look like in a futures backtest?
7. How do you backtest momentum across markets?
8. Frequently asked questions
9. References

How do you trade with a momentum indicator?

Momentum indicator trading means buying when the speed of price change turns up and selling when it turns down. Which trigger to use (a zero cross, a 30/70 level, a signal-line cross) cannot be decided from the chart. Write it as a fixed rule, add commission and slippage, and backtest it on more than one futures market.

A momentum indicator is a line calculated from recent closes that shows how quickly price has moved over a set number of bars. Some traders will disagree with this: the entry trigger you pick matters less than the exit you pair with it and the cost you pay per trade. A momentum entry on its own decides very little. The way to find out what it decides on your market is a backtest with costs, which is the same route the guide to technical analysis tools and indicators you can test recommends for any indicator.

The working order is always the same:

1. Choose the indicator and its lookback. The lookback is the number of bars the indicator reads, for example 10 bars for momentum or 14 bars for RSI.
2. Write the entry trigger as a yes-or-no condition. "RSI crosses above 50" is testable. "RSI looks strong" is not.
3. Write the exit before you look at any result. An opposite signal, a fixed stop, a time stop or a trailing stop; one of these, written down.
4. Set commission and slippage for the contract you trade.
5. Run the rule unchanged on several markets and compare the average trade after costs with the cost of one round trip.

What do momentum indicators actually measure?

A momentum indicator measures how fast price is changing over a lookback window, not which way the longer trend points. Think of a car: the trend is the direction on the compass, momentum is the speedometer. A car can still be driving north while the speedometer falls, and it slows down before it stops and turns. That is why a momentum line usually peaks before price peaks and flattens while price is still rising.

This has two practical consequences. First, a momentum turn is an early warning, not a reversal signal: the car has slowed, but it has not turned. Second, a strong trend can keep the indicator pinned near its top or bottom for weeks, because price keeps moving at a steady fast speed. An indicator that stays overbought for a month is reporting a fast, steady rise, not a mistake; the article on what it means when KD or RSI stays stuck overbought shows how to test that situation instead of fading it.

There is research behind the idea that past speed carries some information. Moskowitz, Ooi and Pedersen (2012) studied a broad set of futures markets and found that the sign of the past twelve-month return was related to the sign of the following month's return in many of them. That is a statement about averages over many markets and decades, measured with specific rules and costs. It tells you the idea is worth testing; it does not tell you what a 10-bar momentum cross does on your contract this year. When recent price moves are close to random, any momentum reading is mostly noise, and the article on measuring market randomness with an entropy indicator explains one way to check for that before you trust a trigger.

ROC, momentum and RSI: what is the difference?

Momentum is the price difference over N bars, ROC is that difference as a percentage, and RSI rescales average gains against average losses onto a fixed 0-to-100 scale, so each one needs a different kind of trigger line. The momentum indicator is the simplest: today's close minus the close N bars ago, in points. The rate of change (ROC) divides that difference by the older close and multiplies by 100, so the reading is a percent. The RSI compares the average size of up closes with the average size of down closes over the lookback and squeezes the result between 0 and 100, with 50 as the neutral point; Wilder (1978) introduced it with a 14-bar lookback, which is still the common default.

The one formula worth keeping in your head is ROC:

ROC = (close now − close N bars ago) ÷ close N bars ago × 100

Here close now is the latest closing price, close N bars ago is the closing price N bars earlier, and the result is a percentage. Dividing by the older close is the whole point: it makes a 1 percent move on a contract trading at 20,000 points equal to a 1 percent move on a contract trading at 5,000 points.

Property	Momentum	ROC	RSI
What it computes	Price minus price N bars ago	That difference as percent of the older price	Average gain against average gain plus average loss, scaled 0 to 100
Scale	Points, no upper or lower limit	Percent, no upper or lower limit	0 to 100
Neutral line	0	0	50
Typical trigger	Zero cross	Zero cross or a percent threshold	30 and 70 levels, or the 50 line
Comparable across markets	No	Yes	Yes

The scale also changes how a threshold behaves over time. A fixed momentum threshold in points becomes easier to reach as a contract's price level rises, so a rule tuned on data from a lower price level fires more often later for no reason connected to the market. Percent and 0-to-100 scales do not drift in that way.

Is MACD a trend or momentum indicator?

MACD is both, and which one you get depends on the part of it your rule reads. The MACD line is the gap between a fast exponential moving average and a slow one (12 and 26 bars by default). A gap between two averages moves slowly and stays positive for the whole length of an up-move, so the MACD line behaves like a trend-following tool. The histogram is the MACD line minus its signal line (a 9-bar average of the MACD line), which means it shows whether the gap is widening or narrowing. The change in a gap is a speed reading, so the histogram behaves like a momentum indicator.

Part of MACD	What it reads	Behaves like	Typical rule
MACD line vs zero	Fast average above or below slow average	Trend filter	Only trade long while above zero
MACD line vs signal line	Gap turning relative to its own recent average	Slow momentum	Enter on the cross
Histogram (MACD minus signal)	Whether the gap is widening or narrowing	Fast momentum	Enter when the histogram turns, exit when it turns back

This matters when you write a rule, because a histogram turn can fire many bars before a MACD-line zero cross, and the two fire in different parts of the same move. The slowness of the two exponential averages is the price you pay for a smooth line; the article on the DEMA indicator and how it cuts exponential moving average lag shows what changes when a faster average replaces the standard ones, and why a faster line is not automatically a better one.

Which entry and exit rules should you test for momentum?

Four rule families cover almost all momentum trading, and each one must be paired with an exit written in advance, because the entry alone decides little about the result. The families are zero-line crosses, threshold crosses, signal-line crosses and divergence. They are not four ways of saying the same thing: on one move they fire at different bars and therefore trade different parts of that move.

1. Threshold cross. The indicator comes back from an extreme, for example RSI crossing up through 30. Fires earliest, often while price is still falling. Usually paired with a fixed stop below the recent low and a target at the neutral line, because a trade entered against the move needs a hard limit.
2. Zero-line or neutral-line cross. Momentum or ROC crosses zero, or RSI crosses 50. Fires after the turn is under way. Usually paired with the opposite cross as the exit, which makes it a stop-and-reverse rule.
3. Signal-line cross. The indicator crosses its own moving average, as in the MACD line crossing the signal line. Fires a few bars after the neutral-line cross. Usually paired with the opposite signal-line cross or a trailing stop.
4. Divergence. Price makes a higher high while the indicator makes a lower high (bearish), or the mirror image (bullish). Divergence is a disagreement between price direction and price speed. Fires latest in the move and needs the most careful definition: how many bars apart the two highs may be, and how much lower the second indicator high must be. Usually paired with a stop above the price high and a time stop.

The exit is where most of the result is decided. A zero-cross entry with a stop-and-reverse exit holds through every pullback; the same entry with a tight fixed stop gets taken out by the first pullback and re-enters later at a worse price. Write the exit as part of the rule, not as a setting you add when the first backtest looks weak. Each entry family and each exit type can be written once as a rule family with a range of lookbacks and thresholds, and in AlphaPrime (Valynox) that is what strategy generation takes as its input, so that every candidate the search evaluates already carries an exit and a cost assumption rather than an entry alone.

What does a momentum rule look like in a futures backtest?

Our own run on NQ daily bars shows what a momentum-style oscillator rule produces after commission and slippage, and why one kept candidate must be read as a flattered example. The run used an OSMA rule as a stand-in for the subject of this article: OSMA is the MACD histogram under another name, the MACD line minus its signal line, so it is a momentum reading in the sense of the previous sections, but it is not ROC, plain momentum or RSI. The run tested NQ (E-mini Nasdaq-100 futures) on daily bars from 2019-01-02 to 2025-06-30, long and short, one contract, with the OSMA rule as the only entry and exit, no stop, no target and no filters, charging $20 commission and $10 slippage per trade.

The important sentence about this table is how the candidate was chosen. The search tried 1,024 candidates (the count recorded in the run) and kept this one for its return against its drawdown. Picking the best of many after the fact flatters the kept one: part of its result is the rule and part is the luck of which setting happened to fit these six and a half years. Nothing here says that OSMA, or any momentum rule, earned this money, and nothing here says a reader can expect it. The article on why the best backtest in a search is usually inflated explains how large that selection bonus tends to be.

What the table can teach, read as an example, is the shape of a momentum-style rule on daily bars. Time in market of 22.69% means the rule was flat about three quarters of the time, which is typical when a histogram turn is both the entry and the exit: it holds the fast part of a move and steps aside in between. The 213 trades over the period give an average trade of $628.8 after the $30 per-trade cost, so the cost was a small share of each trade here; a rule on 5-minute bars with the same cost would have a far thinner margin. The win rate of 54.93% with a profit factor of 1.418 says the result came from winners being somewhat larger than losers, not from a high hit rate. The max drawdown of $46,600 against one contract is the number to set your account size by, not the net profit.

How do you backtest momentum across markets?

Run the same rule with the same settings on several unrelated futures markets, compare the average trade after costs with the round-trip cost on each, and keep only a rule whose neighbouring settings behave alike on most of them. The round-trip cost is commission plus slippage for one entry and one exit together. A rule that only works on one market with one setting is more likely a description of that market's history than a repeatable effect, and Park and Irwin (2007), reviewing decades of studies of technical rules, found that reported results vary a great deal with market and period and that many studies did not account for how many rules had been tried before the reported one.

1. Write the rule and the exit in the percent or 0-to-100 scale, so one setting means the same thing on every market.
2. Fix the lookback and the trigger and write them down. They do not change between markets.
3. Set commission and slippage per market. Each contract has its own tick value, so the same one-tick slippage is a different dollar amount on each.
4. Run unchanged on four unrelated markets, for example an equity index, a bond, an energy and a metal. Unrelated matters: two equity indices move together, so passing on both is close to passing on one.
5. Compare average trade after costs with round-trip cost on each market. Decide in advance how many of the four must be positive.
6. Look at the neighbouring settings too. Run the lookback a few bars shorter and longer and the threshold a few units higher and lower. A rule that passes at 14 and fails at 12 and 16 passed by luck; the article on telling a parameter plateau from a lucky spike covers how to read that surface.

The next thing to do is pick one trigger from the four families, pair it with one exit, and run it with your real commission and slippage on four unrelated futures markets you have daily data for. Write down, before the run, how many markets must show an average trade after costs above twice the round-trip cost. Then compare that number with what comes back.

Frequently asked questions

What is the difference between a momentum indicator and a trend indicator?

A trend indicator, such as a moving average, tells you which way price has been going. A momentum indicator tells you how fast it has been going over a lookback window. Speed changes before direction does, so a momentum line usually turns before a moving average does, and it can stay near its extreme for a long time while a steady trend continues.

Which momentum indicator is best for futures?

No single one tests best across markets, because ROC, plain momentum and RSI measure the same idea on different scales. For a rule you want to run on several contracts, ROC or RSI is more practical than plain momentum, since a percent or 0-to-100 reading means the same thing at any price level. The choice of exit and the cost per trade usually change the result more than the choice between these three.

What is a good setting for a momentum indicator?

Common defaults are 10 or 12 bars for momentum and ROC and 14 bars for RSI, and they are reasonable starting points because they are widely used, not because they are optimal. Test the setting a few bars shorter and longer. If the result changes sharply between neighbouring settings, the setting that looked good was a lucky spike rather than a plateau.

Can you trade with a momentum indicator alone?

You can write a complete rule from one momentum indicator, for example entering on an RSI cross of 50 and exiting on the opposite cross. Whether it holds up after commission and slippage is a question for a backtest on more than one market, not for the chart. Most momentum rules that pass such a test do so with a clearly defined exit and a cost per trade that is small compared with the average trade.

References

• Wilder, J. W. (1978). New Concepts in Technical Trading Systems. Trend Research.
• Moskowitz, T. J., Ooi, Y. H., & Pedersen, L. H. (2012). Time Series Momentum. Journal of Financial Economics, 104(2), 228–250.
• Park, C.-H., & Irwin, S. H. (2007). What Do We Know About the Profitability of Technical Analysis? Journal of Economic Surveys, 21(4), 786–826.