Maximum Adverse Excursion: How Much Heat Do Winning Trades Actually Take?

Maximum Adverse Excursion — MAE — measures the worst drawdown a trade experiences before it resolves. Win rate tells you how often a trade works. MAE tells you how much pain it caused first.

A framework can have a high win rate and still be untradeable. If most of your trades dip 8 points against you before eventually closing in profit, the win rate is academic. Your account is gone before the statistics have a chance to play out. Your prop firm evaluation failed on the second trade. Your trailing drawdown rule triggered before lunch.

I needed a better question than "does it win?"

I needed to know: when it wins, how much heat does it require first? That is the question Maximum Adverse Excursion was built to answer.

What MAE actually measures

Maximum Adverse Excursion is the worst drawdown a trade experiences before it resolves. If you enter long at 5,800.00 and price dips to 5,798.50 before rallying to your target, your MAE on that trade is 1.50 points.

That is the heat. That is the moment where the trade is working against you and you are deciding whether to hold or cut.

MAE is not a theoretical metric. It is the direct, measurable answer to the question every trader asks in real time: how wrong is this going to get before it gets right?

The companion metric is Maximum Favorable Excursion — MFE. That measures the best the trade gets before it resolves. If price rallied 6 points in your favor before pulling back and hitting your 4-point target, your MFE was 6 points. MFE tells you how much opportunity was available, whether or not you captured all of it.

Together, MAE and MFE give you the full picture of what happens between entry and exit. Not just the outcome. The experience.

Why most indicators never report MAE

The answer is uncomfortable but simple.

Most indicators do not report MAE because the results would undermine the marketing. If your signal fires and price immediately moves 4 points against you before eventually reaching a 3-point target, you technically "won." The win rate column looks clean. But the trader who took that signal just sweated through a 4-point drawdown for a 3-point gain. The risk-reward was inverted the entire time.

MAE makes that visible. It strips away the flattering abstraction of win rate and shows you the raw, bar-by-bar reality of what the trade demanded from your account — and from your composure.

That is why I measured it.

How we measured it (and the bug we caught)

Methodology matters here, because MAE is easy to measure badly.

The dataset: 89,774 first-visit signal interactions across 18 years of ES 1-minute data, from January 2008 through early 2026. Stop loss range tested: 2 to 6 points. Target range: 3 to 15 points. Every combination. Every signal.

The measurement: MAE was recorded at the exact bar of trade resolution — the specific 1-minute bar where price hit either the target or the stop. That is the correct methodology. You measure the worst the trade got during the period it was actually open, not during some arbitrary window after the fact.

I mention this because we caught a bug during development where MAE was being measured across the full 480-bar observation window regardless of when the trade actually resolved. That inflated the MAE figures by approximately 9 times. A trade that resolved in 12 minutes was being measured as if it were still open for 8 hours.

That is an easy mistake to make. And it is the kind of mistake that goes undetected when you are not looking at the numbers with adversarial intent. We caught it, fixed it, and recomputed every result in the dataset.

I share this not because the bug is interesting, but because it illustrates the standard. If the measurement methodology is wrong, nothing downstream from it matters.

What the data showed: MAE on winning trades

Here is what 18 years of correctly measured MAE revealed about winning trades in the Algorithmic Suite framework.

49% of all winning trades experienced zero adverse movement.

Nearly half. Price moved immediately in the direction of the signal and never came back. No drawdown. No heat. No moment of doubt. Entry to target in a straight line.

78.1% of winning trades saw less than 1.0 point of adverse movement. On ES, one point is four ticks. On MES, that is $5.00. More than three out of four winners asked for almost nothing from your account before they worked.

94.5% of winning trades saw less than 2.0 points of adverse movement. Fewer than 1 in 20 winners ever dipped more than 2 points against you.

The median MAE on winners was 0.25 points. One tick.

Read that again. The typical winning trade in this framework dipped one tick against you before reaching its target. That is not marginal. That is a fundamentally different experience than what most entries produce.

Why this matters for prop firm traders

If you are trading a funded account or a prop firm evaluation, you already know what trailing drawdown rules do to your psychology.

Every point of adverse movement on an open trade is a point closer to your maximum loss threshold. It does not matter that the trade "would have worked" if you had held through a 5-point dip. The trailing drawdown does not care about eventual outcomes. It cares about the worst moment.

This is where MAE becomes the most important metric in your evaluation.

A framework that regularly produces 3-5 points of heat on winners is a framework that will trigger trailing drawdown rules on losing trades at a rate that makes the math impossible. Even if the winners outnumber the losers, the losers will be deep enough to end the evaluation.

A framework where 78% of winners see less than 1 point of heat is a framework designed for the constraints you actually operate under. Not because it was optimized for prop firm rules. Because the signals fire at genuine inflection points near levels where price actually turns.

That is not a coincidence. It is what happens when entry quality is the priority, not signal frequency.

The other side: Maximum Favorable Excursion

MAE tells you about the pain. MFE tells you about the opportunity.

If a framework produces tight MAE but also tiny MFE, you are looking at marginal scalps. Trades that barely go against you but also barely go in your favor. The kind of trades where one tick of slippage or one extra tick of spread erases the entire edge.

That is not what we found.

87.2% of all trades saw price move at least 2 points in the favorable direction. Not just winners. All trades. Even many of the trades that ultimately stopped out first visited favorable territory worth multiples of the average MAE.

65.3% of all trades saw at least 4 points of favorable movement. That is the available excursion. The framework is not producing marginal entries at noisy, directionless price zones. It is identifying locations where price has conviction — where the move after the signal tends to be real and measurable.

The combination of tight MAE and wide MFE is the signature of a genuine inflection point. The signal fires. Price moves. It does not wander. It does not chop sideways for 45 minutes before eventually drifting somewhere. It goes.

What tight MAE actually means about signal quality

There are two ways a framework can achieve a high win rate.

The first is to set a wide stop and a tight target. Give the trade 10 points of room to be wrong, aim for 2 points of profit. You will win most of the time. And when you lose, you will lose five times what you made on a winner. The math does not work, but the win rate looks impressive on a slide deck.

The second is to enter at the right place. Identify the price level where buying or selling pressure is genuinely concentrated. Fire the signal at the moment where the probability of immediate continuation is highest. Win not because you gave the trade room to breathe, but because the trade did not need room to breathe.

MAE is the metric that distinguishes between those two approaches.

When 49% of winners see zero heat, the framework is not winning because of generous stop placement. It is winning because the entry itself is clean. Price arrives at a level, the signal fires, and the move begins. The stop is there as a risk management boundary, not as a crutch the trade routinely leans on.

That distinction is invisible in a win rate number. It is invisible in a profit factor number. MAE is where it lives.

The numbers in context

These MAE and MFE results did not come from a favorable subset of dates or a single target/stop configuration.

They come from 89,774 signals across 4,721 trading sessions spanning every market regime from the 2008 financial crisis through 2026. Bull markets and bear markets. Low volatility and extreme volatility. Trending environments and choppy ones.

The stop loss range tested spans 2 to 6 points. The target range spans 3 to 15 points. Forty-five combinations, every signal, every session. The MAE characteristics held across all of them.

I have written previously about the scope of the research behind the Algorithmic Suite and about the real transaction costs that every backtest must account for honestly. MAE is the third pillar. You need all three — research rigor, honest cost accounting, and measured entry quality — before you can say anything credible about whether a framework works under live conditions.

The question you should be asking

If you are evaluating any trading framework, for any market, on any timeframe, ask for the MAE distribution on winners.

Not the win rate. Not the equity curve. Not a screenshot from a good Tuesday.

Ask: when this framework wins, how much heat does the average winner require? What percentage of winners go immediately in my favor? What does the worst 5% look like?

If the answer is not available, ask why.

If the answer is that winning trades routinely dip 3, 4, 5 points against you before eventually reaching target, ask yourself whether your account can survive that. Whether your prop firm evaluation can survive that. Whether you can survive that at 6:45 in the morning with real money on the line.

The Algorithmic Suite was built to produce entries where nearly half of all winners never go against you at all. Where 78% see less than a point of heat. Where the median winner dips one tick.

That is not marketing language. That is 89,774 measured trades across 18 years of data.

The Algorithmic Suite

Midnight Grid. Quantum Vision. Turning Points.

Three indicators. One framework. Built on entries where the heat is measured, not hoped away.

Start Your 7-Day Free Trial

Algorithmic is charting software for decision support on TradingView. It is not financial advice. Trading involves risk. Outcomes depend on your rules, risk management, and execution. Past performance does not guarantee future results.