MAE: Use Maximum Adverse Excursion to Place Stops

Most traders set stops by feel, round numbers, or how much they can stomach losing. Your own trade history already contains a better answer — you just have to measure the heat.

What MAE captures: peak heat per trade

Maximum Adverse Excursion (MAE) is the worst point a trade reached against you before it closed — the deepest unrealized loss, the maximum heat you sat through. A trade that you closed for a profit might have been down significantly at some point first; MAE records that low-water mark whether the trade ultimately won or lost.

This is different from your actual loss. Your realized result tells you where the trade ended. MAE tells you the worst place it visited. That distinction is the entire value of the metric: it exposes how much risk you were genuinely exposed to, not just how much you booked. Two trades that both closed at +1R can have wildly different MAE — one sailed straight to target, the other nearly stopped you out first. Only the second is telling you your stop is doing real work.

Building an MAE scatter of winners vs losers

The insight lives in a single chart. Plot every trade with MAE on one axis and colour it by outcome — winners one colour, losers another. A scatter of a few hundred trades reveals a pattern no summary statistic can.

What you are looking for:

Where winners cluster. Most winning trades experience only a little heat before working. They pile up at low MAE values.
Where losers live. Losers are spread across higher MAE — they are the trades that kept going against you.
The overlap zone. The MAE range where winners and losers mix is where your stop decision actually matters.

If your journaling is manual, building this chart is a weekend of spreadsheet pain and you will never redo it. Shibiki auto-journals every fill and reconstructs MAE from the price path, so the scatter stays current without data entry — and it plots winners against losers automatically as new trades close.

Finding the stop distance winners rarely violate

Now read the chart for a number. Look at where your winning trades sit and find the MAE level that only a small fraction of them ever exceeded. Call it the point where, say, the large majority of your eventual winners had already turned around.

That level is a data-driven stop candidate. The logic is direct: if almost none of your winners ever went that far against you, a trade that does go that far probably is not going to be a winner. Cutting it there costs you very little real upside — you are only stopping out trades that were statistically already lost — while it caps the losers that would otherwise bleed past.

Contrast that with a stop set wider than any winner’s MAE. Every extra tick of stop distance beyond your winners’ heat zone is pure donated risk: it never saves a winner, it only enlarges losers.

Cutting stop size without cutting your win rate

This is the counterintuitive payoff. Traders resist tightening stops because they fear getting stopped out of good trades. MAE tells you whether that fear is real.

If your winners rarely reach the current stop distance, you can pull the stop in toward their heat zone and lose almost no winning trades — the ones you stop out were headed for red anyway. The result:

Smaller average loss, because losers get cut sooner.
Roughly unchanged win rate, because winners were never using that extra room.
A better payoff ratio and higher expectancy, from the same setups.

The one caution: tighten into the data, not past it. Pull the stop inside your winners’ heat cluster and you will start knocking out trades that would have worked. MAE shows you the exact edge of that cliff so you stop at it instead of guessing.

MAE in R so it’s comparable across trades

Raw MAE in ticks or pips is useless across instruments — a 20-pip excursion on gold and on a currency pair mean nothing side by side. Normalize everything to R-multiples: express each trade’s MAE as a fraction of the risk you took on it.

An MAE of 0.4R means the trade went 40% of the way to your stop before resolving. Now every trade — different instruments, different account sizes, different position sizes — lives on one comparable scale. You can ask clean questions like “what MAE-in-R do 80% of my winners stay under?” and get an answer that holds across your whole book. This is also what makes MAE portable when you run the same strategy across several funded accounts.

Turning MAE insight into position sizing

MAE closes the loop back to sizing. Once you know the stop distance your winners respect, that distance becomes the input to your risk-per-trade math. Tighter, evidence-based stops mean you can hold your dollar risk constant while carrying a larger, more responsive position — or hold size constant and risk less per trade.

Feed the stop distance into a position size calculator to convert it into exact lots for your account and risk tolerance. Then make the number stick: Shibiki pushes hard risk limits down to the broker-side EA, so the per-trade loss you derived from MAE is enforced at the broker rather than left to willpower in the moment you most want to widen it. The whole point of measuring the heat is to stop sitting through more of it than your own winners ever needed — and enforcement is what turns that measurement into a habit the account can rely on.

If you trade on MetaTrader 5, the price-path data behind MAE comes straight from your fills, so the analysis reflects exactly what happened in the account, not a manual reconstruction.