Loss Taken (MAE Capture)

Loss-side mirror of Profit Captured: average share of the adverse move taken as a realized loss = mean(|loss| / |MAE|), clipped to 0–100%. ~100% = losers close near the worst point (hard stop / held to the bottom); low = adverse move recovered before exit. Read together with Avg MAE. (a.k.a. MAE Capture)

Computed from
Trades list
Scope
Single report
Range
0 – 100%
Direction
Context-dependent

Loss Taken (MAE Capture) measures how much of a losing trade's worst moment you actually realized as a loss. That worst point is called the MAE — Maximum Adverse Excursion, a fancy name for how far against you the trade ever travelled before you closed it. Loss Taken is the share of that adverse peak you ended up eating: 100% means you closed right at the bottom, lower means the trade recovered some of the way back before you exited.

Crucially, there is no universal good or bad here — the verdict flips depending on whether you trade with a stop (a rule that automatically closes a trade once it loses a set amount, so it can't get any worse). A disciplined hard-stop trader will structurally show high Loss Taken, because the stop fires near the adverse extreme by design. The same number for a discretionary trader (one who closes by hand, with no fixed rule) means you held losers all the way down. Read it as a behavioural diagnostic, not a score.

How it's calculated

Every trade traces a path while open (see MAE / MFE). MAE is the trough of that path — the deepest the trade ever went against you. Loss Taken asks how much of that trough you carried through to the exit.

Loss Taken = average ( realized loss / MAE )   over losers
MAE
peak unrealized loss while the trade was open (the worst point)
realized loss
the net loss you actually booked at close
In this product: Computed per trade, then averaged, as a percent. EquityTruth uses the money basis and includes only losing trades with a non-zero MAE; the code takes the absolute value of both the realized loss and the MAE (which is itself ≤ 0), so the ratio is positive. Each per-trade ratio is clipped to 0–100% (the close can print beyond the in-window M1 extreme). A single value, not split by axis. See the MAE / MFE article for the excursion engine.

What it tells you

The verdict on every band below depends on one thing: do you run a hard stop? With a stop, high Loss Taken is the mechanical signature of the stop firing near the adverse extreme — exactly what it's supposed to do. Without one, high Loss Taken means you sat in losers until they bottomed out. The bands are diagnostic prompts, not a good-to-bad ladder.

ValueReadingNotes
~95–100%Stops firing — or holding to the floorYour losers close at their worst point. With a hard stop, this is expected and healthy. Without one, you are riding losers all the way down. Check which it is.
70–95%Mostly the full adverse moveYou eat most of the drawdown before exiting. Fine for a stop-driven system; a flag for a discretionary one that should be cutting sooner.
40–70%Cutting losers before the worstA meaningful chunk of the adverse move was recovered before you closed. Could be smart cutting — or a stop placed wide of the typical trough.
< 40%Losers recover a lot before exitYour losers bounce back substantially before you bail. Can mean disciplined exits, or the disposition effect — holding a loser hoping it returns, then closing once it partly does.

Worked example

A trade sinks to −$200 at its worst (MAE = $200) but you exit at −$120. Its Loss Taken is 120 / 200 = 60%. Average that ratio across every losing trade and you get the figure for the whole record. Now read the 60% two ways:

  • If you ran a $200 hard stop, you'd expect Loss Taken near 100% — the stop should fire at the floor. Getting 60% means the trade bottomed and bounced before hitting your stop, so the stop was never touched.
  • If you exit by hand, 60% means your losers recover about a third of the way before you cut them. That could be smart cutting at a better point — or a sign your stop sits too wide of where price typically turns.

How to read it with your stop policy

This metric only means something once you overlay your own rules:

  • Hard-stop trader — expect high Loss Taken, and treat near-100% as normal. Your stop is designed to fire at the adverse extreme. A low reading here is the interesting one: it says price keeps bouncing before your stop, so your stop may be wider than it needs to be.
  • Fixed-risk (R) trader — if every loser is meant to cost the same "1R" (one unit of risk = your stop distance), Loss Taken is pinned near 100% almost by definition: 1R is the adverse extreme. The headline number stops being diagnostic and just confirms the stop fired — look at the spread and the low outliers instead.
  • Discretionary trader — high Loss Taken is the danger sign. It means you hold losers to the floor with nothing forcing you out. Lower is generally healthier here, because it shows you're exiting before the bottom.
  • Either way, pair it with Max Drawdown — Loss Taken tells you how much of each loser's adverse move you ate, drawdown tells you what those losses did to the account.

Pitfalls

Pitfalls & caveats
  • Ambiguous by design. A high number can mean a disciplined hard stop or no stop at all. You must know your own rules to read it — the metric cannot tell you which.
  • It moves with stop placement, not just skill. A tighter stop mechanically raises Loss Taken, a wider stop lowers it — independent of how well you trade. Don't read the raw number as a discipline score until you account for where your stops sit.
  • Survivorship in the denominator. Loss Taken only averages over trades that closed as losers. A trade that sank deep, then recovered all the way to break-even or profit, counts as a winner and drops out — so the metric never sees your best adverse-trade recoveries, which biases the figure upward. A low reading therefore understates how often your trades actually claw back. Read it alongside Profit Captured, which catches some of those recovered trades on the win side.
  • It's an average. One trade held all the way to its trough can skew the figure; one early cut can flatter it. Look at the spread, not just the mean.
  • It says nothing about loss size. Eating 60% of a huge MAE is far worse than eating 90% of a tiny one. Pair it with the dollar MAE and expectancy.
  • Winners excluded. The profit-side mirror is Profit Captured (MFE capture) — read both together.
  • Reconstructed from M1. MAE comes from minute-bar reconstruction of the intraday path; gaps or missing symbols make it unreliable.

Loss Taken vs Profit Captured

Profit Captured grades your exits on winners — did you keep the gain the trade offered? Loss Taken grades the downside — how much of the adverse move you actually ate. Together they describe your exit-and-stop discipline:

  • High Profit Captured + low Loss Taken = the ideal: keep most of your upside, cut your downside early.
  • High Profit Captured + high Loss Taken = potentially the disposition effect — snatch profits early, sit in losers to the bottom. But mind the subtlety: with a hard stop, high Loss Taken is expected, so this disposition-effect read only applies when there's no stop discipline. Check your stop policy before calling it a defect.

MAE / MFE · Profit Captured · Max Drawdown · Expectancy

Related metrics