Two trades close at +$50 each. In the trade list they're identical. But one entered, moved straight to +$60, and exited cleanly; the other ran to +$400 of open profit and then surrendered nearly all of it before the exit fired. Same realized P&L, opposite stories — and the standard Strategy Tester summary can't tell them apart, because it only looks at where trades ended, never at where they went.
The data to tell them apart is already in your export. TradingView's "List of trades" sheet includes Run-up and Drawdown columns for every trade — the maximum favorable excursion (MFE) and maximum adverse excursion (MAE) while the position was open. Two derived statistics turn those columns into a description of exit behavior: capture ratio and giveback.
The definitions
- Run-up (MFE): the highest open profit the trade reached at any point during its life. A trade that closed at +$50 after touching +$400 has a run-up of $400.
- Capture ratio = realized net P&L ÷ run-up. The trade above captured 50 ÷ 400 = 12.5% of its best open profit. A trade that exits at its high-water mark captures 100%.
- Giveback = run-up − realized P&L. It's the dollar amount of open profit the trade surrendered before closing: $350 in the example. Giveback can exceed the run-up itself when a trade goes from open profit to a realized loss.
Capture ratio is most informative aggregated rather than per-trade: summed realized P&L divided by summed run-up, computed per period (month, quarter) and — importantly — split by direction. Long and short trades often show very different capture profiles on the same strategy, because the exit logic interacts differently with how the instrument trends and mean-reverts in each direction. A strategy showing 70% capture on longs and 25% on shorts is telling you its exit rules behaved very differently on the two sides of the market during the sample.
What the numbers describe (and what they don't)
Resist the reflex that low capture = bad exits. Capture ratio describes a trade-off the strategy has chosen, not a defect:
- High capture (say >70%) describes exits that fire near the favorable extreme — typical of tight targets and quick profit-taking. The cost usually shows up elsewhere: capped winners and a distribution with no right tail.
- Low capture (say <40%) is characteristic of trend-following exits that deliberately hold through retracements to catch occasional large moves. Giving back open profit on most trades is the price of the few that keep running.
So a single aggregate capture number is only interesting in context. What makes the analysis genuinely useful is looking at where the giveback concentrates — which trades, which periods, which direction — and whether the pattern is consistent with the strategy's design or a surprise. A trend strategy with low capture spread evenly across trades is behaving as designed; the same aggregate number driven by a handful of catastrophic round-trips is a different historical fact. As with all backtest statistics, these are descriptions of one historical sample, with all the caveats about persistence covered in Is My Backtest Overfit? A Practical Stability Test.
Giveback alerts: flagging the round-trips
The single most informative subset of trades is the round-trip losers: trades that reached substantial open profit and still closed negative. These are the trades where the giveback wasn't partial — it went through 100% of the run-up and kept going.
"Substantial" needs a threshold, and a fixed dollar number is wrong for every strategy but one. A self-calibrating threshold works better: flag trades where run-up was at least 2× the average losing trade and the realized P&L was negative. Scaling the threshold to the strategy's own average loss means the same rule adapts to a scalping system and a swing system without hand-tuning — a trade that had open profit worth two average losses and still closed red is notable in any system. This is the rule behind DataViz Studio's giveback alert table, which lists each flagged trade with its run-up, realized P&L, and giveback amount.
Here's what the flagged subset of a sample export might look like:
| Entry | Direction | Run-up | Realized P&L | Giveback |
|---|---|---|---|---|
| Mar 14, 09:45 | Long | $620 | −$180 | $800 |
| Apr 02, 10:30 | Short | $410 | −$95 | $505 |
| May 21, 14:15 | Long | $530 | −$240 | $770 |
A table like this converts a vague feeling ("this strategy seems to give a lot back") into specific, inspectable historical events. From here you can look at the flagged trades on the chart and see what the sample actually contained: did they cluster in one volatility regime? One time of day? One direction? Clustering by entry time is common enough that it's worth cross-checking against an hourly breakdown — the method in How to Find Your Strategy's Most Profitable Trading Hours applies directly, with capture ratio per hour instead of P&L per hour.
Computing it from your export
- Export the Strategy Tester report to .xlsx and locate the Run-up and Drawdown columns in the List of trades sheet.
- Exclude open positions first — TradingView writes them into the list as synthetic exit rows dated at export time, and an open trade's run-up is still changing. The detection checklist is in The Open-Position Trap in TradingView Backtest Exports.
- Per closed trade, compute capture = net P&L ÷ run-up and giveback = run-up − net P&L.
- Aggregate: total capture ratio per month and per direction (sum of P&L ÷ sum of run-up — not the average of per-trade ratios, which lets tiny-run-up trades dominate).
- Compute the average losing trade, then flag every trade with run-up ≥ 2× that figure and negative realized P&L.
Edge cases to handle: trades with zero or near-zero run-up make the per-trade ratio undefined or explosive (another reason to prefer the aggregated form), and losing trades that never showed open profit have capture that isn't meaningful — the statistic describes what happened to open profit, so it's conventionally computed over trades that had some.
DataViz Studio computes all of this automatically from an upload — capture ratio per period split by direction, plus the giveback alert table — as part of the risk metrics suite. For a first look at a single file, the free instant report runs in the browser session with no signup and nothing stored.
Where this fits among the other metrics
Most backtest statistics — net profit, win rate, profit factor, Sharpe — are computed from realized outcomes only, as if each trade teleported from entry to exit. Capture ratio and giveback are among the few that use the path each trade took, which is why they surface behavior the summary page structurally cannot: two strategies with identical realized statistics can have completely different capture profiles. How the path-aware metrics complement the standard ones is covered in TradingView Backtest Metrics Explained.
The takeaway in one line: realized P&L tells you what a trade kept; run-up tells you what it briefly had; capture ratio is the historical relationship between the two. None of it says what your exits should be — that's a design decision the data can inform but not make. What it gives you is a precise, checkable description of how much open profit your strategy's exits surrendered in the sample you have, instead of a hunch about it.