General

What is Edge Audit?

Reviewed by Karthikeyan Anbarasan (Founder, TradeLyser) · Updated 19 Oct 2026 · Editorial policy

DefinitionAn edge audit is a scheduled check, usually monthly or quarterly, of whether each setup you trade still has positive expectancy after charges on a large enough sample. An illustrative audit might keep an opening range breakout at +0.32R over 84 trades and pause expiry-day scalps at −0.18R over 63 trades.
Formula
Expectancy (R) = (Win rate × Average win in R) − (Loss rate × Average loss in R), using net results after charges
Example
45% wins averaging 2.0R and 55% losses averaging 1.0R: (0.45 × 2.0) − (0.55 × 1.0) = 0.90 − 0.55 = +0.35R per trade
+0.4R0RTrade 20Trade 120Edge gone: review the setup
Figure 1. Illustrative rolling 20-trade expectancy for one setup, plotted from trade 20 to trade 120. It starts near +0.4R, slides steadily and dips below zero in the last stretch before trade 120. An audit that only looked at the all-time average would still show a positive number.

Running the audit

  1. Pull every closed trade for the period and group it by setup tag. Use net P&L after brokerage, STT and other charges.
  2. Check the sample. Under 30 trades a setup is "not enough data"; around 100 the numbers become dependable.
  3. For each setup, work out expectancy in R, win rate, average win and loss, profit factor and maximum drawdown.
  4. Compare with the last audit and with the most recent 20 or 30 trades, as in Figure 1, to see the direction of travel.
  5. Check rule adherence on the setup’s trades. If adherence was low, the audit is measuring execution, not the edge.
  6. Write a decision for each setup, keep, resize or pause, with the date, and review it at the next audit.

Worked example: three setups

SetupTradesExpectancyNet P&L (1R = ₹2,000)Decision
Opening range breakout84+0.32R+₹53,760Keep at current size
VWAP reclaim41+0.11R+₹9,020Keep; recheck next audit
Expiry-day scalps63−0.18R−₹22,680Pause; paper trade only

Illustrative. At ₹2,000 per R, +0.32R is ₹640 per trade, so 84 trades make ₹53,760; +0.11R is ₹220 per trade, ₹9,020 over 41 trades; −0.18R is −₹360 per trade, −₹22,680 over 63 trades. The account made +₹40,100 overall; without the scalps it would have made +₹62,780. Note the sample sizes as well: the VWAP reclaim has 41 trades, enough for a rough read but not a confident one, so it stays at current size rather than being scaled up.

Account net
+₹40,100
Best setup
+0.32R
Setup to pause
−0.18R
Minimum sample per setup
30 trades
Keep
Positive expectancy over 30 or more trades, stable or rising against the last audit, adherence high.
Watch or resize
Positive but falling, or fewer than 30 trades. Hold or reduce size and recheck at the next audit.
Pause
Negative over 30 or more trades while the rules were being followed. Move to paper trading and re-test.

Edge problem or execution problem

Edge decay

Rules followed, entries and exits as planned, and expectancy still falling. The market conditions the setup relied on may have changed.
Response: resize or pause, then re-test

Execution leak

The setup’s A-grade trades are still positive, but late entries, moved stops and oversize trades drag the total down.
Response: fix the rule break, keep the setup

A quick test is to filter the setup’s trades by execution grade. If the A-grade trades alone are still positive and the C-grade trades drag the total down, the setup is fine and the leak is execution. If even the A trades have turned negative over 30 or more trades, the edge itself has weakened. The mistakes below make either answer harder to read.

  • Auditing setups with 10 or 15 trades and treating the result as final.
  • Using gross P&L. A setup with a thin edge can turn negative after charges.
  • Mixing two setups under one tag, so a good one and a bad one average out.
  • Changing the setup’s rules halfway through the period, so the audit measures two different things.
Rule of thumb
Audit on a fixed date with the market closed, such as the 45-minute review on the first weekend of the month, never straight after a bad day. Keep a dated log of each decision, and at the next audit check whether paused setups would have recovered and kept setups held up; over a year that shows whether your thresholds act too fast or too slowly.
For F&O traders in India
Split audit periods at contract changes: NSE moved expiries from Thursday to Tuesday in September 2025, lot sizes are revised periodically, and STT on options rose to 0.15% of the sell-side premium from 1 April 2026. A setup that cleared its costs before April 2026 may not after. The free Setup Edge Tracker by Strategy Tag template lays out these columns, and TradeLyser’s strategy mapping shows performance per strategy, including win rate, expectancy and R-multiple distribution.

Reference guide

ContextValueReading
Per-setup numbers, net of charges, 30+ trades eachMeaningful auditEnough data to separate a working setup from a lucky run
One account-level number for all tradesHides the answerA strong setup can carry a losing one for months; split by setup tag
Setup paused after five losses in a rowPrematureLosing streaks are normal; judge on expectancy over a full sample

Questions

What is an edge audit in trading?

A scheduled review of each setup’s results, net of charges, to check that it still has positive expectancy. It ends with a keep, resize or pause decision for each setup.

How often should I audit my trading edge?

Monthly for active intraday or F&O traders, quarterly for slower swing traders. What matters is that each setup has enough new trades between audits to say something.

How many trades do I need for an edge audit?

At least 30 per setup for a rough read, and closer to 100 before trusting the numbers. With fewer, record the result as "not enough data" rather than acting on it.

What is a good expectancy in an edge audit?

There is no universal figure. It should be positive after charges, hold up over 30 or more trades and stay roughly stable between audits. A small positive expectancy over many trades can be worth more than a large one over a few.

What should I do if a setup has lost its edge?

First check adherence; if the rules were not followed, fix execution before judging the setup. If they were, reduce size or pause it, paper trade it, and only bring it back if a fresh sample is positive.

Related terms