Audit My TradesThis is a sample built from made-up trades. Audit your own file →
You are ahead, but the evidence of an edge is weak
Over 132 trades (2025-03-04 to 2026-07-03) you kept ₹77,701 after ₹58,944 in estimated charges. The biggest leak below cost you ₹58,944.
The gain is small against the swings from trade to trade. A run like this turns up by chance often enough that it proves little yet.
The test behind it: the gains are too uneven to separate from luck (steadiness -1.68 per block; 1.5 is the least that counts as moderate evidence).
Read it top to bottom: the verdict, then the fixes worth the most money, then where the money went. Everything after that is the detail behind those.
Weakest: Edge. How every point was lost →
If you had no edge at all, a per-block gain this large would still turn up by chance about 1 in 7 times. Take out what NIFTY did while each trade was open and the same trades average -0.234 % per block -- most of the gain was the market, not the strategy.
The whole book as one: 132 trades on 12 symbols (132 long, 0 short), Cash delivery. Read from a trade list.
Charges are estimated: each trade's own turnover at today's statutory rates (2026-10 (STT per Budget 2026, eff 1-Apr-2026)) plus the brokerage you set and a modelled slippage. Your broker's contract notes are the actual figure; if the file carries them, the audit uses those instead.
Early access: this is the full audit. Sections marked Full audit will be paid later; the free report is everything else.
Fix these first
The rupee figures on these cards overlap: one trade can sit in a losing name, a big-loss day and the charges at once. They are the size of each leak, not savings that add up.
Costs take 43 % of what you make-₹59 k
Each trade makes +0.695 % before costs and pays 0.300 % in charges and slippage. If your real fills are a little worse than modelled -- twice the costs -- the average block makes +0.482 %.
Next: A trade should make at least two to three times its cost. Check the brokerage you entered, and if your fills are worse than the default slippage, believe the doubled-costs line. The fix is more profit per trade, not more trades: every extra trade pays the same charges.
Ten trades are 38 % of all your losses-₹54 k
Your ten worst trades lost ₹54,161 of the ₹1,41,212 lost in all. The three worst: 2026-05-07 HDFCBANK long (-₹8,737); 2026-05-25 INFY long (-₹6,291); 2026-05-08 INFY long (-₹6,114).
Next: Read these ten before anything else. What they share -- a name, a strategy, a hold, a size, a day -- is your first rule; the rest of the book is noise next to them.
3 days cost you 15 % of all your losses-₹21 k
A typical losing day costs ₹1,992; these lost at least three times that: 2026-05-07 (-₹8,737); 2026-05-25 (-₹6,291); 2026-05-08 (-₹6,114).
Next: The book is decided on a handful of days. A daily loss limit at about twice the typical losing day would have cut these; test it on the next month before trusting it.
6 more findings
Names you keep losing on: INFY, LT, BHARTIARTL-₹19 k
INFY (9 trades, -₹11,543, won 33 %); LT (10 trades, -₹5,085, won 40 %); BHARTIARTL (15 trades, -₹2,838, won 33 %). Together these 3 names cost ₹19,466.
Next: This is what these names did in your hands, not a view on the stocks. Look at your losing trades on them side by side -- entry time, hold, size, the setup you used -- and ask whether your rules treat them differently from the names that worked for you.
The market made the money, not the rules
Your trades averaged +0.782 % per block, but NIFTY was moving your way while you were in them (132 of 132 trades were long). Take NIFTY's move out and the same trades average -0.234 % per block: the rules added nothing.
Next: Judge the strategy on what it earns beyond the market. If it earns nothing beyond NIFTY, the rules are not doing the work; holding the index would have done the same with no rules at all.
Ten trades are 46 % of all your gains
Your ten best trades made ₹1,00,755 of the ₹2,18,913 made in all; the other 58 winners averaged +₹2,037.
Next: A book that lives on ten trades is fragile. Ask whether those ten could repeat by design or arrived by luck; the answer decides how much of this record to trust.
No edge shows on these trades
After costs the average block makes +0.782 %, -0.234 % once NIFTY's move is taken out: losing, and could still be a bad run. An edge this small against swings this big would take more trades to prove than any record holds.
Next: Nothing here says the rules work. Before more trades, look at the slices below for where the losses sit.
Symbol: INFY trades drag the book
9 trades (6 % of the book) averaged -0.861 % and won 33 % of the time. Without them the rest average +0.487 % instead of +0.395 %. A random set of 9 trades would do this well only 5.0 % of the time.
Next: A lead, not a rule: 21 slices were tried, so a few will look bad by chance. Keep taking these trades for now and re-check on the next 50; if it holds there, it is a rule you can write down.
Expect a losing run of at least 8 trades
Longest losing run 8 trades; deepest drawdown -₹48,060 (2026-06-11).
Next: Runs this long will happen again. Any size that cannot sit through this drawdown twice is too big for this record.
Where your money went
Costs ate 43 % of the gross profit.
Month by month
Running total
what you had kept, by exit day how far below your best point you were
More numbers
Where the points went Full audit
The rules are fixed and the same for everyone, so next month's score is comparable with this one. Confidence is low: it rests on 14 separate results (trades on the same day count as one), and a score on a few months of trading is a first read, not a verdict.
Edge 25 % of the score
Is the result better than luck, and is it yours rather than the market's
Starts at 0 from the steadiness of the gains (-1.68, where 2 counts as strong evidence).
- The market made the money, not the rules−20
Discipline 20 % of the score
The habits: jumping back in after a loss, flipping sides, names that keep losing
Starts at 100.
- 3 names you keep losing on: 14% of your losses−14 -₹19,466
Risk control 20 % of the score
How deep the holes get: big-loss days, the worst fall, losing runs
Starts at 100.
- 3 big-loss days: 15% of your losses−22 -₹21,141
- Worst fall 4% of capital−6 -₹48,060
Costs 15 % of the score
How much of what you make you keep after charges and slippage
Starts at 100.
- Costs took 43% of what you made before costs−43 -₹58,944
Consistency 10 % of the score
Does it hold month to month, recently, and on trades it was not tuned on
Starts at 100.
- 4 of 13 months lost money−18
Sizing & holds 10 % of the score
Where the money sits: your biggest trades, how long you hold losers, overnight vs same-day
Starts at 100.
No points lost.
This month against last month Full audit
2026-06 against 2026-05: better on net result, trades won, big-loss days; worse on names that lost last month. These are the last two months with at least 10 trades; 6 thinner months are left out of the comparison.
| 2026-05 | 2026-06 | ||
|---|---|---|---|
| Net result | -₹18,505 | +₹7,904 | Better |
| Trades won | 31 % | 47 % | Better |
| Re-entries after a loss | 0 | 0 | No change |
| Flip-flop trades | 0 | 0 | No change |
| Big-loss days | 2 | 1 | Better |
| Names that lost last month INFY -₹11,289 on 2 trades | 3 names | 1 lost again | Worse |
Every month
| Month | Trades | Net | Won | Re-entries | Flip-flops | Big-loss days |
|---|---|---|---|---|---|---|
| 2025-05 | 14 | +₹12,149 | 71 % | 0 | 0 | 0 |
| 2025-06 | 14 | -₹3,458 | 50 % | 0 | 0 | 0 |
| 2025-11 | 13 | -₹3,822 | 46 % | 0 | 0 | 0 |
| 2025-12 | 15 | -₹13,559 | 20 % | 0 | 0 | 0 |
| 2026-04 | 10 | +₹5,691 | 40 % | 0 | 0 | 0 |
| 2026-05 | 13 | -₹18,505 | 31 % | 0 | 0 | 2 |
| 2026-06 | 17 | +₹7,904 | 47 % | 0 | 0 | 1 |
Re-run the audit each month on your new trades; this section is how you see whether the habits named above actually stopped.
Your habits, in rupees Full audit
Your own trades, in rupees from your file. The fixes at the top name the ones that matter; these are the tables behind them.
Ten worst trades
| Date | Name | Side | Held | Net |
|---|---|---|---|---|
| 2026-05-07 | HDFCBANK | long | 6d | -₹8,737 |
| 2026-05-25 | INFY | long | 8d | -₹6,291 |
| 2026-05-08 | INFY | long | 7d | -₹6,114 |
| 2026-04-08 | ICICIBANK | long | 5d | -₹5,436 |
| 2025-03-25 | MARUTI | long | 10d | -₹5,322 |
| 2026-05-29 | INFY | long | 5d | -₹4,998 |
| 2025-05-26 | HDFCBANK | long | 7d | -₹4,555 |
| 2025-10-27 | SBIN | long | 8d | -₹4,451 |
| 2026-04-22 | BHARTIARTL | long | 7d | -₹4,170 |
| 2026-05-06 | HDFCBANK | long | 9d | -₹4,088 |
Ten best trades
| Date | Name | Side | Held | Net |
|---|---|---|---|---|
| 2025-04-07 | ITC | long | 15d | +₹17,540 |
| 2025-04-09 | ICICIBANK | long | 15d | +₹15,877 |
| 2025-04-11 | RELIANCE | long | 11d | +₹14,478 |
| 2025-03-11 | HDFCBANK | long | 13d | +₹10,245 |
| 2026-04-07 | KOTAKBANK | long | 9d | +₹8,740 |
| 2026-06-10 | BHARTIARTL | long | 8d | +₹7,705 |
| 2026-06-09 | ITC | long | 10d | +₹7,141 |
| 2026-04-02 | HDFCBANK | long | 7d | +₹6,792 |
| 2026-04-06 | KOTAKBANK | long | 7d | +₹6,125 |
| 2025-03-19 | AXISBANK | long | 5d | +₹6,113 |
Month by month
| Month | Trades | Net | Won |
|---|---|---|---|
| 2025-03 | 8 | +₹28,559 | 75 % |
| 2025-04 | 9 | +₹44,343 | 56 % |
| 2025-05 | 14 | +₹12,149 | 71 % |
| 2025-06 | 14 | -₹3,458 | 50 % |
| 2025-07 | 5 | +₹3,202 | 60 % |
| 2025-10 | 7 | +₹11,596 | 86 % |
| 2025-11 | 13 | -₹3,822 | 46 % |
| 2025-12 | 15 | -₹13,559 | 20 % |
| 2026-01 | 1 | +₹109 | 100 % |
| 2026-04 | 10 | +₹5,691 | 40 % |
| 2026-05 | 13 | -₹18,505 | 31 % |
| 2026-06 | 17 | +₹7,904 | 47 % |
| 2026-07 | 6 | +₹3,493 | 83 % |
Names that keep losing
Traded four times or more, net negative, won 40 % or less.
| Name | Trades | Net | Won |
|---|---|---|---|
| INFY | 9 | -₹11,543 | 33 % |
| LT | 10 | -₹5,085 | 40 % |
| BHARTIARTL | 15 | -₹2,838 | 33 % |
What you keep doing
- Size: the largest tenth of trades by margin (14) made +₹29,520 and won 64 %; the rest made +₹48,181 and won 50 %.
- Big-loss days (3): 2026-05-07 -₹8,737 on 1 trades; 2026-05-25 -₹6,291 on 1 trades; 2026-05-08 -₹6,114 on 1 trades.
When you trade: entries, exits, holds Full audit
From the trade list alone. 2 slices were tried against the random-skip null, so about 0 could pass by luck; a flag is a lead to re-check on new trades, never a rule.
Your file carries dates without times, so the entry-hour and exit-hour tables are empty. Add times to the entry and exit columns (2026-10-24 09:31) and they fill in.
Hold length
| Held | Trades | Share | Average | Won | Kept trades' average | Change | Luck p |
|---|---|---|---|---|---|---|---|
| 3-5 days | 20 | 15 % | +0.102 % | 55 % | +0.447 % | +13 % | 0.324 |
| 6-20 days | 112 | 85 % | +0.447 % | 51 % | +0.102 % | -74 % | 0.683 |
Next: no holds stand out against a random skip; nothing to change here.
Median hold: winners 8 days, losers 7 days (0.9x). The 54 trades held past the winners' median averaged +1.217 % and won 61 %.
Held to the close
Needs the instrument's close on each trade's exit day. Add a close column to a trade list and every trade is re-priced as if held to the close, same costs. A broker tradebook cannot carry it; the broker-data phase of the full audit will. What no trade list can show: the best and worst price each trade saw, so whether a time stop or an earlier exit would have helped.
The numbers behind the verdict skill or luck, the untouched test, capital, charges, every slice
Skill or luck: the numbers Full audit
The whole book. Your holds overlap: 90 % of trades are held 12 days or fewer, so trades are grouped into 25-day blocks and each block is one result (14 blocks from 132 entry days); counting each entry day would count the same market move several times. t per block is the score to quote. What the t columns mean: t is the steadiness score -- how many times bigger the average gain is than the swings around it. Under 1: could easily be luck. 1 to 2: maybe skill. Over 2: too steady to be luck. t per trade counts every trade; t neutral is the score after NIFTY's own move during each trade is taken out.
| Blocks | Trades | Gross / block | Net / block | t per block | t per trade | t neutral | Blocks won | |
|---|---|---|---|---|---|---|---|---|
| All trades | 14 | 132 | +1.082 % | +0.782 % | 1.53 | 1.06 | -1.68 | 64 % |
| Last 12 months | 8 | 73 | +0.661 % | +0.361 % | 0.52 | -0.28 | -1.84 | 50 % |
| Without best and worst day | 14 | 130 | +1.030 % | +0.729 % | 1.60 | 1.17 | -1.91 | 64 % |
| Half: 2025-03-04 .. 2025-11-12 | 8 | 66 | +1.436 % | +1.135 % | 1.89 | 1.76 | -0.45 | 75 % |
| Half: 2025-11-13 .. 2026-06-30 | 7 | 66 | +0.747 % | +0.447 % | 0.56 | -0.29 | -1.47 | 57 % |
Best day 2025-04-07 (+₹17,540), worst day 2026-05-07 (-₹8,737).
Trading more than one strategy? Download the trades at the end of this page, fill the strategy column, and upload it with your file: each one is then graded on its own.
Return on your capital Full audit
Capital here is the peak margin the book needed -- the least capital that would have carried it. Enter your real capital on the upload page for the true percentages. Margin per trade is a modelled fraction of turnover (the fractions are in "How this was read"); an option position uses its own base.
The file carries dates without times, so every trade of a day counts as open together; the peak is conservative.
| Segment | Margin assumed | Trades | Avg margin | Net ₹ | RoM / trade, mean | median | RoM / margin-day, median | Won |
|---|---|---|---|---|---|---|---|---|
| Cash delivery | 100.0 % | 132 | ₹1,48,943 | +₹77,701 | +0.39 % | +0.11 % | +0.01 % | 52 % |
RoM per margin-day = net / (margin × days held, at least one): the unit that compares segments on capital rather than on turnover.
Did it keep working on untouched trades Full audit
No split date was given, so the last 30 % of trading days (from 2026-04-16) are held out and judged on their own. If you tuned the strategy on all of these trades, this is not a true out-of-sample test -- tell us the date you stopped tuning. PASS needs a positive net per block, a steadiness of 1.5 or more after the market's move is taken out, and a positive market-neutral average.
| Blocks | Trades | Gross / block | Net / block | t per block | t per trade | t neutral | Blocks won | |
|---|---|---|---|---|---|---|---|---|
| FAIL Before | 11 | 92 | +1.285 % | +0.985 % | 1.57 | 1.57 | -1.21 | 64 % |
| FAIL After | 4 | 40 | +0.028 % | -0.271 % | -0.49 | -0.52 | -4.24 | 50 % |
Year by year Full audit
| Blocks | Trades | Gross / block | Net / block | t per block | t per trade | t neutral | Blocks won | |
|---|---|---|---|---|---|---|---|---|
| 2025 | 9 | 86 | +1.108 % | +0.807 % | 1.44 | 1.40 | -1.03 | 67 % |
| 2026 | 5 | 46 | +1.037 % | +0.736 % | 0.66 | -0.03 | -1.39 | 60 % |
What charges do to you Full audit
Average cost 0.300 % per trade against an average gross +0.695 %. If your real fills are worse than modelled, the 2x line is the one to believe.
| Net / block | t per block | Blocks won | |
|---|---|---|---|
| 0x costs | +1.082 % | 2.11 | 71 % |
| 1x costs | +0.782 % | 1.53 | 64 % |
| 2x costs | +0.482 % | 0.94 | 64 % |
Which trades to question Full audit
1 of 21 slices pass the bar (skipping them lifts the kept trades' average by 10 % or more, refuses no more than 40 % of trades, and beats a random skip of the same size at 5 %). 21 slices were tried, so about 1 could pass by luck alone: treat a flag as an idea to test on new trades, never a rule.
| Slice | Trades | Share | Its average | Won | Kept trades' average | Change | Luck p |
|---|---|---|---|---|---|---|---|
| Symbol: INFY | 9 | 7 % | -0.861 % | 33 % | +0.487 % | +23 % | 0.050 |
| After: a losing trading day | 64 | 48 % | -0.143 % | 45 % | +0.901 % | +128 % | 0.007 |
| Symbol: LT | 10 | 8 % | -0.343 % | 40 % | +0.455 % | +15 % | 0.169 |
| Symbol: BHARTIARTL | 15 | 11 % | -0.126 % | 33 % | +0.462 % | +17 % | 0.206 |
| Hold: 3-5 days | 20 | 15 % | +0.102 % | 55 % | +0.447 % | +13 % | 0.292 |
| Symbol: MARUTI | 8 | 6 % | -0.057 % | 50 % | +0.424 % | +7 % | 0.321 |
| Weekday: Thursday | 25 | 19 % | +0.166 % | 52 % | +0.448 % | +14 % | 0.323 |
| Symbol: HDFCBANK | 13 | 10 % | +0.070 % | 38 % | +0.430 % | +9 % | 0.355 |
| Symbol: AXISBANK | 9 | 7 % | +0.022 % | 44 % | +0.422 % | +7 % | 0.361 |
| Weekday: Friday | 25 | 19 % | +0.364 % | 56 % | +0.402 % | +2 % | 0.496 |
| Symbol: SBIN | 18 | 14 % | +0.410 % | 61 % | +0.392 % | -1 % | 0.501 |
| Symbol: TCS | 8 | 6 % | +0.438 % | 62 % | +0.392 % | -1 % | 0.560 |
| Weekday: Monday | 27 | 20 % | +0.432 % | 52 % | +0.385 % | -2 % | 0.561 |
| Weekday: Wednesday | 30 | 23 % | +0.431 % | 50 % | +0.384 % | -3 % | 0.562 |
How this was read
- costs in rupees from each trade's own turnover: statutory stack (2026-10 (STT per Budget 2026, eff 1-Apr-2026)) + brokerage Rs 20 per order + GST + slippage Cash delivery 0.045 % RT (measured defaults)
- MARKET-NEUTRAL: each trade minus the NIFTY move over its hold, signed for its side (a same-day trade uses the previous close -> close, a proxy for an intraday window)
- MARGIN per trade = Cash delivery 100.0 % of turnover (echoing the measured fractions behind this audit) -- ASSUMED; your broker's margin statement is the truth
- OVERLAP: 90 % of trades are held 12 days or fewer, so trades are grouped into 25-day calendar blocks (14 blocks from 132 entry days) and each block counts as one result -- trades open on the same days are not independent
- CONCURRENT MARGIN: the file carries dates without times, so every trade of a day counts as open together (conservative)
Not checked here
- Whether the rules used any information that was not available at the moment of the trade (look-ahead). Only you can confirm this.
- Whether these exact trades were chosen after trying many variants. A strategy picked as the best of 50 needs a far higher bar.
- Fills: limit orders that would not have filled, circuit limits, and impact on thin names are not modelled beyond the slippage line.
- Margin, leverage and position sizing: returns are per trade on its own turnover.
Your trades, as read
Every trade this audit counted, one row each, with a strategy column to fill in. Name your strategies in it, then upload it together with your trades file and each strategy gets its own grade.
Name your strategies
Your file carries no strategy names, so the audit grouped trades into types it can see. Name your own: describe each strategy the way you think of it and every trade that fits attaches to it; a trade that fits two goes to the first one listed; the rest stay "not named". Then re-grade and each strategy gets its own skill-or-luck read, its own habits, its own rupees. Nothing is saved on our side: download the strategies and upload them with next month's trades to keep the same names.
Or tag a few trades and let the audit work out the rule
Type the strategy name against five or six trades of each kind below, press Work out the rules, and the audit says what those trades have in common and how many more fit. A name typed against a trade always wins over a rule.