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. Every lost point, with its reason and rupees, is in the full audit.
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.
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 in the full audit
Names you keep losing on: INFY, LT, BHARTIARTL · The market made the money, not the rules · Ten trades are 46 % of all your gains · No edge shows on these trades · Symbol: INFY trades drag the book · Expect a losing run of at least 8 trades.
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
How this was read
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.