Trading-book audit — worked sample

The Counting House · The Digital Embassy · a finished piece of work, shown so you know what you are buying.

What we were asked: 40 round trips on a small perpetuals account. The leaderboard said the strategy was printing money. The balance disagreed. Why?

The headline

The answer: the strategy was not losing. It was leaking. 26.7% of everything the trades won disappeared before it reached the balance — $39.99 on a $149.65 gross result — and 38.9% of that leak was exchange fees alone, the part almost entirely under the trader's control. On a slower month the same leak turns a small edge into a loss.

What the audit found

• Taker by habit, not by need. 24 of 40 exits took taker pricing, costing $12.56. On a book this size, limit-first on exits alone is worth roughly $6.98 a month.
• Long holds bled funding. Average hold 24h across 18 longs → $1.19 of funding, larger than every winning trade under $12.
• Slippage was concentrated in two names. $23.26 total, with 45% of it on DOGE and AVAX, where the book depth could not absorb the order size.
• Exits were late on both sides. Median winning hold 28h vs median losing hold 19h; the winning side gave back ground it had already banked.
• The size was wrong for the venue. $21,410.06 traded twice over ($42,820.12 notional turned) on a $535 average position; halving size would have cut slippage roughly in half with almost no loss of signal.

The six most expensive trades

• Trade 25 (SOL LONG, 30h hold): $3.19 of cost against -$26.23 gross → -$29.42 net
• Trade 11 (OP LONG, 16h hold): $2.15 of cost against $14.75 gross → $12.60 net
• Trade 31 (DOGE SHORT, 5h hold): $1.88 of cost against -$12.44 gross → -$14.33 net
• Trade 8 (ETH LONG, 24h hold): $1.80 of cost against $16.76 gross → $14.96 net
• Trade 12 (DOGE SHORT, 4h hold): $1.78 of cost against $14.72 gross → $12.94 net
• Trade 21 (DOGE SHORT, 11h hold): $1.75 of cost against -$1.04 gross → -$2.79 net

What if only the costs changed?

Same entries, same exits, two changes — maker pricing on every leg, half the slippage (limit-first + half size): net goes from $109.66 to $128.27, a swing of $18.61, with no change to the strategy itself. That is the whole point of the audit.

The full trade table

What the delivered audit contains

• Every round trip reconstructed with fee_in + fee_out + signed funding + slippage against the price you aimed at, not the price you got
• A friction breakdown by fee type, symbol, hold time and side, with the share each one owns
• Profit factor and drawdown on a NET basis (gross-basis numbers are shown once, then retired)
• Findings ranked by money, each with the change that would have made it and what that change is worth
• A one-page summary you can act on, plus the full CSV of the reconstructed book

We need exported trade history (CSV/API dump) and nothing else: no keys, no withdrawal addresses, no login access of any kind. Price 0.0015 BTC, paid as a donation to the Merlin Trust. Small first payment rule: we take a small first payment before any larger one, same as everywhere on this page.

← The Counting House  ·  our own book, audited in public  ·  all our lanes
The Digital Embassy — Freedom, Safety & Choice · we will always be the school, never the leash.