Bottom line up front
Only the most recent public fills are visible, so this audit covers the data covered rather than full account history. This account is loss-making in the data covered: -$188.50 realised PnL across 3,299 closed episodes over two days of trading. The headline pattern is systematic: no edge on any single instrument, a 34.71% win rate, and a profit factor of 0.56 that signals losses outpace wins by a factor of nearly two. The account exhibits acute behavioural distress—revenge trades, FOMO re-entries, and oversized losers dominate the record. Fees of $95.48 consumed the entire residual edge; without execution costs, the account would still be underwater by $93.71.
What the data shows
The account opened on 18 May 2026 and has traded continuously through 20 May. In the data covered, it has cycled through 3,306 total episodes (3,299 closed, 7 open), with a highest balance in this window of $44,858,968.03 on 19 May and a lowest balance in this window of $44,014,623.63 at the current snapshot. The deepest decline in this window was negligible in percentage terms, but the account's realised loss of $93.71 before fees reveals a structural problem: the account is losing money on the core trading logic, not just on friction.
Money is lost across every major instrument. TON, the most-traded coin with 320 episodes, lost $22.89 realised. LIT (244 episodes) lost $22.52. BIO (135 episodes) lost $20.07. INJ, ZEC, VVV, and ONDO all show no edge. The only instrument with a positive verdict is XMR, which generated $4.29 realised PnL across 59 episodes—a rare bright spot in an otherwise uniform loss distribution. Long positions lost $78.12; short positions lost $110.37. Neither side shows an advantage; both are underwater.
Fees paid total $95.48 gross, representing a net fee drag of $95.48 after rebates. Against a realised loss of $93.71, fees nearly doubled the damage. The account's gross volume was $906,638.55, with a maker ratio of 64.9%, suggesting reasonable execution intent, but the underlying trades are unprofitable before costs are even applied.
Trade quality
Win rate is 34.71%. Profit factor is 0.56. Expectancy is -$0.06 per trade. Win/loss ratio is 1.06, meaning the account wins slightly more often than it loses, but when it loses, the average loss (-$0.20) exceeds the average win (+$0.21) by a meaningful margin. This is a losing system. The 10-trade win streak and 16-trade loss streak both occurred in the data covered, indicating volatility in sequence but no underlying consistency.
Post-mortems
BIO short, 18 May, 0.04 entry to 0.04 exit, -$6.35 loss. This trade is flagged as both an oversized loser (6.14x the median loss) and a revenge trade following a $0.164556 loss on BIO long minutes earlier. The position notional was $618.62 at maximum, held for 0.02 hours. The structural stop was 4% away; the account hit the loss without triggering it, suggesting the position was closed manually or the stop was not active. This is a textbook revenge trade: after losing on BIO long, the account immediately flipped to short, sized up, and lost again.
LIT short, 19 May, 1.08 entry to 1.09 exit, -$4.11 loss. Also flagged as oversized loser and revenge trade. The position notional was $457.74. Held 0.03 hours. This followed a series of losses on other instruments and represents a second attempt to recover via a different coin and direction. The loss is 4.54x the median loss size. The pattern repeats: loss, then a larger, faster trade in a different instrument, then another loss.
What the risk simulation reveals
Under a 1% risk rule applied historically to this account's trade sequence, the simulated result would have been a -$7,714,397.67 loss with a maximum decline of -17.81%. Under 2%, the simulated loss would have been -$15,428,795.35 with a -35.58% decline. Under 4%, the simulated loss would have been -$30,857,590.70 with a -70.99% decline. These are gross-of-fees figures. The simulator shows that if this account had been sizing positions to fixed risk percentages rather than fixed notional amounts, the losses would have been catastrophic. The account's actual loss is small only because position sizes are modest relative to the $44M balance; the underlying trade logic is deeply negative.
Open positions
Five positions remain open, all small and unhedged:
- kPEPE short, 10x leverage, entry 0.003632, unrealised -$0.011688. No stop in place.
- NEAR short, 10x leverage, entry 1.6079, unrealised +$0.00432. No stop in place.
- ONDO long, 10x leverage, entry 0.36969, unrealised +$0.07824. No stop in place.
- BIO short, 3x leverage, entry 0.033997, unrealised +$0.004194. No stop in place.
- VVV long, 3x leverage, entry 16.4899, unrealised -$0.015839. No stop in place.
None of the open positions have stops. The largest unrealised loss is on kPEPE at -$0.011688; the largest unrealised gain is on ONDO at +$0.07824. These are immaterial in absolute terms, but the absence of stops across all five positions is a structural risk gap.
Honest summary
- Strength: The account maintains a 64.9% maker ratio, indicating an intent to provide liquidity and reduce slippage costs. This is the only visible execution discipline in the record.
- Weakness: Revenge trading is endemic. Five separate revenge-trade episodes are flagged, all occurring within hours of prior losses on different instruments. The account cycles through coins and directions rapidly after losses, a pattern that guarantees compounding of losses rather than recovery.
- Weakness: Oversized losers dominate the loss distribution. MORPHO lost $2.15 (23.78x median loss), DYM lost $0.92 (10.13x median loss), PENDLE lost $0.56 (6.14x median loss). The account does not scale down after losses; it scales up or maintains size, then loses more.
- Weakness: No instrument shows an edge. The only positive instrument is XMR at +$4.29 across 59 episodes. Every other coin is negative. This is not a matter of execution or timing; the underlying signal is absent.
- Data scope: The sample is only two days old and covers only the most recent 10,000 fills. Behavioural conclusions drawn from this window should be treated as snapshots of current state, not as stable patterns. The account may have been trading differently before the data covered began.