Bottom line up front
Only the most recent public fills are visible, so this audit covers the data covered rather than full account history. The account is exceptional in the data covered: +785k USD realised PnL on a starting estimate of -353k, a 74% win rate, and a profit factor of 50.71. The edge is almost entirely short-biased: shorts generated 788k while longs lost 3.7k. The fragility is acute: the deepest decline in this window reached -99.82%, and two of the three largest losses were preceded by averaging down and revenge trading after smaller losses. Without position sizing discipline, this account would have been wiped.
What the data shows
The account opened into a volatile window and made its money on three directional shorts: BTC short from 93,478 to 90,678 (248k), ETH short from 3,151 to 2,977 (241k), and HYPE short from 38.7 to 37.29 (192k). These three trades account for 682k of the 790k realised PnL. The highest balance in this window was 1,988k on 19 November, reached after the BTC short closed profitably. The lowest balance in this window was 3,557 on 3 December—a 99.82% decline from highest balance in this window.
The short side worked because the account sized aggressively into directional conviction and held through volatility. The long side failed because it became a vehicle for revenge and averaging down. SOL long opened 18 November at 140.37, accumulated to 2,063 contracts across 67 add events, and closed 25 November at 136.93 for a -10,894 loss. BTC long opened 24 November at 87,710, reached a notional of 1.126m, and closed 25 November at 87,554 for a -3,186 loss. Both trades carried structural stops (ATR 14 1H) that were never triggered, implying the account held through them or disabled them mid-position.
Fees consumed 5,059 USD gross, or 0.64% of realised PnL. The account is a 94.5% maker, so fees are minimal relative to volume. Net fee drag is immaterial to the headline.
Trade quality
Win rate of 74.19% across 31 closed episodes. Profit factor of 50.71—meaning gross wins are 50.71x gross losses. Average win of 34,817 USD; average loss of -1,974 USD. Win/loss ratio of 17.64. Expectancy of 25,323 USD per closed trade.
These numbers describe an account with genuine edge on the short side and severe discipline failure on the long side. The profit factor is inflated by the three mega-wins; strip those and the account is much tighter. The win rate is real, but it masks concentration: 23 BTC episodes at 78% win rate, 2 ETH episodes at 100%, 1 HYPE at 100%, 4 SOL at 25%. The long/short split is stark: shorts 76.47% win rate, longs 71.43% win rate, but shorts +788k and longs -3.7k.
Post-mortems
SOL long, 18–25 November, 140.37 to 136.93, -10,894 USD. Opened 18 November at 140.37. Averaged down 67 times, reaching a max notional of 288k. Structural stop (ATR 14 1H) was 2.21% below entry; the account held through it. Closed 25 November at 136.93. This is the largest loss in the data covered and was flagged as averaging down, oversized loser. The account added into a losing position across 160 hours, violating basic risk discipline.
BTC long, 24–25 November, 87,710 to 87,554, -3,186 USD. Opened 24 November at 87,710. Max notional reached 1.126m. Closed 25 November at 87,554 after 33.69 hours. Flagged as averaging down, FOMO re-entry, oversized loser, and revenge trade. The previous BTC long closed at 88,343 on 21 November; this re-entry came at 82,242 (lower), then the account re-entered again at 87,710 on 24 November. Structural stop was 1.44% away; the account held through it. The position size relative to the loss magnitude suggests the account was chasing back a prior deepest decline in this window.
BTC long, 22–24 November, 83,913 to 86,408, +9,110 USD. This was a win, flagged as averaging down. Opened 22 November at 83,913, reached 461k notional, closed 24 November at 86,408. The account added into this position and it worked. This is the only long-side win of material size in the data covered.
What the risk simulator reveals
Under a 1% stop rule, the account would have realised 185,698 USD with a max decline of -9.52%, stopping out 2 episodes early. Under a 2% stop rule, 371,396 USD with -12.82% max decline. Under a 4% stop rule, 742,792 USD with -15.51% max decline. These are gross of fees.
The simulator shows that mechanical stops would have prevented the 99.82% decline and still left the account with substantial profit. The SOL and BTC long losses would have been capped at 2–4% of notional rather than allowed to run to full loss. The short-side winners would have been largely unaffected because they moved in the account's favour quickly.
Open positions
No open positions at the time of this analysis.
Honest summary
- Short-side edge is real. Three shorts generated 682k of 790k realised PnL. Win rate on shorts is 76.47%. The account sized into directional conviction and held through volatility. This is the core skill.
- Long-side discipline is absent. Four long episodes, 25% win rate, -10.9k realised PnL. The two largest losses were preceded by averaging down across 67 and 82 add events respectively. Revenge trading and FOMO re-entries are documented in the behavioural flags. Structural stops were set but not honoured.
- Position sizing is asymmetric and fragile. The account held 1.126m notional in a single BTC long that lost 3.2k. The same account sized 1.51m in the winning BTC short. Without the three mega-wins, the account would be underwater. The 99.82% decline in this window is not a deepest decline in this window—it is evidence of near-total loss followed by recovery on a single trade.
- Sample is capped at 10k fills. This audit covers only the most recent public fills. Earlier history is not visible. Behavioural patterns observed here may not persist or may be artefacts of the analysed window.