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 down 99.4% in the data covered, having collapsed from a highest balance in this window of $736,642 to a lowest balance in this window of $1.59. The deepest decline in this window was -100%. BTC was the only instrument with edge; everything else—kPEPE, ETH, DOGE, ASTER, HYPE—was loss-making. The account exhibits severe behavioural dysfunction: five separate revenge trades, five averaging-down episodes that compounded losses into catastrophic positions, and two trades (kPEPE on 27 July and ETH on 7 January) that alone account for $391,300 in losses. Fees consumed $55,354 of gross volume, but the core problem is position sizing and loss recovery discipline, not execution cost.
What the data shows
The account opened on 16 July 2025 with approximately $207,862 and reached $736,642 by the same day—a 255% intraday move driven by a single BTC long that ran from entry at $118,309 to exit at $119,002, netting $232,130. This was the account's only clean, unaveraged win of material size. From that highest balance in this window, the account entered a systematic deepest decline in this window. By 6 August, it had fallen to $1.59, a complete wipeout within three weeks.
The loss sequence reveals two distinct failure modes. First, the kPEPE trade opened 27 July at $0.01 per token, was averaged down 12 times to a maximum notional of $3.28 million, and closed 29 July at $0.01, realising -$214,549. The trader then re-entered kPEPE on 25 July (after the 29 July close, suggesting the data window includes overlapping or corrected fills), averaged down again, and lost a further -$37,802. Second, the ETH trade opened 7 January 2026 at $3,252.31, was averaged down twice, reached a notional of $5.33 million, and closed 14 January at $3,138.66, realising -$176,751.
Between these two catastrophes, the account executed five revenge trades: ETH on 23 August following a DOGE loss, BTC on 2 October following a HYPE loss, kPEPE on 5 October following a BTC loss, BTC on 16 October following a kPEPE loss, and kPEPE on 17 October following another kPEPE loss. Three of these revenge trades (BTC on 2 October, BTC on 16 October, kPEPE on 5 October) were oversized losers in their own right, totalling -$137,465 in realised losses.
Long positions account for -$211,880 of realised PnL across 32 episodes with a 21.9% win rate. Short positions account for +$5,297 across 19 episodes with a 36.8% win rate. The asymmetry is stark: the account has no edge on the long side and a marginal edge on the short side, yet 84% of capital was deployed long. BTC alone generated +$212,962 realised PnL across 23 episodes at 39.1% win rate—the only instrument with positive expectancy. Every other instrument was loss-making: kPEPE -$197,893 across 10 episodes at 20% win rate, ETH -$189,626 across 4 episodes at 25% win rate, DOGE -$22,384 across 2 episodes at 0% win rate.
Gross fees paid were $55,354 on $141.8 million in gross volume. The maker percentage was 0.68%, indicating mostly taker fills. Fee drag alone would have been manageable; the realised PnL before fees was -$174,490, meaning fees added only $32,000 to the net loss. The core damage was position sizing and averaging into losers.
Trade quality
Win rate was 27.5% across 51 closed episodes. Profit factor was 0.72, meaning every dollar won generated $1.39 in losses. Expectancy was -$4,051 per trade. Win/loss ratio was 1.89 (average win $37,120 vs. average loss -$19,629), a decent ratio that was overwhelmed by the frequency of losses and the size of the largest losses. The longest loss streak was 10 consecutive closed trades. The longest win streak was 3 trades.
These numbers describe an account with no statistical edge. A 27.5% win rate with a 1.89 win/loss ratio requires an expectancy of approximately -$3,000 per trade to break even after fees; this account's actual expectancy of -$4,051 confirms the account is underwater on a per-trade basis.
Post-mortems
kPEPE long, 27 July – 29 July 2025. Entry at $0.01, exit at $0.01, -$214,549 realised loss. The position was averaged down 12 times to a maximum notional of $3.28 million. The trade lasted 49 hours. A structural stop was set at 5% below entry (instrument default), but the position was allowed to grow to 328x the initial notional, consuming the entire account equity multiple times over. This is not a trade; it is a liquidation event. The averaging-down flag and oversized-loser flag are both present.
ETH long, 7 January – 14 January 2026. Entry at $3,252.31, exit at $3,138.66, -$176,751 realised loss. The position was averaged down twice to a maximum notional of $5.33 million. The trade lasted 180 hours. A structural stop was set at 1.36% below entry (ATR 14 1H), but again the position was allowed to balloon far beyond the initial risk envelope. This trade followed a sequence of losses and exhibits the averaging-down and oversized-loser flags.
BTC long, 2 October – 16 October 2025. Entry price not recorded, exit at $112,192.45, -$65,058 realised loss. Maximum notional $3.87 million. This trade was flagged as a revenge trade following a HYPE loss of -$431.79. The account risked $65,000 to recover $432. Duration 344 hours. Oversized-loser flag present.
BTC long, 16 October – 19 October 2025. Entry price not recorded, exit at $109,585, -$40,365 realised loss. Maximum notional $1.13 million. This was a revenge trade following the previous BTC loss. Duration 75 hours. Oversized-loser flag present.
kPEPE long, 25 July – 25 August 2025. Entry price not recorded, exit at $0.01, -$37,802 realised loss. Maximum notional $381,369. Duration 645 hours. This was the second kPEPE position, opened after the first kPEPE catastrophe closed. Oversized-loser flag present.
What the risk simulator reveals
Under a 1% hard stop rule applied retroactively to all 51 closed episodes, the account would have realised -$3,905 with a maximum decline of -3.32% and a win rate of 9.5%. Under a 2% rule, the loss would have been -$