Bottom line up front
Open MU short dominates this wallet: $469,444.54 notional, $86,390.31 unrealised, 155.42% unrealised ROE, 10x cross margin, liquidation at $4,950.73. This is still open exposure; the unrealised gain is not locked in, and the final outcome is unknown until the position closes. The liquidation level sits far above current mark price, but the position remains the single largest risk vector on the account. Behind that open trade, the closed book is profitable: +1.11% net in the data covered, with $61,440.58 realised PnL across 69 closed episodes. However, only the most recent public fills are visible, so this audit covers the data covered rather than full account history. The account's strength lies in long-side edge (VVV, ORCL, MSTR, EWY all won cleanly), but the closed losses cluster heavily on a single day of revenge trades and oversized shorts that turned sour.
What the data shows
The account opened on 6 May 2026 and has been active for 65 days. The highest balance in this window was $7,023,691.40 on 6 May; the lowest balance was $4,688,300.05 on 3 June. The deepest decline in this window was -33.25%, a sharp deepest decline in this window that maps to the day of heavy losses (6 May) when five separate revenge trades and oversized positions blew up in quick succession.
Closed realised PnL stands at $59,786.46 after $545.06 in net fees. Fees consumed 0.91% of gross PnL—a modest drag, but material given the account size. Long trades generated $59,150.39 of the $61,440.58 total profit; short trades contributed only $2,290.19. This asymmetry is stark: long win rate was 51.43%, short win rate 55.88%, yet longs carried the entire edge. The short-side wins were scattered and small; the short-side losses were concentrated and severe.
The account's edge is narrow and instrument-specific. VVV (7.99 hours, +$59,540.67), ORCL (3.65 hours, +$6,534.74), MSTR (2.3 hours, +$3,003.24), EWY (0.67 hours, +$2,584.83), and TSLA (100% win rate across 2 episodes) form a clean primary-edge cluster. These five instruments account for roughly $72,663 of the $61,440 realised profit. Everything else—SNDK, XYZ100, MU, CRCL, CL—was a no-edge grind: 11 episodes in SNDK for -$2,913.73 realised, 14 in XYZ100 for -$2,855.05, 7 in MU for -$2,761.41. The account traded these instruments repeatedly despite consistent losses, a pattern consistent with the revenge-trade and averaging-down flags.
Trade quality
Win rate: 53.62% across 69 closed episodes. Profit factor: 3.63. Expectancy: $890.44 per closed trade. Win/loss ratio: 3.14 (average winner $2,293.02, average loser -$731.28).
These metrics are strong in isolation. A 3.63 profit factor and 53.62% win rate with a 3.14 win/loss ratio represent genuine edge. However, the distribution is heavily skewed: the top five wins account for $72,663 of the $59,786 realised PnL. Remove those five trades and the account is underwater. The median loss ($731) is small, but five trades exceeded 6x the median loss (MU long -$3,933.50, CL long -$2,658.64, SNDK long -$2,126.43, SNDK long -$2,033.58, SNDK long -$1,686.56). These were all opened on 6 May, all flagged as oversized losers and revenge trades, and all closed within minutes to hours.
Post-mortems
CL long, 6 May 10:50–10:57 UTC, entry 90.90, exit 90.34, -$2,658.64
Opened immediately after a -$308.29 loss in SP500. Max notional $288,365.86 on 40x leverage. Duration 4 minutes. MAE -1.54%, MFE +1.29%. Structural stop (ATR 14 1H) was 3.18% away; the trade hit -1.54% and reversed slightly, but exited at the lows. Flagged as averaging down, oversized loser, and revenge trade. The position was sized aggressively into a micro-timeframe move after a small prior loss, a textbook revenge setup.
SNDK long, 6 May 13:29–13:52 UTC, entry 1,428.32, exit 1,419.76, -$2,126.43
Opened after a -$210.28 loss in NVDA. Max notional $332,365.41 on leverage. Duration 13 minutes. MAE -3.36%, MFE +2.73%. Structural stop was 2.75% away; the trade moved against it immediately and never recovered. Flagged as averaging down, oversized loser, and revenge trade. The account then re-entered SNDK twice more that day, losing an additional -$2,033.58 and -$1,686.56 in subsequent episodes.
What the risk simulator reveals
Under a 1% hard stop rule, the account would have realised -$560,279.41 PnL with a deepest decline in this window of -11.35%. Under 2%, -$1,120,558.82 and -22.37%. Under 4%, -$2,241,117.64 and -43.44%. These are gross-of-fees simulations. The actual account ran without hard stops and achieved +$61,440.58 with a -33.25% deepest decline. The simulator reveals that the account's profitability is entirely dependent on the absence of mechanical stops. The largest winners (VVV +$59,540.67, ORCL +$6,534.74) had no entry prices recorded and no structural stops defined, suggesting they may have been inherited positions or filled outside the standard order flow. The revenge trades and oversized losers that triggered the -33.25% deepest decline in this window would have been arrested by a 1–2% rule, but so would the account's edge. This is not a compliment to the account's discipline; it is a statement of fragility.
Open positions
The dominant open position is MU short: $469,444.54 notional, $86,390.31 unrealised gain, 155.42% unrealised ROE, 10x cross margin, liquidation at $4,950.73. No stop in place. This position is still open; the unrealised gain is not locked in. The liquidation price sits far above current mark, but the position remains the account's largest single exposure and the primary driver of current account value.
Four other open positions are present: BTC short ($24,384.01 notional, +$59.64 unrealised, 9.76% ROE, 40x, liquidation $2,489,500.84); SOL long ($9,739.30 notional, -$825.76 unrealised, -156.32% ROE, 20x); SUI