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 profitable in this window: $313,586 realised PnL on 23 closed episodes, but the path was violent. The highest balance in this window reached $8.45M on 24 December; the lowest balance was $48.5k on 22 October. The deepest decline in this window was 81.35%, driven by a single $1.18M loss on an ETH long opened 3 November and closed 26 November. The core edge is short-side ETH—90% win rate, $788k profit—but long-side attempts have destroyed more than half that gain. Averaging down is the dominant behaviour: five major positions show repeated size additions, and the two largest losses both exhibit this pattern.
What the data shows
This is a single-instrument account trading ETH across 23 closed episodes in a 73-day window. The short side generated $788k profit on 10 wins from 11 episodes (90% win rate); the long side lost $475k across 13 episodes despite an 84.62% win rate, a paradox explained by one catastrophic position. The account opened with modest capital, scaled aggressively into winning positions, and then suffered a structural failure in position sizing discipline.
The largest win came early: an ETH short opened 14 October at $4,026, closed 30 October at $3,972, yielding $881k. The position was built through 77 averaging-down events, reaching a notional of $16.5M. This trade worked because the direction was correct and the account had room to compound. The second-largest win—$589k—came from an ETH long opened 27 November at $2,880, closed 2 December at $3,007, also built through averaging. Both trades show the same pattern: directional conviction, size accumulation, and profitable exits before the position became unmanageable.
The largest loss inverted this logic. On 3 November, the account opened an ETH long at $3,298, averaged down 54 times, reached a notional of $15.5M, and exited on 26 November at $3,094 for a loss of $1.18M. This position was 5.88 times larger than the median loss and consumed all profits from the short side and more. The structural stop was set at 1.23% away (ATR 14 1H), but the position was allowed to run 6.6% underwater before closure. A second major loss followed: an ETH short opened 2 December at $3,084, averaged down 4 times, reached $19.3M notional, and closed 26 December at $3,150 for a loss of $201k.
Fees consumed $37.8k net of rebates, or 10.83% of realised PnL. This is material drag but not catastrophic; the account made $351k gross before fees, leaving $313.6k net. The maker rate of 68% suggests some passive liquidity provision, which helped offset execution costs.
Long versus short asymmetry is the defining weakness. Shorts won at 90% and generated positive expectancy; longs won at 84.62% but destroyed capital through oversized losers. The account appears to have a directional bias toward short-side conviction but lacks the discipline to cut long-side losses before they metastasize.
Trade quality
Win rate of 86.96% is exceptional; profit factor of 1.23 is weak. This inversion—high win rate, low profit factor—is a textbook signature of a few large losses overwhelming many small wins. Average win was $84.8k; average loss was $460.8k. The win/loss ratio of 0.18 confirms the pattern: losses are 5.5 times larger than wins on average. Expectancy of $13.6k per episode is positive but fragile; it depends entirely on the short-side edge. Removing the single largest loss would double expectancy; removing the two largest losses would triple it.
Post-mortems
ETH long, 3–26 November, $3,298 entry to $3,094 exit, −$1.18M loss. This position was opened with conviction and built through 54 averaging-down events to a highest balance in this window notional of $15.5M. The structural stop was set at 1.23% (ATR 14 1H), implying a liquidation boundary near $3,258. The position ran 6.6% underwater before closure, well beyond the stop distance, suggesting either a stop was not enforced or the account had sufficient margin to absorb the draw. The loss is flagged as an oversized loser—5.88 times the median loss—and the averaging behaviour indicates the account was fighting the position rather than accepting the initial thesis failure. This is the single largest destroyer of capital in the data covered.
ETH short, 2–26 December, $3,084 entry to $3,150 exit, −$201k loss. Opened with averaging-down behaviour (4 events), reached $19.3M notional, and closed after 563 hours. The structural stop was 1.55% away; the position ran 2.1% underwater. This loss is smaller in absolute terms but followed immediately after the account had recovered to near-highest balance in this window balance. The averaging pattern suggests the account was again adding to a losing position rather than cutting it early.
What the risk simulator reveals
Under a 1% fixed stop-loss rule applied historically, the account would have realised $93.5k with a deepest decline in that scenario of 57.75%. Under 2%, the result would have been $186.9k with a deepest decline of 73.08%. Under 4%, the simulated outcome would have been $373.9k with a deepest decline of 84.27%. The simulator stopped 3 episodes early across all three rules, indicating that some positions would have been force-closed before natural exit. The actual result of $313.6k sits between the 2% and 4% simulations, suggesting the account's effective stop discipline was loose—closer to 4% than to 1% or 2%. A tighter stop regime would have reduced the deepest decline significantly at the cost of cutting some winning positions early.
Open positions
No open positions at the time of this audit.
Honest summary
- Short-side edge is real. 90% win rate on 11 short episodes, $788k profit, and consistent execution across multiple positions. The averaging-down behaviour worked on shorts because the direction was correct and the account exited before the position became unmanageable.
- Long-side discipline is absent. 84.62% win rate masks the fact that long positions generated a net loss of $475k. The two largest losses are both long positions built through averaging down. The account appears to lack a clear exit rule for longs and instead compounds into underwater positions.
- Position sizing is the critical failure mode. The largest loss reached $15.5M notional; the second-largest loss reached $19.3M. These are 50–60x the account's starting capital estimate. Without a hard position-size limit or a sub-1% stop rule, the account is vulnerable to a single adverse move wiping out months of gains, which nearly happened on 24 December.
- Sample is constrained by the 10k-fill cap. Only the most recent fills are visible; earlier account history is not available. The patterns observed here are real within this window, but the account's longer-term behaviour and whether these patterns persist or have been corrected is unknown.