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 -0.01% in the data covered, down $524.76 on a $4.98M balance, but the headline obscures the real problem: ETH shorts consumed $721.84 in realised losses while BTC and SOL longs generated modest edges. The deepest decline in this window was -0.02%. Behavioural flags dominate the loss trades—averaging down, revenge trading, and FOMO re-entries appear across the five largest losers, while the account ran 1,060 closed episodes in under 48 hours with a 67.45% win rate that masks a 0.8 profit factor and negative expectancy of -$0.50 per trade.
What the data shows
This is a high-frequency scalp account operating on 9–20x leverage across BTC, ETH, and SOL. The data covered spans 35 hours (18–20 May 2026) and captures 1,060 closed trades plus 3 open positions. Realised PnL is -$1,015.55; net fee drag is -$489.36, meaning fees consumed 48% of the gross loss. The account paid $54.10 in explicit fees against $18.4M in gross volume, achieving 99.54% maker fill rate—execution is clean, but volume and leverage are not protecting against directional bias.
By instrument, the split is stark. BTC generated +$96.30 across 474 episodes (71.73% win rate, long-biased at +$87.79 vs short at +$8.51). SOL added +$100.78 across 190 episodes (65.26% win rate, balanced). ETH lost -$721.84 across 396 episodes (63.38% win rate)—a win rate above 60% that still produced a $721 loss signals severe position sizing or averaging-down behaviour on losers. The long/short split confirms this: longs lost -$115.70 (66.98% win rate), shorts lost -$409.05 (67.92% win rate), meaning short-side attempts were the primary leak.
The five largest losses are all ETH or BTC shorts opened on 18–19 May. The single worst trade was an ETH short entered at 2082.90 on 18 May, exited at 2098.52 after 23 minutes, for -$196.12 on a $37.2k notional position. The second-worst was the same coin, same day, entered at 2118.48, exited at 2129.83 after 32 minutes, for -$140.33 on $29.7k notional—this trade carried three flags: averaging down, oversized loser (37.78x the median loss), and revenge trade following a SOL short loss of -$15.95. The pattern repeats: small losses trigger re-entries into larger positions in different coins, which then fail.
Trade quality
Win rate of 67.45% is above market noise, but profit factor of 0.8 and expectancy of -$0.50 per trade reveal the real picture: the account wins more often than it loses, but when it loses, it loses larger. Average win is $2.86; average loss is -$7.45. The win/loss ratio of 0.38 means the account is taking 2.6x larger losses than wins on average. Fees are not the primary driver—net fee drag of -$489.36 is material but secondary to the -$1,015.55 realised loss. The account is losing on edge, not on execution.
Post-mortems
ETH short, 18 May 14:29–14:32 UTC, entry 2118.48, exit 2129.83, -$140.33 on $29.7k notional.
This trade carried three behavioural flags: averaging down (added to the position), oversized loser (37.78x median loss), and revenge trade (opened after a -$15.95 SOL short loss). The account held the position for 32 minutes, allowing a 0.53% adverse move to crystallise into a $140 loss. The structural stop distance was 3.0%, meaning the account was risking $891 to make a small profit—a 6:1 risk/reward on a coin that was moving against the short thesis.
BTC long, 19 May 14:54–15:05 UTC, entry 76901.92, exit 76557.20, -$69.83 on $31.5k notional.
Opened after a -$0.026 BTC loss (revenge trade), this long carried averaging-down and FOMO re-entry flags. The account entered at 76901.92, held for 20 minutes, and exited at 76557.20 after a 0.45% move. The position was sized at $31.5k notional on 9x leverage, risking $945 to capture a small scalp. The trade closed at a loss, adding to the sequence of oversized bets following micro-losses.
What the risk simulator reveals
Under a 1% hard stop rule applied historically to this window, the account would have realised -$87,014.16 in PnL with a -2.7% deepest decline, win rate collapsing to 56.52%. Under 2%, the loss would be -$174,028.32 with -5.37% deepest decline. Under 4%, the loss would be -$348,056.65 with -10.62% deepest decline. These are gross-of-fees figures. The simulator shows that without structural stops, the account's high win rate is masking catastrophic tail risk: a 1% stop would have prevented the largest losses but also forced the account to exit winning trades early, net-negative in this window. The account is currently running no stops on any of the three open positions.
Open positions
Three open longs are live: BTC at 76805.90 on 9x leverage with +$0.0543 unrealised; ETH at 2110.46 on 9x leverage with +$0.3824 unrealised; SOL at 84.2563 on 20x leverage with +$0.0616 unrealised. All three are micro-profitable and carry no stops. The SOL position is the most leveraged at 20x; the BTC and ETH positions are at 9x. None of these positions are material to the account balance, but the absence of stops across all three is consistent with the behavioural pattern observed in closed trades.
Honest summary
- Strength: BTC and SOL edges are real. BTC generated +$96.30 on 474 trades with a 71.73% win rate; SOL added +$100.78 on 190 trades with 65.26% win rate. The account can identify directional bias in these two coins.
- Weakness: ETH is a consistent loss generator (-$721.84 across 396 trades), and the account does not exit the position class. Worse, losses in one coin trigger revenge trades in another, escalating position size after losses rather than reducing it. The five largest losses all carry averaging-down or revenge-trade flags.
- Weakness: Position sizing is inverted to risk. The largest losses occur on the largest notional positions ($29.7k–$37.2k), while wins cluster around $2–$45. The account is sizing up into losers and sizing down into winners—the opposite of sound risk management.
- Data scope: The data covered is 35 hours and 1,060 trades. This is a