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 -33.9% in the data covered, down $81.7k on a starting balance of $241k. The headline loss masks a structural problem: a 67.8% win rate on 90 closed episodes was obliterated by five oversized losses that each exceeded the median loss by 8× to 58×. Revenge trades, FOMO re-entries, and averaging down into losing positions consumed the edge. The deepest decline in this window reached -53.9%, and the account sits at $159k. Position sizing exploded after losses—the largest losing trade carried $4.1m notional on a $241k account—and stops were consistently ignored or set too wide to matter.
What the data shows
The account opened on 16 April 2026 and closed 90 episodes over four days, trading only ETH. The initial balance was $241k; the highest balance in this window reached $287.7k before a catastrophic deepest decline in this window sequence. Long trades lost $43.9k; short trades lost $37.8k. Neither direction worked, but the damage was concentrated: five trades accounted for $155.3k in losses, while the remaining 85 trades netted $73.6k. This is not variance. This is a sizing and discipline failure.
The win rate of 67.8% is a mirage. Sixty-one of 90 trades were winners, but the average winner was $2,421 and the average loser was $7,910. The profit factor of 0.64 means every dollar won generated $1.56 in losses. The expectancy was -$908 per trade. Fees added $5.5k in net drag, but they are noise compared to the core problem: position sizing scaled into losses instead of away from them.
Averaging down occurred across five distinct ETH long episodes, with the largest opening on 17 April at $2,327.90 and scaling to 839 contracts. That trade closed at $2,353.65 for a $29.5k win—the largest winner in the window. The same pattern then reversed: on 18–19 April, two long positions were averaged down into losses of $32.5k and $20.4k respectively. The account did not learn from the win. It applied the same mechanic to the next loss and got crushed.
Revenge trades are visible in five instances, all following losses on the same day. The largest revenge trade opened on 19 April with a notional of $4.08m—17× the starting balance—after a $55.9k loss on a short position. This trade closed 1.83 hours later for a $40.6k loss. The account was chasing the loss with maximum leverage and no structural discipline.
Trade quality
Win rate: 67.78%. Profit factor: 0.64. Expectancy: -$907.95 per trade. Win/loss ratio: 0.31.
A 67.8% win rate is excellent. It is also irrelevant. The account won two-thirds of its trades and lost one-third of its capital. The profit factor of 0.64 is the core metric: for every $1 of gross profit, the account generated $1.56 in gross losses. Expectancy of -$908 per trade means the account was underwater on average before fees. Fees of $5.5k net drag were a secondary problem; the primary problem was that winners were too small and losers were too large.
The max win streak was 10 consecutive winners. The max loss streak was 4. Neither streak prevented the account from sizing into the next loss as if the previous loss had not occurred.
Post-mortems
Trade 1: ETH long, 18–19 April, entry $2,325.51, exit $2,317.75, -$32.5k loss
This trade was flagged as averaging down and an oversized loser. It opened on 18 April and closed 8.53 hours later. The position reached $2.79m notional—11.6× the starting balance. The account added to a losing long position at $2,341.5, $2,340.4, and $2,340.9 across five separate fills. The structural stop was set at 0.88% below entry, far too tight to survive normal volatility. The account closed the position at a $32.5k loss, 30.76× the median loss size.
Trade 2: ETH long, 20 April, entry $2,306.85, exit $2,295.72, -$40.6k loss
This trade was flagged as a FOMO re-entry, oversized loser, and revenge trade. It opened on 20 April at 13:17 UTC, 1.83 hours after a $20.4k loss closed on the same coin. The position reached $4.08m notional—16.9× the starting balance. The structural stop was 1.29% away, again too wide to enforce discipline. The account closed the position 1.83 hours later for a $40.6k loss, the second-largest loss in the window. This was a direct revenge trade: the previous loss was on ETH, the re-entry was on ETH, and the sizing was maximum.
What the risk simulation reveals
Under a 1% stop-loss rule, the account would have closed at +$5.5k with a maximum decline of -10.2%. Under a 2% rule, +$11.0k with -19.1% decline. Under a 4% rule, +$21.9k with -34.1% decline. The actual result was -$81.7k with a -53.9% decline.
The simulation stopped early on three episodes due to data quality issues, but the pattern is clear: mechanical stops at any reasonable distance would have turned this account profitable. The account's actual structural stops were set at 0.88% to 1.64% away from entry—tighter than the 1% rule—but were not enforced. The account overrode stops or ignored them entirely when positions moved against it.
Open positions
No open positions at the time of this audit.
Honest summary
- Visible strength: Win rate of 67.8% and a 10-trade win streak demonstrate the account can identify profitable setups. The largest winner ($29.5k on 17 April) was executed with averaging down and closed at the right time.
- Visible weakness: Position sizing is the critical failure. The account scaled notional to 17× starting balance after losses and ignored or overrode structural stops. The five largest losses account for 190% of total realised losses; the remaining 85 trades were collectively profitable.
- Visible weakness: Revenge trades and FOMO re-entries are documented in five instances. The account re-entered ETH within minutes of closing a loss on the same coin, at maximum leverage, with no reset period.
- Data scope: Only the most recent 10,000 fills are visible. This audit covers four days of trading. The patterns observed—averaging down, revenge sizing, stop override—are acute and repeating, but the data covered is short.