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 -100% in the data covered, having liquidated from a highest balance in this window of $5.13M to a current balance of -$3.58M. The headline loss masks a catastrophic pattern: three oversized ETH long positions—each flagged as revenge trades or averaging-down episodes—consumed $3.36M of the $3.58M loss. The account opened with edge (a 2-hour ETH short at 4610 for +$190k), then abandoned discipline entirely. Structural stops were set at 3% but ignored; positions grew to $100M notional on single trades while the account held $5M in highest balance in this window balance.
What the data shows
The account was active for 11 days in the data covered, from 12 August to 24 August 2025. It opened with modest size and a clean short: ETH short on 12 August at 4610.29, closed at 4578.28 in 2 hours for +$190k. That trade had no behavioural flags and established a template for what worked. The account then reversed course entirely.
On 13 August, the account opened an ETH short at 4696.69 with 1,197 averaging-down events, reaching $110M notional, and closed 4 hours later at 4723.91 for -$370k. Minutes later, it opened an ETH long at 4731.32 as a revenge trade (following the short loss), added 438 times, reached $100M notional, and closed 19 hours later at 4705.89 for -$2.21M. This single trade consumed 62% of the total loss. A third ETH long opened on 14 August at 4566.43, closed in 31 minutes at 4474.02 for -$776k. A BTC long opened on 13 August at 122957.14 as a revenge trade following the second ETH loss, reached $61M notional with 438 averaging events, and closed 1 hour later at 122963.26 for -$269k.
The account made three wins totalling $237k and twelve losses totalling $3.66M. Fees paid were $173.7k gross. Long positions lost $3.41M on 0% win rate across five episodes; short positions lost $168k on 50% win rate across two episodes. The deepest decline in this window was 99.24%, from $5.13M to $38.8k, over seven days.
Trade quality
Win rate was 25% (3 wins in 12 closed trades). Profit factor was 0.06—for every dollar won, the account lost $16.67. Expectancy was -$298k per trade. Win/loss ratio was 0.19: average win was $79k, average loss was $423k. The max loss streak was 7 consecutive losses. Fees of $173.7k represented 5% of gross realised losses, a material but secondary drag; the core problem was position sizing and averaging into losers.
Post-mortems
ETH long, 13–14 August, opened 4731.32, closed 4705.89, -$2.21M. This was flagged as both oversized loser and revenge trade. It opened immediately after the ETH short loss of -$370k. The position reached $100.78M notional—20x the highest balance in this window account balance—with 438 averaging-down events. Structural stop was set at 3% but never executed. The trade lasted 19 hours and lost 0.55% on entry, a modest adverse move that should have triggered the stop. Instead, the account added repeatedly into a losing position until it closed manually at 4705.89. This single trade is the fulcrum of the account's collapse.
ETH short, 13 August, opened 4696.69, closed 4723.91, -$370k. Flagged as averaging down and oversized loser. Opened with 1,197 add events, reaching $110.3M notional. Closed in 4 hours 23 minutes at 4723.91, a 0.58% adverse move. The structural stop at 3% was not triggered. This loss directly preceded the revenge long that consumed $2.21M.
BTC long, 13 August, opened 122957.14, closed 122963.26, -$269k. Flagged as averaging down and revenge trade. Opened after the second ETH loss (the -$2.21M long). Position reached $61.48M notional with 438 averaging events. Closed in 1 hour 10 minutes at 122963.26, a +0.005% move—the position was in profit but closed at a loss, suggesting forced liquidation or margin call. This was a revenge trade chasing recovery after the catastrophic ETH long.
ETH long, 14 August, opened 4566.43, closed 4474.02, -$776k. Oversized loser. Opened at 4566.43, closed in 31 minutes at 4474.02, a 1.99% adverse move. Position reached $33.3M notional. This was a final capitulation trade after the account had already lost $2.85M in the prior 24 hours.
What the risk simulation reveals
Under a 1% stop-loss rule applied historically, the account would have realised -$99.6k with a deepest decline in this window of 2.76%, stopping out 3 trades early. Under 2%, the loss would have been -$199.2k with a 5.51% deepest decline. Under 4%, the loss would have been -$398.5k with an 11.02% deepest decline. In all three scenarios, the win rate remained 27.27% because the three winning trades were small and unaffected by the stops. The simulator is gross of fees. The core finding: disciplined stops would have reduced losses by 94–97% in this data covered. The account had stops defined but did not execute them.
Open positions
No open positions remain. The account is fully liquidated.
Honest summary
- One visible strength: The opening trade (ETH short, 4610–4578, +$190k in 2 hours) was clean, sized appropriately, and closed with profit. It demonstrates the account can identify direction and execute without emotional override.
- One visible weakness: Revenge trading and averaging into losers are structural, not occasional. Three of the four largest losses were flagged as revenge trades or averaging-down episodes. The account opened positions at $100M notional on a $5M account and added 400+ times into each loser. Structural stops existed but were ignored.
- One visible weakness: Long-side bias was catastrophic. All five long trades lost; zero won. Short trades won 50% of the time. The account's directional conviction was inverted.
- Data scope caveat: Only the most recent 10,000 fills are visible. Earlier account history is not available. The 11-day data covered may not reflect longer-term patterns or account genesis.