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 lost $3.58M in the data covered, a -100% realisation on an estimated $3.58M starting balance. The highest balance in this window was $5.13M on 13 August; the lowest balance in this window was $38.8K on 20 August. The deepest decline in this window was -99.24%. The core failure is structural: the account opened with a profitable short ETH trade on 12 August (+$190K), then immediately reversed into oversized long positions, averaged down aggressively across both ETH and BTC, and closed out three catastrophic losses totalling $2.36M in a 36-hour window. Revenge trading and position sizing without stops destroyed an account that briefly held $5M in equity.
What the data shows
The account opened on 12 August with a clean short ETH trade: entry 4618.65, exit 4579.53 in 2 hours, +$190K on a $23M notional position. This was the only trade executed with discipline. Immediately after, the account reversed into a long ETH position at 4589.35 on 12 August, which was averaged down 20 times to a maximum size of 10,000 ETH, closing at 4569.5 for -$105K. This was flagged as averaging down but the loss was contained.
The critical breakdown occurred on 13 August. After the short ETH loss of -$370K (entry 4694.55, exit 4724.47 in 4.4 hours on a $110M notional position), the account opened a revenge trade: a long ETH position at 4727.0 on 13 August at 19:15 UTC with a $100.78M notional size. This position was held for 19.3 hours and closed at 4712.98 on 14 August for -$2.21M. This was the single largest loss and is flagged as both an oversized loser (21× the median loss) and a revenge trade.
Simultaneously, after the ETH short loss, the account opened a long BTC position at 122,925.89 on 13 August at 21:31 UTC with a $61.5M notional size, averaged down 438 times, and closed at 122,980.59 after 1.17 hours for -$269K. This was flagged as both averaging down and a revenge trade.
On 14 August, a third oversized loss materialised: a long ETH position closed at 4506.25 for -$776K on a $33.3M notional, the second-largest loss and 7.39× the median.
By 20 August, the account had declined from $5.13M to $38.8K. The remaining activity (14 August to 24 August) shows two small short BTC wins (+$46.5K and +$404) but these are noise against the structural damage.
Fees paid were $173.7K on $769.9M gross volume. The net fee drag of $173.7K is material but secondary to the position-sizing and revenge-trading catastrophe. Realised PnL was -$3.42M; fees consumed an additional $173.7K, for a total realised loss of -$3.58M.
Trade quality
Win rate was 25% (3 wins from 12 closed episodes). Profit factor was 0.06, meaning losses were 16.7× larger than wins. Expectancy was -$298K per trade. The win/loss ratio was 0.19: average loss was -$423.8K, average win was +$79K. These numbers describe an account with no edge, no discipline, and no risk management.
The longest losing streak was 7 consecutive closed trades. The longest winning streak was 1 trade. There were no consecutive wins.
Post-mortems
ETH long, 13–14 August, revenge trade, -$2.21M:
Opened at 4727.0 on 13 August at 19:15 UTC following a -$370K short loss. Maximum notional exposure was $100.78M. Closed at 4712.98 on 14 August after 19.3 hours for -$2.21M. This position was sized at 21× the median loss and was explicitly a revenge trade. The structural stop was set at 3% distance but was not respected; the position was held through a 0.3% adverse move and closed at a loss. This is the defining trade of the account's collapse.
ETH short, 13 August, averaging down, -$370K:
Opened at 4694.55 on 13 August at 12:51 UTC. Averaged down 1,197 times to a maximum size of 23,500 ETH. Closed at 4724.47 after 4.4 hours for -$370K on a $110.3M notional. The position moved 30 basis points against the entry and was closed without recovery. This loss triggered the revenge trade that followed 4.5 hours later.
BTC long, 13 August, averaging down and revenge trade, -$269K:
Opened at 122,925.89 on 13 August at 21:31 UTC following the -$2.21M ETH loss. Averaged down 438 times to a maximum size of 500 BTC. Closed at 122,980.59 after 1.17 hours for -$269K on a $61.5M notional. This was a revenge trade opened after the largest loss of the window. The position was closed at a 55-basis-point loss.
What the risk simulation reveals
Under a 1% stop-loss rule applied historically, the account would have realised -$951.60 with a maximum decline of -0.04%. Under a 2% rule, the loss would have been -$1,903.21 with a maximum decline of -0.09%. Under a 4% rule, the loss would have been -$3,806.41 with a maximum decline of -0.18%. All three simulations show 40% win rate and 1 episode stopped early. These counterfactuals demonstrate that even a basic 1% hard stop would have reduced losses by 99.97% in the data covered. The account's actual losses were 940× larger than what a 1% stop would have permitted.
Open positions
No open positions at the time of the latest fill.
Honest summary
- One visible strength: The opening short ETH trade on 12 August was clean—entry, exit, profit, no averaging, no revenge. The account demonstrated it could execute a disciplined trade.
- One visible weakness: Position sizing was completely uncontrolled. The three largest losses were 21×, 7.39×, and 3.53× the median loss. The account opened positions with $100M+ notional on a $3.58M starting balance, then averaged down into them. No position had a working stop.
- One visible weakness: Revenge trading was systematic. After the -$370K short loss, the account immediately opened a $100.78M long position. After the -$2.21M long loss, it opened a $61.5M long BTC position. Both were closed at losses within hours.
- Data scope caveat: Only the most recent 10,000 fills are visible. Earlier account history is not available. The data covered spans 11 days and 12 closed episodes; this is sufficient to identify structural failures but not sufficient to assess whether these patterns are account-wide or window-specific.