Bottom line up front
This account is -99.92% in the analysed window, having lost $5.01M on a starting balance of approximately $5.01M. The wallet peaked at $5.17M on 9 July before declining to a trough of $1.33M on 10 September—a 74.3% drawdown—and has since eroded to $3,951. The headline pattern is unambiguous: the account exhibits severe position-sizing discipline failure, revenge trading after losses, and a complete absence of edge on the two coins that dominate the loss ledger (ETH and BTC). Five of the top six losses are oversized positions, and three of those were opened immediately after prior losses on the same instrument.
What the data shows
The account opened on 12 May 2025 and has executed 36 total episodes across eight coins, closing 29 and leaving 7 open. Realised PnL after fees is -$5.21M against gross closed trade volume of $111.96M. The account is not capital-constrained in the traditional sense—it has sufficient balance to trade—but it is discipline-constrained. Position sizing scales with conviction or desperation, not with risk.
ETH accounts for 12 closed episodes and -$3.96M of realised loss. BTC accounts for 7 episodes and -$1.04M. Together they represent 19 of 29 closed trades and 78% of total realised loss. The only coins with positive edges are HYPE (2 episodes, +$35k) and SOL (4 episodes, +$129). These are noise against the core bleed.
The long side has lost -$3.53M across 20 episodes at a 15% win rate. The short side has lost -$1.47M across 9 episodes at a 22.2% win rate. Neither direction works. The account has generated a 17.24% win rate overall, a 0.06 win/loss ratio, and a profit factor of 0.01—meaning for every dollar of gross profit, the account has lost $99. Expectancy is -$172,596 per episode. Fees paid total $32,161, a minor drag relative to the underlying losses, but the account was already deeply unprofitable before execution costs.
The largest single loss is an ETH short on 22–23 August, closed at $4,604.85, for -$931,454 on a $15.99M notional position. The second-largest is an ETH long opened on 17 June and closed on 21 June at $2,314.14 for -$845,040 on a $6.85M notional—this trade was flagged as both an oversized loser and a revenge trade, opened immediately after a -$675,911 loss on ETH on 5 June. The third-largest is that 5 June loss itself: -$675,911 on a $16.03M notional ETH long, opened and closed the same day. The pattern repeats: loss, then a larger position on the same coin within hours or days.
Trade quality
Win rate of 17.24% is well below breakeven. Profit factor of 0.01 means the account is generating $1 of gross profit for every $100 of gross loss. Win/loss ratio of 0.06 indicates that when the account wins, it wins small ($11,692 average), and when it loses, it loses large ($210,989 average). Expectancy of -$172,596 per episode is the core metric: on average, each trade destroys $172k of capital. This is not variance or bad luck. This is the expected value of the trading system.
Post-mortems
ETH long, 17–21 June, closed at $2,314.14, -$845,040 on $6.85M notional.
This trade was opened on 17 June at 20:13 UTC, 102 hours before closure. It was flagged as both an oversized loser and a revenge trade—the prior loss on ETH was -$675,911 on 5 June. The position size doubled down on the same instrument after a catastrophic loss. The trade bled $845k in just over four days. This is not a risk management failure; it is the absence of risk management.
BTC long, 3 August–10 October, closed at $114,720.82, -$461,752 on $5.66M notional.
This trade was opened on 3 August at 00:37 UTC, 1,651 hours (69 days) later closed on 10 October. It was flagged as both an oversized loser and a revenge trade, opened after a -$603,835 loss on ETH on 1–22 August. The account rotated from a failed ETH long into a BTC long at similar notional size. The position was held for nearly two months and still lost $461k. No structural stop was in place.
What the risk simulation reveals
The risk simulator applies historical 1%, 2%, and 4% stop-loss rules to the actual trade history. Under a 1% rule, the account would have realised -$193,839 with a 48.35% deepest decline in this window and a 10% win rate. Under a 2% rule, -$387,678 with a 96.69% deepest decline. Under a 4% rule, -$775,355 with a 193.39% deepest decline. Six episodes were stopped early under each rule, indicating that even mechanical stops would not have prevented the core pattern: the account enters positions too large relative to its conviction and risk tolerance, and when stopped, it re-enters at the same or larger size.
Open positions
No open positions are currently held. The wallet is flat.
Honest summary
- Visible strength: SOL and HYPE generated small positive edges. The account can identify profitable setups; the problem is not pattern recognition but position sizing and emotional discipline after losses.
- Visible weakness: Revenge trading after losses is the dominant behavioural signature. Five of the top six losses are flagged as oversized, and three of those were opened within hours of prior losses on the same coin. Position size is not calibrated to account equity or risk per trade; it scales with desperation.
- Visible weakness: The account has zero structural stops on the majority of losing trades. Even when stops are present (HYPE long, BTC long), they are not tight enough to prevent six-figure losses. The account is trading with conviction but without conviction-appropriate position sizing.
- Data scope: The account is 288 days old and has closed 29 episodes. The sample is sufficient to diagnose the problem: the account does not have an edge on its core instruments, and it compounds losses through revenge sizing. The issue is not sample size; it is behaviour.