{"market":"ETH","direction":"short","leverage":750,"take_profit_bp":16,"take_profit_price":2379.24,"bust_distance_bp":8.53,"bust_price_estimate":2385.08,"expected_net_return_on_margin_pct":89.1,"confidence":0.4,"regime":"trending_down","key_levels":{"support":2378.5,"resistance":2386.8},"entry_mid":2383.05,"data_time_utc":"2026-10-22T20:24:36.150Z","reasoning":"Facts (Hyperliquid info API, 20:24 UTC): ETH is down 437 bp on the day versus BTC's 372 bp, trails BTC by 11 bp over 60m and 18 bp over 240m, has 15m EMA9 12.5 bp under EMA21 with the EMA21 falling, trades 16 bp under its 1h VWAP with mark 3.8 bp under oracle, and its last 10 prints were 68% sell-side by notional; BTC is sitting 2 bp above its 1h low into a bid-heavy book, so shorting BTC there has less room than shorting ETH, which still has 19 bp down to its 1h low of 2378.5. Inference: the trend continues after a small 5m bounce, so short ETH. Leverage is at the 750x floor because in an 8h 1m-candle first-passage backtest the win and bust odds both get worse as leverage rises; there, ETH short at 750x hit 15 bp before busting 43% of the time and 20 bp 36% of the time. The 16 bp target sits just above the 1h low, clears most of the haircut (keeps 74%) and pays about 89% on margin, while the historical EV was flat (about -18% to -20%) across 10-20 bp targets.","risks":"ETH's 1m realized volatility is 4.4-4.9 bp per minute, so the 8.5 bp bust distance (2385.08) is less than two typical one-minute moves, and a continued 5m mean-reversion bounce toward the 15m high of 2386.8 or the 1h VWAP of 2386.5 would liquidate the position. Treat this as a negative-expected-value lottery: every leverage and target combination tested had negative historical EV, and the trade only pays if the downtrend has real directional edge."}