Original Research Lead hypothesis (not seeded). It is a breadth replication of a frozen rule, not a new parameter search. Lesson L262 says the absorbed-flush rule transferred best of any frozen rule: it was promoted on 13 Bybit perps (run 958bf541: 200 trades, PF 1.72, Sharpe 0.82, DD 10.2%) and again on 15 Hyperliquid perps with no shared names (151 trades, PF 1.70, Sharpe 0.73). It ended underpowered on a 24-name Binance SPOT panel (PF 1.44, Sharpe 0.48, DD 22.7%). The analyst there read that as a sign the effect relies on leveraged-perp liquidation flow. This test goes back to perps, keeps the rule frozen, and uses 18 Bybit USDT perps that share NO names with any of the three earlier panels: SOL, XRP, ADA, LINK, BNB, AXS, GALA, ICP, IMX, KSM, DASH, ZEC, ENS, MASK, AR, LRC, FLOW, CRO. Excluded because they were used before: DOT LTC BCH AVAX ATOM NEAR UNI FIL ETC DOGE TRX XLM AAVE (Bybit panel); ARB OP SUI LDO CRV SNX RUNE DYDX GMX STX WLD BLUR PEPE SEI TIA (HL panel); and the 24 Binance spot names. Bybit is under quota (3.4% vs a >=5% target). Single venue, but a different universe. MECHANISM: on a leveraged-perp alt, a day whose range blows out to >=2x normal volatility while the close still finishes in the upper 40% of that range shows forced selling (long liquidations, stop runs) that was absorbed intraday by resting demand. The forced flow is over, and the price that cleared it sits below fair value. Over the next few days it recovers as liquidity providers who absorbed the flush are paid out. PRE-STUDY (frozen rule, gross close-to-close 3-day forward return, no costs, no concurrency cap; measured 2026-10-31 on local Bybit daily bars from 2020-10 / 2021 to 2026-10-30): 297 qualifying name-events on only 121 DISTINCT event dates (events cluster on market-wide flush days). Mean +1.72% per event, median +0.37%, win rate 52%. By year: 2021 -2.2% (15 events), 2022 +0.4% (51), 2023 +1.6% (48), 2024 +3.8% (106), 2025 +2.2% (48), 2026 -2.0% (28). The edge is right-skewed (mean far above median) and not positive every year. That is pre-registered here, not hidden. EXECUTION: Bybit is a 24/7 venue, so 'enter at the next bar's open' is written as its faithful equivalent, a market order at the signal bar's close (the next daily bar's open). Exit is a market order at the close of the 3rd daily bar after entry. No stop orders are required. SURVIVORSHIP: Bybit serves no bars for closed contracts, so the panel is today's survivors that were listed by 2022-01. Each leg is eligible only from its own listing plus 60 bars of warm-up (point_in_time_universe=true, universe_at()). Survivorship biases a long-only rebound upward, so the analyst should weigh that against the result.