A long-short mean-reversion strategy on BTCUSDT.BINANCE USD-M perpetual futures that trades counter to forced-liquidation cascades. Uses BINANCE USD-M perp bars (proven clean data) + liquidations supplementary data (PostgreSQL — distinct stream from any prior pipeline experiment). The hypothesis: when hourly liquidation volume on BTCUSDT spikes >3 standard deviations above its 30-day baseline AND price has moved ≥2% in the same direction over the prior 4 hours, the market is in a FORCED-FLOW OVERSHOOT — the marginal seller (in a long-liquidation cascade) or marginal buyer (in a short-squeeze) is mechanical, not natural. Once forced flow exhausts (typically within 4-12 hours), natural participants step in at the dislocated price and the move partially reverses. The strategy goes LONG when long-side liquidations cascade (price overshoots DOWN) and SHORT when short-side liquidations cascade (price overshoots UP). This fills FIVE under-represented buckets: long-short direction (14.2% → toward 45%), liquidations-driven mechanism (zero strategies in 459 experiments have used this data stream), crypto-specific microstructure (a TradFi-impossible edge), mid-term hold (4-12h cycles), and supplementary-data-driven (which has been clean in our catalog unlike BINANCE_SPOT bars). The mechanism is documented in crypto microstructure research (Kaiko 2022, Coin Metrics 2023) — liquidation cascades systematically produce 1.5-3% price overshoots that partially revert within 6-18 hours as the forced-flow component of order flow dissipates. Uses ONLY BINANCE USD-M perp bars + Postgres-stored liquidations data — no BINANCE_SPOT bars (which have repeatedly failed Layer 3 verification in this session). Position sizing: 15% of equity per direction. Risk per trade: 1.5% of equity capped by stop-loss.