Kalman Velocity Trend
A two-state Kalman filter estimates price and its velocity. The strategy goes long when the filtered velocity rises above a minimum speed (measured in ATR per bar) with price above the filtered line and above a slow trend EMA, and short on the mirror. Exits: opposite velocity turn, or an ATR trailing stop that only tightens. Noise is measured on a rolling window, so the same settings work on any s
Member rating—no ratings yet
Used by0members
Best forward test—none running yet
Shared settings0by members, by market
How to use it
4h BTC / ETH / SOL perpetuals. Lower "Responsiveness" = fewer, longer trades. Position size is a % of equity (Properties). Commission 0.06%. Back-test BTCUSDT 4h (Bybit, 0.06% fee/side, funding not included): 2021–2025 +164% (PF 1.49, max DD 40%, 105 trades); unseen 2025–2026 +20% (PF 1.28, 46 trades). Past results don't predict future ones. Concept: R. E. Kalman (1960), constant-velocity tracking filter; ATR trailing stop (J. W. Wilder, 1978).
Try Kalman Velocity Trend
Open in the terminalGet accessAdd it to your chart, back-test it on any pair, compare it with other strategies, forward-test it on paper, or run it as a bot.
Back-tests and forward tests are simulations; past results do not predict future ones. Nothing here is financial advice. Trading carries a high risk of loss.