Crypto Trend Forward Test day 0

Live forward test of the eval-verified Donchian long-only trend strategies · inception 2026-10-25 · data: Binance daily klines (public REST)

loading prices...

BTC — Donchian 200d long-only era-stable pick

—

ETH — Donchian 20d long-only highest backtest SR, decaying edge

—

Forward-test equity (start = 1.0 at inception) — strategy vs buy & hold

Fees 5 bps taker + sqrt-impact slippage (floor 2 bps) + 11%/yr funding drag on held longs, applied daily, same as the backtest harness.

Daily log (inception → now)

date	BTC close	BTC pos	ETH close	ETH pos

Method: position decided at each daily close using data through that close (20d/200d close-channel breakout, long-only, held until opposite channel break) — identical to the audited harness in Avi-L/crypto-quant-eval. The forward test is stateless: every load recomputes positions from live Binance klines since inception, so the record is exactly reproducible. Backtest evidence: BTC Donchian 200d long-only OOS SR ~0.9, era-stable; ETH Donchian 20d full-sample SR 1.56 with a decaying recent edge (0.5). Max drawdowns of −60/−80% are inherent to these strategies. This is research output, not trading advice.