H-004: Best daily trend rule, long/short futures, beats buy-and-hold

Weak Family: trend · Status: weak

Do long/short trend rules on crypto futures beat holding? The idea did a bit better than simply buying and holding, but not by enough to rule out luck. Treat it as unproven.

Result

significant after Bonferroni (p < 0.0167) on 0/3 assets (need 2); beats buy-and-hold on 2/3

Sharpe: return per unit of risk (higher is better). CAGR: average yearly growth. Max drawdown: worst fall from a peak. p-value: chance the result is luck (lower is better). “Hold” columns are buy-and-hold for the same asset and days. Glossary

Asset	Period	Rule (chosen in-sample)	Sharpe	Hold Sharpe	CAGR	Hold CAGR	Max drawdown	Hold drawdown	In market	p-value	Beats hold	Counts as win
BTCUSDT	oos (2022-01-01 → 2026-10-13)	SMA50	0.56	0.50	+16.7%	+13.1%	-62.6%	-66.9%	100%	0.080	yes	no
ETHUSDT	oos (2022-01-01 → 2026-10-13)	DC20	-0.00	0.22	-21.3%	-7.8%	-87.1%	-74.0%	100%	0.399	no	no
SOLUSDT	oos (2022-01-01 → 2026-10-13)	TSMOM180	0.38	0.36	-7.8%	-9.2%	-86.2%	-94.6%	100%	0.180	yes	no

Equity curves

Growth of 1 unit, log scale, fees included. Only days on which the rule is fully defined are counted.

Asset and period:

Why it might work

Same as H-003 with native shorts on USD-M futures (taker 0.05%).

Verdict history

2026-09-23: Fail (F1MEBOT (2026))
2026-10-14: Weak (this engine, data to 2026-10-13)

Pre-registered specification

kind: rule_search
candidates:
  SMA20:
    type: price_above_sma
    window: 20
  SMA50:
    type: price_above_sma
    window: 50
  SMA100:
    type: price_above_sma
    window: 100
  SMA200:
    type: price_above_sma
    window: 200
  SMA20/50:
    type: sma_cross
    fast: 20
    slow: 50
  SMA50/200:
    type: sma_cross
    fast: 50
    slow: 200
  TSMOM30:
    type: tsmom
    lookback: 30
  TSMOM90:
    type: tsmom
    lookback: 90
  TSMOM180:
    type: tsmom
    lookback: 180
  DC20:
    type: donchian
    entry: 20
  DC55:
    type: donchian
    entry: 55
mode: long_short
fee: 0.0005
universe:
- BTCUSDT
- ETHUSDT
- SOLUSDT
in_sample:
  start: '2018-01-01'
  end: '2021-12-31'
out_of_sample:
  start: '2022-01-01'
significance:
  method: bootstrap
  n: 400
  alpha_family: 0.05
criterion:
  min_strict_wins: 2