Time-Series Momentum, Long Only
A built-in strategy on Central Command, tested on BloFin history. Thin edge in testing. 1D candles, long only, category Momentum.
Rules
Entry. At the close of each Monday and Thursday daily candle, go long if the close is higher than 56 days earlier. Long only.
Managing the trade. At each check, close a long whose 56-day return has turned negative. Otherwise hold.
Stop and target. Stop 4 x ATR(14) on the daily chart from the entry, as a disaster stop. No fixed target.
Exact rules
A hand-built variant of tsmom-weekly (not from the sibling rule: the parent's longs alone had too few trades in-sample). A stopped-out position can come back at the next check if the return still agrees.
[EXACT RULES]
timeframe: 1D
direction: long only
long entry: (DAY_OF_WEEK == 1 OR DAY_OF_WEEK == 4)
long entry: ROC(56) > 0
long exit: (DAY_OF_WEEK == 1 OR DAY_OF_WEEK == 4) AND ROC(56) < 0
stop: 4 ATR(14)
take profit: none
How it tested
Settings were chosen on 2023-10-01 to 2025-08-01, then judged once on 2025-08-01 to 2026-10-19, which was not looked at while choosing. Figures are after fees, on BTC, ETH, SOL, XRP, DOGE, BNB.
| Trades | Return | Profit factor | Worst drop | Win rate | |
|---|---|---|---|---|---|
| Chosen on | 43 | +40.6% | 4.04 | 10.9% | 41.9% |
| Judged on | 25 | +0.2% | 1.03 | 12.9% | 28.0% |
| Whole period | 68 | +40.9% | 2.70 | 12.9% | 36.8% |
Holding Bitcoin over the same judged period returned -29.0%.
A backtest is not a result. Past results, paper or live, do not guarantee future returns.
Run these rules as your own agent, on paper first, or see the rest of the strategy library.