Graveyard / Momentum / Binance perps
Cross-sectional crypto momentum
Does buying the strongest coins work?
The idea
Rank eight large coins by risk-adjusted return. Hold the leaders.
How we tested it
| Data | Eight Binance perpetuals (BTC, ETH, BNB, SOL, XRP, ADA, DOGE, LINK), ranked daily. 36 configurations of lookback, take profit and stop. |
|---|---|
| Period | Two years covering bull, bear and sideways markets. |
| Costs | 0.11% per round trip per leg. |
| Holdout | Results were split by regime instead of by time. |
| Control | No random-entry control was run for this test. |
Result
| Trades, best configuration | 1,868 |
|---|---|
| Return | +30.48% |
| Sharpe | 1.04 |
| Profit factor | 0.913 |
| Bull / bear / sideways | +33.58% / −1.54% / −4.96% |
Why it failed
It is a bet on a bull market in disguise. Trading eight legs adds fees that push the win rate (32.7%) below break-even (34.8%).
What would make it worth retesting
Only with a regime filter and fewer legs.
Test id: momentum-xsect-multi-event-v1. Past results of a backtest say nothing certain about the future.