Statistical arbitrage ("stat-arb") trades the relationship between related assets rather than the direction of either one individually. A classic crypto example: if ETH and BTC historically move together, a stat-arb strategy tracks the ETH/BTC ratio, and when it stretches unusually far from its historical range, goes long the underperformer and short the outperformer, betting the ratio reverts, largely independent of whether the overall market goes up or down.
Because it's expressed as a spread (long one leg, short the other), a well-constructed stat-arb position has much lower exposure to broad market direction than either leg alone. The risk is specifically that the historical relationship breaks down (a "regime shift" in the pair itself), not that the market falls.
Like cross-sectional momentum, this is a roadmap category for zengtrade. The paired long/short mechanics need a different execution model than the single-asset long-only engine currently running.
Educational content, not investment advice. zengtrade is paper-first and non-custodial.