Mean Reversion Strategies

Unless you trade spreads, these strategies work poorly for crypto.

Mean reversion strategies

Economic Intuition: After extreme moves, crypto tends to partially retrace due to profit-taking and positioning adjustments.

Why mean reversion works

• Retail traders take profits after big moves

• Institutional rebalancing (selling winners, buying losers)

• Funding rate pressures (high funding leads to longs exit which in turn leads to price drops)

• Psychological anchoring to recent price ranges

Signal calculation

$$\text{MeanRev}_N(t) = \frac{\text{EMA}_N(P) - P_t}{\sigma_t}$$

• N: Long lookback

• $\sigma_t$ is realized volatility at time t

Intuition tells us, that we need to use longer lookbacks to avoids conflict with momentum strategies (which use 10,30,60 day lookbacks). Using ETH/USDT price history from 2017 to 2025 we have found that the half-life of mean reversion for ETH/USDT daily price series is around 487 days and is impractical for our portfolio management, so we won’t be using signals from this strategy type in our framework.