Two Opposing Bets
Momentum says: what has been going up will continue going up. Mean reversion says: what has gone up too far will come back down. These are logically contradictory strategies. Both cannot be right at the same time on the same asset. Yet both have extensive academic support and long track records of profitability. The resolution to this apparent contradiction is that they work in different market regimes.
When Momentum Works
Momentum strategies outperform during trending market regimes. These are periods characterized by persistent directional moves, increasing participation (volume), and expanding volatility. Trending regimes are driven by gradual information diffusion (the market slowly pricing in a fundamental change), sustained capital flows (institutional rebalancing, ETF inflows), or reflexive feedback loops (rising prices attracting new buyers who push prices higher).
In crypto, momentum regimes are particularly powerful because of the narrative-driven nature of the market. When a new narrative takes hold (DeFi summer 2020, NFT mania 2021, AI tokens 2024), the momentum can persist for months as each new wave of attention brings fresh capital into the theme.
When Mean Reversion Works
Mean reversion strategies outperform during range-bound, choppy, or oscillating market regimes. These are periods where price moves in one direction are quickly reversed, volatility is stable or declining, and no strong directional trend is established. Range-bound regimes often occur after a significant trend has exhausted itself and the market needs time to consolidate and establish a new equilibrium.
In equities, mean reversion is well-documented at the individual stock level. Stocks that decline sharply on no fundamental news (perhaps due to forced selling from a fund liquidation) tend to recover within days or weeks. In prediction markets, contracts that spike sharply on low-volume trades often revert toward their prior level as the market corrects the temporary displacement.
Identifying the Current Regime
Several indicators help identify whether the market is in a momentum or mean-reversion regime. The ADX (Average Directional Index) measures trend strength. Readings above 25 suggest a trending regime favorable to momentum. Readings below 20 suggest a range-bound regime favorable to mean reversion.
Volatility regime is another signal. Expanding volatility (measured by ATR or Bollinger Band width) tends to favor momentum. Contracting volatility favors mean reversion until the eventual volatility breakout reestablishes a trend.
Autocorrelation of returns provides a direct statistical measure. Positive autocorrelation (today's return predicts tomorrow's direction) is a momentum regime. Negative autocorrelation (today's return predicts the opposite tomorrow) is a mean-reversion regime. Near-zero autocorrelation is a random regime where neither approach has a reliable edge.
The Adaptive Approach
The best systematic trading frameworks do not choose one approach permanently. They identify the current regime and deploy the appropriate strategy. In a strong uptrend, the momentum system is active while the mean-reversion system is sidelined. In a range-bound market, the mean-reversion system is active while the momentum system is quiet.
This adaptive approach is harder to backtest and implement than a single-strategy system, but it dramatically improves performance consistency across market cycles. You are no longer dependent on a specific market environment to make money. You have tools for multiple environments.
Explore these tools on Blockcircle: Prediction Markets Mispricing Engine | Momentum Trading Engine