Alpha decay is the erosion of a strategy’s excess return over time. An edge that earned 2% a month last year earns 1% this year and nothing next year. The signal that predicted price no longer does.
Every edge decays. The questions are how fast, and whether you notice before the market does.
Why it happens
Crowding. Other participants find the same inefficiency and trade it, and the mispricing closes. Regime change: the market condition that made the signal work stops occurring. Adaptation: the behaviour the strategy exploited — a predictable pattern in how others trade — changes because those others got smarter.
Sometimes the edge was never real. It was noise that looked like signal in the backtest, and live trading is the decay of an illusion.
How it’s measured
Rolling performance windows. If a strategy’s 90-day Sharpe has trended down for three consecutive quarters, that’s decay, not variance. Signal correlation — how well the signal predicted returns — tracked over time shows the same thing more directly.
Why it matters for capacity
The more capital a strategy runs, the more its own trades move the market it’s trying to predict, and the faster its edge decays. Every strategy has a capacity beyond which it destroys itself. Strategies that publish capacity limits are acknowledging this; ones that don’t are pretending it doesn’t apply to them.
What a systematic process does about it
Monitors continuously, retires strategies whose decay is confirmed, and keeps discovering new ones. A single strategy is a depleting asset. A research process that produces strategies is not.