Quant trading strategies are systematic approaches to trading where every decision comes from a model or rule that’s been specified and tested in advance. They fall into a handful of families, each exploiting a different regularity in how markets behave.
The main families
Momentum and trend following: recent winners keep winning, at least for a while. Mean reversion: extremes revert. Statistical arbitrage: related assets that diverge reconverge. Market making: earn the spread for providing liquidity. Factor investing: characteristics like value or quality predict returns over long horizons. Event-driven: predictable price behaviour around scheduled events.
Most serious operations run several at once, because they earn in different regimes.
What every strategy needs
A reason the edge exists — a behavioural bias, a structural constraint, a risk premium someone else won’t bear. Without one, a pattern in the data is probably noise.
A test that could have failed. If the backtest was tuned until it worked, it’s fitted rather than validated.
A cost model. Fees, slippage and funding turn many profitable ideas into losing ones.
How strategies are combined
By regime. A model classifies the current market state and decides which strategies should be active and at what size. That layer — deciding which strategy, not just running one — is where most of the value in a modern quant operation sits.
How strategies die
Alpha decay. Every edge fades as it’s discovered and traded. A quant operation is a process for finding new edges faster than the old ones disappear.