A market regime is a persistent state of the market that determines how prices behave and which strategies work. Trending or ranging. High volatility or low. Risk-on or risk-off. A strategy tuned for one regime typically loses in another.
The concept matters because most strategy failures aren’t the strategy being wrong. They’re the regime changing.
The simplest classification
Two axes: direction and volatility. Trending-calm, trending-volatile, ranging-calm, ranging-volatile. Mean reversion earns in ranging-calm and dies in trending-volatile. Trend following is the reverse.
Richer classifications add liquidity conditions, funding regimes, correlation structure and momentum of volatility itself. AlphaNet’s engine distinguishes 27 states.
Why classification is the hard part
Regimes are obvious in hindsight and ambiguous in real time. The transition from range to trend looks, for its first several bars, exactly like a range that’s about to revert. Every classifier lags, and the cost of the lag is the losses taken while the old regime’s strategy is still running in the new one.
What regime awareness changes
Instead of one strategy running always, a set of strategies with a model deciding which should be active, in what size, given the current state. The decision about what to do is made by something that has looked at the conditions — not a rule that fires regardless.
That’s the practical difference between automation and systematic trading, and it’s what the AI-trading-bot-versus-quant-system entry is about.