Market Regime
Classification
Before asking which direction the market is moving, you need to know what kind of market you are in. Regime classification answers that question — and it changes which tools you should trust.
What is a market regime?
A market regime is a persistent statistical state of price behaviour. Most classification frameworks distinguish two primary regimes: trending and ranging. In a trending regime, price moves directionally with positive serial autocorrelation — today's move is correlated with tomorrow's. In a ranging regime, price oscillates around a mean, previous moves predict reversals, and support/resistance levels have explanatory power that they lack in trends.
The key word is persistent. Regimes are not random flip-flops; they tend to last from days to weeks before transitioning. Detecting which regime is active is useful precisely because the same market data looks very different depending on which lens you apply.
The two primary regimes and what each means for options
Directional / Momentum
Price makes sustained highs or lows. Momentum signals (MACD, VWAP slope, velocity) are informative. Options strategies that benefit from sustained moves — directional spreads — have structural tailwinds. OI walls often get breached rather than defended.
The math: three tools that classify regimes
1. The Hurst Exponent
The Hurst exponent (H) is a single dimensionless number derived from the rescaled-range (R/S) analysis of a price series. Given a window of N price changes, the R/S statistic is the ratio of the cumulative range to the standard deviation of those changes. Hurst showed that this ratio scales as N^H:
- H > 0.5: the series is persistent — past up-moves predict future up-moves. Trending market.
- H < 0.5: the series is anti-persistent — past up-moves predict reversals. Mean-reverting market.
- H ≈ 0.5: the series is a geometric random walk with no exploitable structure.
Practitioners roll this over a 20–60 day window, giving a continuously updating reading of the current regime rather than a one-time classification.
2. Hidden Markov Models (HMM)
A Hidden Markov Model treats the regime as a hidden state that cannot be observed directly, but can be inferred from observable data — returns, volatility, volume. The model specifies a small number of latent states (typically two or three) and learns, from historical data, how likely each state is to transition to another and what the statistical distribution of returns looks like in each state.
Once trained, the Viterbi algorithm or forward-backward recursion assigns a probability to each regime at each point in time. Crucially, HMMs update continuously: as new data arrives, the posterior probability shifts. This makes them well-suited to intraday use where regime transitions during a session matter for real-time reads of NIFTY or Bank Nifty option chains.
3. Volatility-Based Thresholds
A simpler, widely-used heuristic: compare realised volatility (or implied volatility via India VIX) against its own rolling median. Elevated, rising volatility often accompanies trending or high-uncertainty regimes. Compressed, stable volatility fits ranging behaviour. While less mathematically rigorous than H or HMM, volatility-based filters are fast to compute intraday and serve as a useful first-pass screen.
Why Indian index options need regime awareness
NIFTY and Bank Nifty have structural features that make regime detection particularly relevant:
- Weekly expiry cycle: as Thursday approaches, option sellers defend strikes and max pain exerts magnetic pull on price — a ranging-regime dynamic that disappears in the days after a fresh expiry when a new weekly series opens with thin OI.
- India VIX regime: when India VIX is below its 30-day median, historical data shows NIFTY tends to respect OI wall levels more reliably. When VIX is elevated, these walls are breached more frequently — a signal that the market has entered a volatility-trending regime.
- FII positioning: large FII index derivative flows can shift regime quickly. A transition from net long to net short futures often precedes a regime break that momentum signals then confirm.
How TradePulse applies regime classification
TradePulse's market commentary engine does not score indicators with fixed weights. Instead, it first classifies the current regime using a combination of rolling Hurst estimation, volatility percentile, and OI-wall integrity checks. The regime label then determines which signal families carry interpretive weight in the commentary output.
In a detected ranging regime, the platform emphasises max pain, PCR trend, and net OI change at strikes nearest spot. In a trending regime, it shifts weight toward intraday price velocity, VWAP relationship and FII futures positioning. The commentary also states the transition trigger — the specific market condition that would constitute evidence the regime has changed, making the read actionable rather than just descriptive.
How to read a regime label as a trader
When you see TradePulse flag the current state as Trending, it means the statistical evidence favours persistent momentum. The actionable implication is to give more credence to breakouts of OI walls and less credence to mean-reversion signals like max pain. Conversely, a Ranging label means OI walls, max pain and PCR carry stronger predictive weight, and breakout signals should be treated with more scepticism until confirmed by volume and IV.
A third label — Transitional — appears when regime indicators disagree. This is the highest-uncertainty state, and it is the most honest answer the system can give: conflicting evidence means position sizing should reflect that uncertainty rather than picking a side.
Regime labels are diagnostic context, not buy or sell signals. This is educational material — not investment advice. All trading involves risk.
Frequently asked questions
What is a market regime in trading?
A market regime is a persistent statistical state of price behaviour — most commonly trending (prices move directionally with momentum) or ranging (prices oscillate around a mean with no sustained direction). Regime classification is the practice of identifying which state the market is currently in so that the right analytical tools and strategies are applied.
How is the Hurst exponent used in regime detection?
The Hurst exponent (H) measures how a price series scales over time. H above 0.5 indicates persistent, trending behaviour where past direction has a positive correlation with future direction. H below 0.5 indicates mean-reverting behaviour. It is calculated from the rescaled-range statistic over rolling windows, producing a dimensionless reading that classifies regime without requiring an arbitrary price-level threshold.
Does regime classification tell me whether to buy or sell?
No — and that distinction matters. Regime classification is a diagnostic tool, not a directional signal. It tells you which type of market behaviour is active, so you can choose strategies and interpret indicators appropriate to that state. Direction — bullish or bearish — comes from the signals applied within the identified regime. Nothing here constitutes trading advice.
See regime detection in action on TradePulse
TradePulse's AI commentary classifies the current NIFTY and Bank Nifty regime in real time, weights signals accordingly, and tells you the specific conditions that would signal a regime shift.