Systematic signals for when volatility is due to revert to its long-run average
Volatility is mean-reverting by nature — periods of extreme fear or extreme calm are always temporary. The Mean Reversion Engine quantifies exactly how far current volatility has deviated from its long-run equilibrium and estimates the probability and timeline for reversion using the Ornstein-Uhlenbeck stochastic process. This creates high-conviction trade signals for both premium sellers (when IV is elevated) and volatility buyers (when IV is suppressed).
Systematic signals for when volatility is due to revert to its long-run average
Unlock institutional-grade portfolio optimization and quantitative modeling tools.
Institutional Plan
Uses the OU stochastic process — the standard academic model for mean-reverting processes — to estimate the speed and magnitude of IV reversion for each ticker. Calibrated on 3 years of daily IV history.
For any current IV level, shows the probability that IV will revert to within 10%, 20%, or 30% of its long-run mean within 30, 60, or 90 days — directly informing expiration selection.
Calculates the statistical half-life of the current IV deviation — the expected time for IV to close half the gap to its mean. A half-life of 15 days suggests a 30-45 DTE trade; 45 days suggests a longer-dated position.
Classifies the current volatility regime (low, normal, elevated, extreme) using a Hidden Markov Model, and adjusts all reversion signals for the current regime. Regime persistence can override mean reversion.
Overlays macro volatility drivers (VIX term structure, credit spreads, macro event calendar) to distinguish regime-driven IV from idiosyncratic spikes — filtering out false reversion signals.
Aggregates mean-reversion signals across your entire watchlist to identify when broad portfolio volatility is at an extreme — the ideal time to add or reduce premium-selling exposure.
Each reversion signal is accompanied by a confidence score (0–100) that combines the OU deviation measure, regime classification, and macro overlay. Only high-confidence signals (70+) are highlighted by default.
Every signal type is accompanied by its historical win rate and average P&L over the past 5 years — giving you the base-rate context to evaluate each opportunity objectively.
The Ornstein-Uhlenbeck model is calibrated to the last 3 years of daily IV data for each stock. This gives three parameters: the long-run mean (theta), the speed of mean reversion (kappa), and the volatility of volatility (sigma).
The deviation is expressed as a standardized OU score: how many standard deviations the current IV sits above or below the estimated long-run mean. This is more precise than a simple Z-score because it accounts for the mean-reverting dynamics.
Using the kappa parameter (speed of mean reversion), the engine calculates the expected half-life: the time for IV to close half the gap to its mean. This directly informs expiration selection.
If the macro environment is in a high-volatility regime (VIX above 25, credit spreads widening), the engine reduces confidence in reversion signals because regime persistence can override mean reversion. Signals are color-coded by confidence.
The final output is a ranked list of tickers where mean reversion is both statistically likely and macro-consistent, with the suggested strategy, expiration, and position size. One click routes to execution.
Volatility Anomaly — Platform Overview
A complete walkthrough of the Volatility Anomaly platform: the screener, backtester, research library, and all Professional modules.
1:12
Full platform access
Upgrade to Institutional to use Volatility Mean Reversion Engine