Clustering Volatility Regimes for Adaptive Risk Avoidance

Summary

The study presents an unsupervised method for identifying a data-driven number of volatility regimes in nonstationary financial time series. It uses change-point detection to divide each series into locally stationary segments, measures distances between the segment distributions, and clusters those segments into discrete regimes through an optimization procedure. The authors apply the method to financial indices, large-cap equities, exchange-traded funds, and currency pairs.

They also describe a dynamic trading strategy that matches the current distribution of a series to its historical regimes and uses that match to make online risk-avoidance decisions. The reported contribution is a simpler descriptive representation of past volatility behavior and a validated strategy based on regime matching. The document gives no performance figures, transaction-cost analysis, or detail on how the validation was conducted. The method reduces reliance on the rigid assumptions of some parametric regime-switching models, but its practical value still depends on stable estimation and useful correspondence between historical regimes and current conditions.

Key ideas

• Change-point detection divides a nonstationary series into locally stationary segments.
• Distances between segment distributions are used to cluster segments into a learned number of volatility regimes.
• The method is applied to indices, large-cap equities, ETFs, and currency pairs.
• A dynamic strategy matches current distributions to historical regimes to guide online risk avoidance.

Tags

# Structural clustering of volatility regimes for dynamic trading strategies


# Structural clustering of volatility regimes for dynamic trading strategies









We develop a new method to find the number of volatility regimes in a nonstationary financial time series by applying unsupervised learning to its volatility structure. We use change point detection to partition a time series into locally stationary segments and then compute a distance matrix between segment distributions. The segments are clustered into a learned number of discrete volatility regimes via an optimization routine. Using this framework, we determine a volatility clustering structure for financial indices, large-cap equities, exchange-traded funds and currency pairs. Our method overcomes the rigid assumptions necessary to implement many parametric regime-switching models, while effectively distilling a time series into several characteristic behaviours. Our results provide significant simplification of these time series and a strong descriptive analysis of prior behaviours of volatility. Finally, we create and validate a dynamic trading strategy that learns the optimal match between the current distribution of a time series and its past regimes, thereby making online risk-avoidance decisions in the present.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.