Time partitioning model finds persistent patterns in financial time series
Time-partitioning Gaussian Process Mixture Model
Machine Learning
Summary
This paper presents a new statistical model to identify lasting patterns and changes in time-based data, like stock market numbers, without assuming simple dependencies within each pattern. The authors developed a method that adapts to different lengths of time periods and can measure uncertainty about when changes happen. Their model helped spot changes in financial indices that coincide with major events in markets and policy. They also found that adjusting certain settings can improve prediction accuracy, but might reduce how clearly the model identifies meaningful patterns.
What this means in practice
- •For financial analysts: Detect and interpret persistent regime shifts in market indicators aligned with major economic events to inform investment decisions.
- •For risk management teams: Identify and quantify shifts in asset volatility regimes to better assess and adapt to changing financial risks.
Authors
Guilherme Piantino, Hedibert F. Lopes
Abstract
We propose the Time-partitioning Gaussian Process Mixture Model (TPGPMM), a Bayesian nonparametric framework for identifying persistent regimes and modelling nonstationary dependence in time series. The model combines an adaptive Pitman--Yor prior on contiguous segmentations with regime-specific Gaussian processes, without imposing Markovian within-regime dependence. Learning the Pitman--Yor discount parameter allows adaptation to homogeneous and heterogeneous regime lengths. A split--merge Metropolis-within-Gibbs sampler with exact Gibbs boundary updates jointly learns change points and model hyperparameters, quantifying uncertainty over segmentations and within-regime dynamics. With an RBF kernel and a fixed, conservative observation-noise level, TPGPMM favours sparse segmentations that distinguish persistent shifts from transient shocks. Simulations demonstrate recovery of the regime structure. Applications to the CBOE Volatility Index (VIX) and the 10-year United States Treasury yield identify regimes aligned with major market and monetary-policy events. Variants that estimate observation noise improve predictive accuracy relative to the main specification, but can produce less sparse or less event-aligned segmentations. A variant with regime-specific means fits better than a fitted stationary Gaussian process and forecasts about as accurately. These results highlight a trade-off between interpretable regime identification and predictive accuracy.