When the Rules Change: Adaptive Signal Extraction via Kalman Filtering and Markov-Switching Regimes

TL;DR

This paper develops an adaptive modeling approach combining Kalman filtering and Markov-switching regimes to analyze how investor behavior and market relationships change across different market states, with a focus on the Korean stock market from 2020 to 2024.

Contribution

It introduces a novel adaptive framework that captures regime-dependent investor behavior and demonstrates the importance of proper validation in microstructure analysis.

Findings

Foreign investor predictive power increases during crises.

Individual investors react asymmetrically to positive and negative shocks.

In-sample patterns do not reliably generalize out-of-sample.

Abstract

Most empirical microstructure research assumes that order flow--return parameters are constant, yet these relationships shift substantially across market regimes. Combining adaptive Kalman filtering, Markov-switching regime identification, and asymmetric response estimation, we characterize regime-dependent investor behavior in the Korean stock market during 2020--2024 using daily transaction data disaggregated by investor type. Three principal findings emerge: foreign investor predictive power increases several-fold during crisis periods relative to bull markets; individual investors chase momentum asymmetrically, reacting far more strongly to positive than to negative shocks; and independent information-theoretic validation corroborates both patterns. Rigorous out-of-sample testing reveals that these in-sample regularities do not generalize reliably, underscoring the need for proper validation methodology in microstructure research.