Phase transition phenomenon: A compound measure analysis

Abstract

This study investigates the well-documented phenomenon of phase transition in financial markets using combined information from both return and volume changes within short time intervals. We suggest a new measure for the phase transition behaviour of markets, calculated as a return distribution conditional on local variance in volume imbalance, and show that this measure successfully captures phase transition behaviour under various conditions. We analyse the intraday trade and quote dataset from the KOSPI 200 index futures, which includes detailed information on the original order size and the type of each initiating investor. We find that among these two competing factors, the submitted order size yields more explanatory power on the phenomenon of market phase transition than the investor type. (C) 2015 Elsevier B.V. All rights reserved.