Pattern recognition in micro-trading behaviors before stock price jumps: A framework based on multivariate time series analysis

TL;DR

This paper introduces a new multivariate time series framework to analyze micro-trading behaviors before stock price jumps, improving abnormality detection and attribute selection for better understanding of pre-jump trading patterns.

Contribution

It presents a novel methodology that incorporates temporal information, evaluates attribute abnormality, and explores joint informativeness to analyze pre-jump trading behaviors.

Findings

Identified highly informative trading attributes for predicting jumps.

Detected stocks with abnormal pre-jump trading patterns.

Provided a set of jump indicators for the Chinese stock market.

Abstract

Studying the micro-trading behaviors before stock price jumps is an important problem for financial regulations and investment decisions. In this study, we provide a new framework to study pre-jump trading behaviors based on multivariate time series analysis. Different from the existing literature, our methodology takes into account the temporal information embedded in the trading-related attributes and can better evaluate and compare the abnormality levels of different attributes. Moreover, it can explore the joint informativeness of the attributes as well as select a subset of highly informative but minimally redundant attributes to analyze the homogeneous and idiosyncratic patterns in the pre-jump trades of individual stocks. In addition, our analysis involves a set of technical indicators to describe micro-trading behaviors. To illustrate the viability of the proposed methodology, an application case is conducted based on the level-2 data of 189 constituent stocks of the China Security Index 300. The individual and joint informativeness levels of the attributes in predicting price jumps are evaluated and compared. To this end, our experiment provides a set of jump indicators that can represent the pre-jump trading behaviors in the Chinese stock market and have detected some stocks with extremely abnormal pre-jump trades.