A new approach to the modeling of financial volumes

TL;DR

This paper introduces a semi-Markov model to analyze high-frequency financial volume data, capturing key empirical features like dependence, periodicity, and asymmetry in stock trading volumes.

Contribution

It proposes a weighted-indexed semi-Markov chain model for intraday volume dynamics, demonstrating its effectiveness on real Italian stock market data.

Findings

Model reproduces empirical volume patterns

Captures intra-daily periodicity and asymmetry

Shows dependence in high-frequency volume data

Abstract

In this paper we study the high frequency dynamic of financial volumes of traded stocks by using a semi-Markov approach. More precisely we assume that the intraday logarithmic change of volume is described by a weighted-indexed semi-Markov chain model. Based on this assumptions we show that this model is able to reproduce several empirical facts about volume evolution like time series dependence, intra-daily periodicity and volume asymmetry. Results have been obtained from a real data application to high frequency data from the Italian stock market from first of January 2007 until end of December 2010.