Study of statistical correlations in intraday and daily financial return time series

TL;DR

This paper investigates correlations and co-movements in intraday and daily stock data, revealing increasing average correlations during the day and exploring market structure dynamics for potential trading strategies.

Contribution

It combines analysis of intraday correlation dynamics with multidimensional scaling visualization of daily market structures, including sector detection and crisis periods.

Findings

Average stock correlation increases throughout the trading day

No significant structural change in market during a day

Visualization aids in identifying stock pairs for trading

Abstract

The aim of this article is to briefly review and make new studies of correlations and co-movements of stocks, so as to understand the "seasonalities" and market evolution. Using the intraday data of the CAC40, we begin by reasserting the findings of Allez and Bouchaud [New J. Phys. 13, 025010 (2011)]: the average correlation between stocks increases throughout the day. We then use multidimensional scaling (MDS) in generating maps and visualizing the dynamic evolution of the stock market during the day. We do not find any marked difference in the structure of the market during a day. Another aim is to use daily data for MDS studies, and visualize or detect specific sectors in a market and periods of crisis. We suggest that this type of visualization may be used in identifying potential pairs of stocks for "pairs trade".