Index-based Investing and Intraday Stock Dynamics

Abstract

We investigate the implications of index-based investing on intraday stock dynamics using a large high-frequency data set, which consists of trade prices sampled every five seconds for all S&P 500 constituents from 2004 to 2019. We estimate multiple aspects of intraday stock dynamics, including trading volume, volatility, correlation, and beta using estimators that are efficient in the high-frequency setting. We find informative intraday patterns that evolve over time. For example, in the recent decade, the correlation and trading volume increase substantially at the end of trading session; the betas of stocks converge towards one during the day. We show these patterns can be attributed to a ''firm-specific open, index-based close'' intraday trading profile. We theoretically support our findings via a market impact model with time-varying informational shocks and trading demands from single-stock and index investors.

Keywords: index-based investing, intraday stock dynamics, high-frequency data, market impact model

JEL Classification: G11, G12