Granular Information and Sectoral Movements

Abstract

This paper shows a strong link between the granular information contained in individual stock prices and sectoral movements. Using machine learning algorithms, we find that a predictor aggregating the price movements of a broad cross section of individual stocks predicts sector ETF returns at intraday and lower frequencies. When we combine the information from structural models with machine learning, the resulting information signals have even stronger return predictability. A trading strategy that exploits the return predictability is profitable after trading costs. These results support theories of granular and network origins of aggregate shocks.

Keywords: Granular Information, Sectoral Movements, Exchange-Traded Funds, Machine Learning

JEL Classification: G10, G14, G40