Improving Industry Momentum with Sentiment Signals

Refining Momentum with News Sentiment and News Dispersion

Introduction

Extensive research highlights the profitability of industry momentum strategies and their role in partially explaining stock momentum returns. However, traditional industry momentum relies exclusively on historical price data, which may overlook other valuable industry signals, such as news sentiment and its variability.

In this post, I explore how incorporating industry-level news sentiment and its dispersion enhances traditional momentum strategies. By combining momentum, news sentiment, and news dispersion, and testing them across various industry classifications, I find significant performance improvements over traditional momentum. These improvements are particularly strong for more granular industry classifications, where the additional signals provide greater predictive power and diversification opportunities.

Table of Contents

• Background

  • Industry Momentum

  • News Sentiment

  • News Dispersion

• Strategies

  • Price Data

  • Industry Data

  • Sentiment Data

  • Strategy Construction

• Results

• Investor Takeaways