Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Dynamic Asset Pricing: Integrating FinBERT-Based Sentiment Quantification with the Fama--French Five-Factor Model

Chi Zhang

TL;DR

This study addresses how time-varying sentiment derived from FinBERT can be integrated into asset pricing. It develops a daily sentiment index $S_t$ and its volatility $HV_t$, and embeds them into the Fama–French five-factor model, including an interaction with $HV_t$ to capture regime shifts, evaluated via rolling regressions and event studies around major shocks such as the June 15, 2022 Fed rate hike. The results show sentiment has a positive effect on returns in normal periods but can reverse or amplify under high uncertainty, with rolling Coefficients $\,\hat{\gamma}_w$ indicating strong time variation and conditional relevance. An event study demonstrates the sentiment-augmented model better explains abnormal returns around policy announcements, supporting the usefulness of NLP-derived sentiment in real-time asset pricing and its potential applications for investors and regulators.

Abstract

This paper presents a comprehensive study on the integration of text-derived, time-varying sentiment factors into traditional multi-factor asset pricing models. Leveraging FinBERT, a domain-specific deep learning language model, we construct a dynamic sentiment index and its volatility from large-scale financial news and social media data covering 2020 to 2022. By embedding these sentiment measures into the Fama French five-factor regression, we rigorously examine whether sentiment significantly explains variations in daily stock returns and how its impact evolves across different market volatility regimes. Empirical results demonstrate that sentiment has a consistently positive impact on returns during normal periods, while its effect is amplified or even reversed under extreme market conditions. Rolling regressions reveal the time-varying nature of sentiment sensitivity, and an event study around the June 15, 2022 Federal Reserve 75 basis point rate hike shows that a sentiment-augmented five-factor model better explains abnormal returns relative to the baseline model. Our findings support the incorporation of high-frequency, NLP-derived sentiment into classical asset pricing frameworks and suggest implications for investors and regulators.

Dynamic Asset Pricing: Integrating FinBERT-Based Sentiment Quantification with the Fama--French Five-Factor Model

TL;DR

This study addresses how time-varying sentiment derived from FinBERT can be integrated into asset pricing. It develops a daily sentiment index and its volatility , and embeds them into the Fama–French five-factor model, including an interaction with to capture regime shifts, evaluated via rolling regressions and event studies around major shocks such as the June 15, 2022 Fed rate hike. The results show sentiment has a positive effect on returns in normal periods but can reverse or amplify under high uncertainty, with rolling Coefficients indicating strong time variation and conditional relevance. An event study demonstrates the sentiment-augmented model better explains abnormal returns around policy announcements, supporting the usefulness of NLP-derived sentiment in real-time asset pricing and its potential applications for investors and regulators.

Abstract

This paper presents a comprehensive study on the integration of text-derived, time-varying sentiment factors into traditional multi-factor asset pricing models. Leveraging FinBERT, a domain-specific deep learning language model, we construct a dynamic sentiment index and its volatility from large-scale financial news and social media data covering 2020 to 2022. By embedding these sentiment measures into the Fama French five-factor regression, we rigorously examine whether sentiment significantly explains variations in daily stock returns and how its impact evolves across different market volatility regimes. Empirical results demonstrate that sentiment has a consistently positive impact on returns during normal periods, while its effect is amplified or even reversed under extreme market conditions. Rolling regressions reveal the time-varying nature of sentiment sensitivity, and an event study around the June 15, 2022 Federal Reserve 75 basis point rate hike shows that a sentiment-augmented five-factor model better explains abnormal returns relative to the baseline model. Our findings support the incorporation of high-frequency, NLP-derived sentiment into classical asset pricing frameworks and suggest implications for investors and regulators.
Paper Structure (61 sections, 9 equations, 2 figures, 2 tables)

This paper contains 61 sections, 9 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Background
  3. Motivation
  4. (a) Dynamic vs. Static Methods
  5. (b) Limitations of Traditional Lexicon-Based Techniques
  6. (c) The Overlooked Role of Sentiment Volatility
  7. (d) Integrating Sentiment into Classical Multi-Factor Models
  8. Literature Review
  9. Sentiment‑Analysis Methodologies
  10. Sentiment and Asset Pricing
  11. Dynamics of Sentiment Effects
  12. Methodology
  13. Data Sources
  14. 1. Financial News Data
  15. 2. Social Media Data
  16. ...and 46 more sections

Figures (2)

  • Figure 1: Rolling $60$‑day coefficient on the sentiment factor $(\hat{\gamma}_{w})$, 2020‑01‑02 -- 2022‑12‑30. The horizontal line at zero aids visual interpretation.
  • Figure 2: Cumulative abnormal return (CAR) around the 15 June 2022 FOMC announcement.