Investor negativity derived from news shapes firm distress dynamics under bounded rationality
Year: 2025
Type: article
Abstract: Purpose This study examines and explains the two-phase mechanisms through which investor negativity derived from news affects firm distress, drawing on prospect theory, bounded rationality and regret theory. Design/methodology/approach We apply prompt-based large language models (LLMs) to over 80,000 Vietnamese-language news articles to measure firm-level negativity sentiment. To test the proposed inverted U-shaped effect, we used ordered probit and logit regressions, which not only match the ordinal structure of distress levels but also enable the identification of threshold turning points in the sentiment–distress relationship. The sample included 80 listed Vietnamese firms that experienced and recovered from at least one distress episode between 2010 and 2022, ensuring sensitivity across distress intensities. Findings The results provide reliable empirical evidence for our proposed hypothesis: negative sentiment independently and nonlinearly influences distress outcomes. This is an inverted U-shaped relationship between negativity and firm distress. Research limitations/implications This study focused only on a tight, specific sample, which included only Vietnamese-listed non-financial firms that have operated continuously from 2008 until now and have experienced at least one instance of distress. Future research can be extended to a larger number of firms in other emerging economies. Practical implications Risk management should take into account the investor sentiment derived from news in their risk analysis and distress prediction models to enhance predictive accuracy. Originality/value This study integrates loss aversion and regret theories to demonstrate novel nonlinear dynamics linking negativity sentiment to firm distress, advancing our understanding of how behavioral responses evolve across different sentiment intensities.
Authors: Xuan Thi Thanh Pham, Dat Tan Huynh, Nhu Thi Kim Nguyen, Thu Bich Luu
Language: English
Subfield: Accounting
Domain: Social Sciences
Open Access status: diamond
Award: G62730574
OpenAlex ID: W4416674875