Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1894
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-19T10:42:13.763522Z
status: published
lens: g_space
sector: business
headline: Corporate Artificial Intelligence and Stock Price Crash Risk: Evidence from Chinese Listed Firms
dek: Using a panel of 45,385 firm-year observations of Chinese A-share listed firms over 2010-2025, this paper examines whether firm-level artificial intelligence (AI) development affects stock price crash risk. Two-way fixed effects estimates show that a higher level of corporate AI development significantly reduces both the negative conditional skewness and the down-to-up volatility of firm-specific returns. The finding survives an instrumental variable strategy, the exclusion of turbulent years, alternative measur…
gain_title: Higher corporate AI development was associated with lower negative skewness and down-to-up volatility in firm-specific returns, indicating reduced extreme downside risk for Chinese listed firms.
problem_title: (none)
trace_subject: (none)
gain_reading: Higher corporate AI development was associated with lower negative skewness and down-to-up volatility in firm-specific returns, indicating reduced extreme downside risk for Chinese listed firms.
gain_evidence: a higher level of corporate AI development significantly reduces both the negative conditional skewness and the down-to-up volatility of firm-specific returns
problem_reading: (none)
problem_evidence: (none)
quick_read: A study of 45,385 firm-year observations from Chinese A-share listed firms between 2010 and 2025 tested whether a firm's AI development changes its risk of a stock price crash. Two-way fixed effects estimates showed that more AI development went with lower negative skewness and lower down-to-up volatility in firm-specific returns.

The result matters because it moves the assessment of AI's economic effects beyond real-economy outcomes to extreme downside risk in capital markets. The abstract reports that the finding held across several robustness checks, but the evidence is specific to Chinese listed firms and the mitigating effect varies across firm types.
limitation: The mitigating effect is concentrated in particular groups, non-high-tech firms and firms in economically developed regions, so the finding does not apply uniformly across all firms in the sample.
tag: Evidence-backed gain
key_points: The sample covers 45,385 firm-year observations of Chinese A-share listed firms from 2010 to 2025. | Higher AI development reduced both negative conditional skewness and down-to-up volatility of firm-specific returns. | The result survived an instrumental variable strategy, exclusion of turbulent years, and alternative measures and specifications. | The mitigating effect is concentrated in non-high-tech firms and in firms in economically developed regions.
rundown: The authors used two-way fixed effects estimates to link firm-level AI development with stock price crash risk, measured through negative conditional skewness and down-to-up volatility of firm-specific returns.

To test robustness, the study applied an instrumental variable strategy, excluded turbulent years, and used alternative measures of both the dependent and explanatory variables as well as alternative specifications.

Mechanism tests indicate AI eases financing constraints and raises total factor productivity. Financing constraints amplify crash risk while productivity dampens it, which the authors use to explain the observed reduction.

Heterogeneity tests show the effect is concentrated in non-high-tech firms and in firms located in economically developed regions.
sources:
- peer_reviewed | Advances in Economics, Management and Political Sciences | https://doi.org/10.54254/2754-1169/2026.37745 | 2026-10-19
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

8f6acebd5b5102fc47c25c93d7d7d8816785a7c9691bcffab0222dc0c5ee842d

previous

0000000000000000000000000000000000000000000000000000000000000000