Corporate environmental pictures information disclosure and investor market reaction: A new perspective from large-scale pictures feature mining

Abstract

Considering the new perspective of pictures, this study uses the new method of large-scale image processing technology based on deep learning to analyze the corporate environmental information picture disclosure strategies and studies the new problem of the impact path of picture disclosure strategies on investor market reaction. Our results show that, among content, color, and channel disclosure strategies, the content of environmental awareness pictures has the highest proportion (35.01%), followed by renewable energy (19.92%) and environmental certification (19.56%). Among 11,035 sample pictures, white and blue pictures are the most used with 2886 and 2700, respectively, these mainly comprise the content of environmental facilities and renewable energy. Compared with social responsibility reports, enterprises prefer to disclose pictures on social media channels. The regression results show that the greater the content of environmental pictures, the wider the disclosure channels, and the more positive the investor market reaction. However, color does not elicit positive investor market reactions. Moreover, analyst attention plays an intermediary role between pictures environmental information disclosure and investor market reactions. In conclusion, this study provides a new perspective for the study of environmental information disclosure, enriches the existing research, and provides referable practical suggestions for the sustainable development of enterprises and the efficient operation of capital markets.