Abstract:
Against the backdrop of data assets increasingly becoming core corporate competitiveness, some enterprises may increase the frequency of data asset information disclosure in annual reports to compete for more financial resources. Using listed companies on Shanghai and Shenzhen A-share markets as samples, this study employs machine learning and text analysis methods to construct a data asset utilization lexicon and uses a regression residual model to measure corporate data asset over-disclosure behavior. The findings reveal that data asset over-disclosure fails to enable enterprises to obtain credit resources as "wished", but rather "counterproductively" inhibits credit availability. This effect is more pronounced in state-owned enterprises, enterprises with lower operating environment uncertainty, and those with stronger bank-enterprise relationships. The mechanisms operate through: (1) The uneven quality of over-disclosed data asset information reduces corporate information transparency, making it difficult for banks to assess enterprises' true financial conditions and future development potential. (2) Data asset over-disclosure increases analysts' cognitive burden, inducing optimistic bias in analyst forecasts, which signals to the market that enterprises are concealing negative information and their development potential is overestimated. (3) Enterprises strategically disclose data asset information, creating "data theater", and accelerating resource misallocation to non-core areas and elevating operational risks. Further research finds that credit contraction induced by data asset over-disclosure and also triggers high-risk investment and financing behaviors such as shadow banking activities and short-term borrowing for long-term investment, exacerbating corporate financial distress. This study provides empirical evidence for strengthening data asset information disclosure regulation, preventing hot concept speculation, and improving credit risk identification capabilities.
下载:
