Abstract
In the era of big data, the significance of data has skyrocketed, transforming into a new type of asset and a significant element of production. This research focuses on A-share listed companies in China from 2007 to 2022, utilizing a Word2Vec neural network model and deep learning text mining to investigate the ramifications of data asset information disclosure on corporate capital market effectiveness. The findings indicate that data asset disclosure significantly enhances market efficiency by improving corporate governance, increasing institutional investors’ shareholding, and boosting stock liquidity. These effects are most pronounced in growing, high-technology, and state-owned enterprises.
Introduction: The Data Asset "Black Box"
In the modern digital economy, data has been elevated to a "fifth factor of production" alongside land, labor, capital, and technology. However, accounting for these assets remains a challenge. While China introduced interim provisions for data resource accounting in 2024, historical data is largely buried in the unstructured text of annual reports.
This study addresses whether the voluntary disclosure of data assets provides a genuine signal to the market or is merely "window dressing." If investors cannot effectively evaluate a firm's data assets, they may misjudge the company's value. The central hypothesis is that quantifying these textual disclosures can predict improvements in capital market efficiency—specifically, helping stock prices reflect firm-specific news rather than broad market trends.
Methodology
The study employs a multi-step empirical approach combining Natural Language Processing (NLP) and econometrics across a sample of 25,322 firm-year observations:
- Quantifying Disclosure via Word2Vec: Rather than simple keyword counting, the study uses a Word2Vec neural network model to build a custom dictionary. Starting with seed words like "data assets," the model identifies semantically similar terms within the corpus of Chinese annual reports.
- Measuring Market Efficiency: The paper uses Stock Price Synchronicity (SYN) as a proxy. High synchronicity suggests a stock moves in lockstep with the market, while low synchronicity suggests the market is successfully "pricing in" details unique to that specific company.
- Regression Analysis: A fixed-effects model tests the relationship between disclosure metrics and market efficiency, controlling for variables like firm size, leverage, and growth.
Theoretical Framework and Mechanisms
Data asset disclosure acts as a positive signal, suggesting a firm is managing its digital transformation effectively. By reducing information asymmetry, firms lower the cost for "arbitrage traders" to collect private information, encouraging them to trade on firm-specific news.
The study identifies three primary channels through which disclosure improves efficiency:
- Corporate Governance: Enhanced transparency leads to better internal oversight.
- Institutional Magnet: Professional investors are more likely to increase holdings in firms with high disclosure as they have the expertise to value digital assets.
- Stock Liquidity: Increased information flow improves the ease with which shares are traded.
Key Results and Insights
- Efficiency Boost: A standard deviation increase in data asset disclosure improves capital market efficiency by an average of 6.57%.
- Life Cycle Heterogeneity: The impact is strongest for firms in growing and mature phases. For declining firms, disclosure has no significant impact, as investors focus more on survival than digital innovation.
- Enterprise Type: The effect is more pronounced in high-technology enterprises, state-owned enterprises (SOEs), and firms with high media or investor focus.
Conclusion
While the study primarily measures the intensity of disclosure rather than its veracity, it provides strong evidence that transparency regarding data assets is a vital component of modern market efficiency. As China moves toward mandatory reporting, the ability of the market to price these "invisible" assets will be crucial for fostering a high-quality capital market. Future research should look beyond frequency to analyze the sentiment and quality of these disclosures.
Summary updated