Chinese FOG Index: the Readability of Information Disclosure in Chinese Listed Companies

The disclosure of non-financial information from listed companies has a significant impact on Chinese market. How can we apply advanced techniques (e.g: big data analysis ) to analyze text-format non-financial information? The FOG index, proposed by Robert Gunning in 1952, is the most commonly adopted text readability index; unfortunately, it is only suitable for English text. In this study, we choose annual financial reports of 200 listed Chinese companies, using machine learning and text-mining methods, to build a new index which is suitable for measuring the readability of Chinese text. Statistical methods such as Cluster Analysis, Ridge Regression, and LARS regression are also utilized to get the final model. Then we apply the proposed index to prospectuses of Chinese listed companies in the Sci-Tech innovation board to measure their readability and thus the quality of their information disclosure. We tentatively build an assessment model for Chinese text readability which may enhance the intelligibility and observability of non-financial information disclosure quality in the Chinese market.