The financial crisis in 2008 made countries recognize the great harm of financial systemic risk. After the crisis, all countries gradually realize the importance of macro-prudential supervision and adopt a variety of macro-prudential supervision tools to restrain systemic risk. As the prevention of financial systemic risk requires high timeliness and foresight, good communication plays a vital role in effective macro-prudential supervision. Therefore, countries should also improve the transparency of relevant macro-prudential policies in addition to using macro-prudential supervision tools. Countries have been using macro-prudential supervision verbal communication to guide the expectations of market entities, and scholars have been beginning to study its effect. However, there is a lack of research on whether verbal communication can effectively reduce the financial systemic risk in China. It is of great practical significance for China to improve the macro-prudential supervision framework and prevent financial systemic risk to measure the macro-prudential supervision verbal communication index reasonably and analyze its role in preventing financial systemic risk.

This paper discusses the above issues. The result shows that the macro-prudential supervision authority will conduct verbal communication on changes in financial systemic risk. And the changes of incoming links and outgoing links in the sectors of bank, security, insurance and trusts have an effect on verbal communication. Macro-prudential supervision verbal communication helps to reduce financial systemic risk with a lag of about half a year. And it mainly affects bank incoming links and insurance outgoing links. Moreover, based on the supervised learning method, we train sub-sample dictionaries to obtain prone phrases and their probability distributions, use text classifiers to automatically classify new communication texts, and finally construct the index of macro-prudential supervision verbal communication on new samples. The result shows that the index results obtained by different samples are relatively consistent, indicating that the quantitative method of macro-prudential supervision verbal communication in this paper is relatively robust, reproducible and malleable.

The expansion of this paper is mainly reflected in two aspects: First, considering the importance of foresight and timeliness in the prevention of financial systemic risk, this paper uses the text analysis method and the supervised learning method to construct the macro-prudential supervision verbal communication index from the perspective of verbal communication. It is more objective and reproducible. Second, this paper discusses the relationship between macro-prudential supervision verbal communication and financial systemic risk, which makes up for the lack of existing studies that mainly analyze the role of macro-prudential supervision tools and the impact of macro-prudential supervision verbal communication on financial markets.