|
|
|
| Regulation Intelligence and IPO Pricing Efficiency—Evidence from IARP |
| LIU Chun, SUN Liang
|
| International School of Business and Finance, Sun Yat-sen University |
|
|
|
Abstract China's capital market has long struggled with the problem of an inadequate initial public offering (IPO) pricing function. Despite improvements in the registration system, there are still many instances of extreme IPO overpricing or underpricing in the science and technology innovation market (STAR) and growth enterprise market (GEM). Furthermore, the underpricing of IPOs in China's capital market is typically significantly higher than that in developed countries such as the United States and United Kingdom. Excessive underpricing significantly raises financing costs, which in turn distorts firms' investment decisions and seriously undermines the investment and financing functions of the capital market. Therefore, investigating practical approaches to improving the efficiency of IPO pricing that are relevant to China's institutional background is a crucial task that must be undertaken in the current reform of China's capital market, especially in light of the complete implementation of the registration system.
In theory, regulation intelligence (hereinafter reg-intel), which heavily leverages artificial intelligence (AI) technologies such as machine learning (ML) and natural language processing (NLP), could be a useful tool for increasing the efficiency of IPO pricing. By reducing the cost of regulators' information integration, reg-intel can assist regulators in identifying and highlighting suspicious information and thus can bolster the credibility of IPO firms' information. Reg-intel also can assist regulators in finding and unlocking idiosyncratic information about the IPO firms, thus reducing information asymmetry for investors and improving the efficiency of IPO pricing. In practice, in September 2020 the SSE investigated the application of the intelligent assisted review platform (IARP), which uses AI technologies such as ML and NLP to facilitate IPO reviews. Using the IARP as a quasi-natural experiment and adopting the difference-in-differences method, this paper systematically analyses and tests the effect of reg-intel on the efficiency of IPO pricing and the underlying mechanism for the first time.
This paper reports the following findings. (1) IPO pricing efficiency improves significantly when the IARP is introduced. (2) The results hold after a series of robustness tests. (3) There is no evidence of an ex-ante trend of changes in pricing efficiency, supporting the validity of the parallel trend hypothesis and a causal interpretation of the results. (4) The IARP exerts its effects through two channels: it strengthens the credibility of the IPO firm's information and reduces information asymmetry for investors. (5) The IARP is more effective for IPOs with poorer information environments and stronger technological attributes. (6) The IARP significantly reduces the first-day stock turnover rate and post-listing price correction volatility of IPOs.
This paper makes three contributions to the field. First, it enriches and extends research on the efficiency of IPO pricing in the context of the registration system. This paper provides the first evidence that reg-intel can effectively improve the efficiency of IPO pricing in this context, thus contributing new ideas and evidence to the relevant literature and providing theoretical references and directions for the practice of reg-intel in China's capital market. Second, this paper enriches and extends research on reg-tech in China's capital market. This paper is the first to discuss the advanced stages of AI-based reg-tech and the resulting impact on the cost of information integration for regulators. It thus adds both a new perspective on the role of reg-tech and new evidence to support the evaluation of reg-tech's real efficacy in China's capital markets. Third, this paper enriches and extends research related to the impact of AI technology on the financial industry. Unlike previous literature, which mainly examines financial intermediaries, this paper focuses on the use of AI by capital market regulators, thus adding new evidence and suggesting new directions.
The results of this paper suggest that reg-intel can significantly improve the efficiency of IPO pricing. Therefore, to improve IPO pricing efficiency, authorities should further expand and deepen the construction of reg-intel in China's capital market and formulate a categorized, stratified, focused, and coordinated review strategy. Simultaneously, the authorities could also consider more imaginative answers to the problem of sensible capital market pricing in China. For example, an interactive platform where regulators and investors could exchange information might directly lower retail investors' costs associated with processing information.
|
|
Received: 06 September 2023
Published: 17 July 2024
|
|
|
|
|
|
|
|
