基于BP神经网络的中国创业板IPO定价模型研究

发布时间：2018-01-03 04:36

本文关键词：基于BP神经网络的中国创业板IPO定价模型研究　出处：《华南理工大学》2012年硕士论文　论文类型：学位论文

【摘要】：随着国内外金融行业的发展，中国创业板市场于2009年9月在深圳交易所推出，使中国的证券市场形成了主板、中小板、创业板为重要组成部分的多层次资本市场。中国创业板市场的推出一方面满足了处于高速发展中小型企业的融资需求，另一方面也将社会闲散资金聚集到资本市场，促进了资本的优化配置。但创业板市场仍是一个新兴的不成熟的市场，其IPO定价机制和方式并不完善，，导致的IPO抑价现象相对于主板市场比较严重。因此，研究中国创业板公司IPO定价问题，提高新股定价效率，对完善中国创业板市场体系、提升中小企业融资效率以及改善中国资本市场运营环境都具有很强的理论和实践意义。创业板IPO是具有高成长性的中小型企业首次公开发行股票并在交易所上市交易的行为。由于创业板上市主体所处的行业和经营业务的创新性，很难寻找到处于同行业、同发展阶段的可比公司作为定价参考对象，同时公司的经营时间较短，大多处于起步期，未经历行业发展周期，历史经营数据较少，难以准确衡量企业未来发展状况。传统的自由现金流贴现法、相对估值法和经济附加值估值法都不再适用于创业板IPO定价，很多学者开始从实物期权法、神经网络等新的角度进行研究。本文立足于创业板IPO定价的动态复杂性，通过分析BP神经网络的原理和算法，力求解决创业板IPO定价需涉及的非线性和动态性的数学问题，突出创业板IPO合理定价的重要性。本文阐述中国创业板现有的IPO定价及发行方式，结合该市场及上市主体的特征，从内外部两方面挖掘影响中国创业板公司IPO定价的因素；基于BP神经网络理论构建中国创业板IPO定价模型，运用上市公司数据进行模型的训练和测试，揭示创业板IPO定价与主板IPO定价的区别，以提高创业板IPO定价效率。最后，进行模型进行优化度分析，并给出模型应用的建议，以便提高创业板定价的合理性和规范性，舒缓创业板市场新股定价高度抑价的现状，为构建有效的中国创业板IPO定价机制提供借鉴意义。
[Abstract]:With the development of financial industry at home and abroad, China's gem market was launched on the Shenzhen Stock Exchange in September 2009, which made China's securities market form a main board, a small and medium-sized board. Growth Enterprise Market is an important part of the multi-level capital market. China's gem market on the one hand meets the financing needs of small and medium-sized enterprises in high-speed development. On the other hand, social idle funds are also gathered into the capital market, which promotes the optimal allocation of capital. However, the gem market is still a new immature market, and its IPO pricing mechanism and methods are not perfect. The IPO underpricing phenomenon is more serious than the main board market. Therefore, the study of the IPO pricing problem of China gem companies, improve the pricing efficiency of new shares, to improve the gem market system. It is of great theoretical and practical significance to improve the financing efficiency of SMEs and to improve the operating environment of Chinese capital market. Growth Enterprise Market (IPO) is a high growth small and medium-sized enterprise initial public offering of shares and trading on the exchange, due to the gem listed in the main body of the industry and business innovation. It is difficult to find comparable companies in the same industry and development stage as pricing reference. At the same time, the operating time of the company is shorter, most of them are in the initial stage, not experienced the industry development cycle, the historical management data is less. It is difficult to accurately measure the future development of enterprises. Traditional free cash flow discount method, relative valuation method and economic value added valuation method are no longer applicable to gem IPO pricing, many scholars began from real options method. Based on the dynamic complexity of IPO pricing in gem, this paper analyzes the principle and algorithm of BP neural network. This paper tries to solve the nonlinear and dynamic mathematical problems involved in the IPO pricing of gem, and highlights the importance of rational pricing of gem IPO. This paper expounds the existing pricing and issuing methods of IPO in China growth Enterprise Market (gem), combines the characteristics of the market and the main body of the market, and excavates the factors that affect the pricing of IPO from the internal and external aspects. Based on BP neural network theory, this paper constructs the IPO pricing model of China gem, and uses listed company data to train and test the model to reveal the difference between gem IPO pricing and mainboard IPO pricing. In order to improve the efficiency of gem IPO pricing. Finally, the optimization of the model analysis, and give recommendations for the application of the model, in order to improve the rationality and standardization of gem pricing. To ease the current situation of high underpricing of new shares in gem market, and to provide reference for the construction of effective IPO pricing mechanism of gem.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;TP18

【相似文献】