Game Analysis And Dynamic Simulation Of The Institutional Investors Regulation In Chinese Stock Market

Institutional investor is one of the most important kinds of investors in the financial market. With the progress of economic globalization and integration of global finance, institutional investors' behaviors are becoming the decisive factor of financial market, and they will play a more and more important role in stabilization and development of the domestic ever the international financial market. The stock market is the most important and active part of the financial market, and the main investing behaviors of institutional investors are operated in the stock market. The stock market in China is still a developing market nowadays, so there are many illegal investing behaviors in Chinese stock market, those illegal investing behaviors are blocking the deeper development of our stock and financial market. So it is an important thing to cultivate the institutional investors and regulate their illegal behaviors for developing Chinese financial and national economy.Cultivating and regulating the institutional investors not only deal with the law, but also the economic theory. Since 1950s, game theory has been applied in economic research; at the same time, along with the development of computer science and the emergence of CAS theory, agent-based modeling economic simulation has become an effective tool and virtual experiment environment for the economic research. The regulation of institutional investors in stock market deals with several kinds of investors and the regulator, the connection of profits between different agents is very complicated, so it's very efficient to apply game theory and economic simulation for researching the regulation policy in order to constitute a more efficient series of policies for cultivating and regulating the institutional investors.In order to found and develop our stock market rapidly, the "Securities Law of the People's Republic of China" defined the organization of institutional investor widely and amphiboly, and release the limit of institutional investor's behavior, by doing this, our stock market developed rapidly as the government's wish, but the looser investing environment breed some illegal behaviors of institutional investors. Nowadays, the research on regulation of institutional investor mostly done in the law field, the operation and effect of the regulation methods are not obvious enough; there is rare research on regulation of institutional investor done form the economic point, so deeper research should be done on the regulation method with stronger operational.A series of scientific, reasonable and operational regulation method of institutional investor is one of important part in the cultivating and developing the institutional investor and stock market in China, which is the crucial factor of the stock even the financial market's efficiency. In this thesis, we study the collusion, capital encroachment, inside trading the three kinds of illegal behaviors of institutional investor by game theory, and establish the dynamic simulation model of the game model in order to experiment in the computer environment for obtaining the ideal measures of the institutional investor's regulation. Form the crucial factors of regulation, which are strength of regulation and punishment and the adaptability of regulator, we study the behavior mode of institutional investors and changes of their investing strategies, and the effect and shaping process of the ideal regulation strategies in different virtual economic environments. Furthermore, we discuss the public character of the regulation resource, and research the economic root of the unperfected impact of the regulation, propose several measures to alleviate this phenomenon through game and simulation analysis.This major research, conclusions and innovation points are as follows:Firstly, a game model and its simulation model on Swarm are established for researching the illegal behavior of controlling the stock price of institutional investors, and experiment in different virtual economic environments to study the institutional investors' behavior mode, and the effectiveness of the self-adaptive regulation. By the game analysis and various simulation experiments, we provide that strengthening regulation punishment and the adaptation of regulation strategy will reduce the possibility the illegal behaviors happen.Secondly, the game and simulation models on Swarm are established for researching the illegal behaviors that inside trading and capital encroachment, and then we provide several effective regulation measures through the simulation experiments. The results show that the inner management of listed companies is the main reason which could make the illegal happen, the regulator should strength the regulation of listed companies in order to reduce the possibility the illegal behaviors happen.Thirdly, by researching on the regulation of institutional investors'market controlling behavior and the reality of Chinese institutional investors, a simulation system with more adaptive has been found in order to study similar questions, by doing this the simulation model is provided with higher compatibility and the research period and cost are reduced effectively. After the collusion research, we study another two kinds of illegal behaviors which are collude controlling and controlling the stock price by only one institutional investor with multiple personal accounts, and obtain reasonable conclusions. Fourthly, making the regulator in the model with the form of independent agent, and enduing it with proper adaptability through genetic algorithm. In order to improve the regulation effect, the organization of regulation should be improved and some encouragement mechanism must be built in the inner regulator in order to make the regulation strategies more adaptive.Fifthly, empirical data and simulation model are applied simultaneously. Usually the simulation models always make some important functions hypothetical, which may reduce the creditability of the simulation experiments and their conclusions. In this thesis, we obtain the real data of the representative stock by econometrics or as the input for the simulation function form in order to do "empirical" research using by the "virtual" simulation model, and the results can be more creditability and scientific.