Graduate School of Business Administration
Research on theory, corporate value and asset price concerning investment and financing has been actively reported in the field of finance since the 1950s, and so far, the theory on the company's capital structure (Modigliani, Miller), the capital asset valuation model Sharpe), portfolio theory (Markowitz), derived securities price theory (Black, Scholes, Merton) and other excellent theories have been proposed. Many of these traditional theories have been proposed on the premise of market efficiency and reasonable decision making, and financial theory and the financial system have greatly improved by the contribution of these theories. Meanwhile, in recent years, behavioral economics (Kahneman, Tversky) casts doubts on the prerequisites of such a traditional theory, the field of finance is also showing new developments, It is being done.
In recent years, the role played by information is increasing in society and business. Under such circumstances, the importance of research utilizing information technology is increasing, and even in the fields of finance, economics, business administration, etc., it is necessary to develop machine learning, natural language processing technology, artificial intelligence technology, Many studies used have been conducted. Our laboratory is engaged in research on the method itself of these information technologies and its application. For example, in the past, research on utilizing machine learning, analysis of text data by natural language processing technology, analysis on big data, research on corporate technology development, relationship between companies using social network analysis Analysis, research that focuses on corporate management, etc.
In the field of computer science, a method called agent-based modeling has been proposed as a powerful method to analyze the relationship between micro rules and macroscopic behaviors. This approach is a bottom-up approach that attempts to explain the macro behavior of the whole system from local rules, and is a method suitable for analyzing a multi-agent system in which a large number of autonomously acting agents. Agent based models are applied in a wide range of fields, and numerous analyzes have been reported that apply this approach to the field of social science (Social Simulation). Our laboratory is conducting research on such agent base model. In particular, in recent years, we are conducting analyzes targeting the social sciences including finance.