The general well-being and happiness of human beings depends on an economy. Well-designed markets are essential to the development of an economy. However, a market works well only if it is well designed. If it is not well designed, it can actually harm an economy. Moreover, designing a market well is never easy because a market is a highly complex system including micro-macro interactions, self-fulfilling prophecy and butterfly effects. Traditional mathematical economic models cannot directly handle complex systems. Only an agent-based artificial financial market model (ABAFMM), a kind of computational model, can handle them directly. ABAFMMs are better at providing insight into the mechanism, but worse at making accurate quantitative predictions, and there is no perfect model for all purposes. Even Schelling’s model cannot predict the final exact distribution, although it has given us great insights about segregations. Since an ABAFMM is a completely virtual and artificial computer-based financial market and requires no empirical data, only an ABAFMM can investigate situations that have never before occurred in actual financial markets, and can isolate the direct effect of changing rules and regulations. Therefore, ABAFMMs are absolutely required to design a market that works well.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Financial Market Design and Agent-Based Model

  • Takanobu Mizuta,
  • Isao Yagi

摘要

The general well-being and happiness of human beings depends on an economy. Well-designed markets are essential to the development of an economy. However, a market works well only if it is well designed. If it is not well designed, it can actually harm an economy. Moreover, designing a market well is never easy because a market is a highly complex system including micro-macro interactions, self-fulfilling prophecy and butterfly effects. Traditional mathematical economic models cannot directly handle complex systems. Only an agent-based artificial financial market model (ABAFMM), a kind of computational model, can handle them directly. ABAFMMs are better at providing insight into the mechanism, but worse at making accurate quantitative predictions, and there is no perfect model for all purposes. Even Schelling’s model cannot predict the final exact distribution, although it has given us great insights about segregations. Since an ABAFMM is a completely virtual and artificial computer-based financial market and requires no empirical data, only an ABAFMM can investigate situations that have never before occurred in actual financial markets, and can isolate the direct effect of changing rules and regulations. Therefore, ABAFMMs are absolutely required to design a market that works well.