Inter-Business Trading Structure Model with Agent-Based Simulation and Its Application to Real Data

Abstract

This paper proposes an inter-business trading structure model with agent-based simulation. The proposed model can deal with the two kinds of dynamic changes: (1) supply and demand volumes and the resulting production lead time, and (2) the combination of both decentralized and centralized inter-business structure. The model is applied to analyzing recent rapid market changes of Japanese firm data. The data is taken from the TDB (Teikoku Databank) business trading database, which contains firm information with four million trading items, business volumes and relations over many years. Using this information, we focused on changes in the trading of firms involved in the Japanese software industry. The experimental results applied to real data suggest that the dynamic structural changes determined by the model have a critical role in various cases of performance of Japanese software industry firms in the TDB database.