An experimental approach to find a suitable simulation method in business economics

Abstract

The main goal of this paper is to compare the results of an agent-based and Monte Carlo simulation experiments in business process negotiation between sellers and customers of a simple trading commodity. The motivation of the presented research is to find suitable method for predicting key performance indicators of a business company. The intention is to develop a software module in the future which might help the management of business companies to support their decisions. Microeconomic demand functions were used as a core element in the negotiation. Specifically, Marshallian demand function and CobbDouglas utility functions is introduced. The paper firstly presents some of the principles of agent-based and Monte Carlo simulation techniques, and demand function theory. Secondly, we present a conceptual model of a business company in terms of a simulation framework. Thirdly, a formalization of demand functions and their implementation in a seller-to-customer negotiation is introduced. Lastly, we discuss some of the simulation results in one year of selling commodities. The results obtained show that agent-based method is more suitable than Monte Carlo in the presented domain, and the demand functions could be used to predict the trading results of a company in some metrics.