The CENKI space economic simulator: Demonstrating agent-based modeling on satellite market data

Expanding investments into space ventures are encouraging a diverse range of opportunities, from in-space manufacturing to missions to Mars. The growing diversity and complex economic interdependencies introduce uncertainty into market and investment projections. The availability of up-to-date market data and market simulation tools are critical to understanding the risks and rewards of a commercial space endeavor. The current study addresses the need for market simulation tools and applies a newly-developed agent-based modeling tool to explore new markets for in-space products and services, evaluate sensitivities to emerging technologies and regulatory policies, and spur investment into future space operations. This study demonstrates the modeling and predictive capability of the agent-based approach to capture sales and acquisition behaviors in historical telecommunications satellite provider data. Particular attention is paid to the role of incorporating model uncertainty, both from insufficient data as well as probabilistic biases, to understanding operational decisions on spending allocation and scheduling. Market competition associated to the SES framework represents a method for simulating the natural evolution of a complex financial system made of many independent models of differing character. In order to explore future topics of space economic development and policy, this framework must be vetted as capable of producing intuitive market dynamics. To do so, a Study has been created including five peripheral Players supplying a single focal Player who is considering the near term trajectory in the communication satellite industry. This Study acts to both verify the function of the SES as well as explore the uncertainty in making predictions based on the available market data used and the interrelationships simulated. The SES is validated by predicting revenue generation for a company from 2006–2009 using empirical models up to 2005 and comparing the results to historical data. Furthermore, results of various merger and non-merger scenarios are compared and the real merger cost of a company acquisition is explained using discounted profits in the merger year of 2005.