Gooding, Tim (2018) Foundational forces of trade : exploring trade in a complex environment using agent-based models. (PhD thesis), Kingston University, .

Abstract

In the short-run, there is not enough time for systemic feedback to organise and become a force. In this instance, it is appropriate to use models and theory underpinned by observed systemic causal links, such as supply and demand determining price. On the other hand, where there is enough time for complexity to become influential, the system might be best explored using complexity science. This work explores the fundamental forces arising from trade in the long-run, where there is enough time for complexity forces to become influential. The complexity research tool used was agent-based modelling. A simple trade model was created that can be fully explained in just a few lines of prose. Three areas of the economy were explored using this model with the following results: 1) In complex systems, long-run prices are shaped by the system's definition of evolutionary fitness regardless of supply and demand. 2) Econophysicists are condtionally correct in their hypothesis that money in a closed system maintains a Boltzmann-Gibbs distribution (0.5 Gini coefficient). 3) Experimentation with trade models support the Rand/Stonedahl experiments indicating that agent computational ability inversely correlates with efficiency. In addition to a simple trade model, a far more complex model was built that includes an environment with raw resources, agents who are labourers, business owners, and consumers. This model requires agents to successfully reproduce in order to avoid extinction. Almost all agent behaviour is governed by evolutionary variables. Using this model, it is also established that: 4) The behaviours established in simple trade model are maintained in a more complex model. 5) Long-term outcomes of economic models change as the nature of the population changes. 6) The nature of the economy can change the nature of the population. These last two points suggest that economic models using representative or eternal populations are not suitable for making conclusions concerning long-term economic theory or recommending policy. However, the most important finding is likely that price can be delinked from supply and demand in the long-run. Explanations for this result lie in the results of the other experiments that, together, build up our understanding of the foundational forces of trade in a complex system.

Actions (Repository Editors)

Item Control Page