Modelling a Supply Chain as a Network of Markets

Abstract

The field of Agent-based Computational Economics (ACE) studies the macroeconomic regularities emerging from the local interactions of agents. In this paper, we describe a model of supply chains inspired by ACE. To this end, we present how we have connected sets of trading agents with markets so that any supply chain structure may be simulated. The basis of our model is a representation of a supply chain as a network in which every edge represents a marketplace and every node a set of companies. The companies in every node buy in the same market, and sell in another market. We implement this using a modification of the JAVA Auction Simulator API (JASA) which simulates a single auction. As an illustration of our model, we outline three possible studies based on results from the ACE literature, and preliminary results about the first of these studies are reported. The common objective of these three studies is to find new methods to control a complex system.

A supply chain as a network of auctions

“…With regard to such adaptation of tools, we have already stabilised the prices in a supply chain modeled as in this paper by means of speculation (Moyaux and McBurney, 2006b). We also plan in (Moyaux and McBurney, 2006a) to adapt the methodology proposed by Mizuta et al (2003) to broadcast different price signals in order to stabilise supply chains. However, adapting such tools to supply chains is not the topic of this paper, which focuses on the model and its dynamics.…”

Network Effects in Double Auction Markets with Automated Traders

“…[3] used such methods to examine the effects of linked markets on financial crises, while [18,19] consider the behavior of supply chains. 2 .…”

“…[3] uses agent-based methods to examine the effects of linked markets on financial crises, while [38] looks at the effect of different trade routes on price convergence. [18,19] study the bull-whip effect [14] 1 in supply chains. In addition, some initial results on multiple auctions that compete for traders were presented in [22] and the design of such auctions is the focus of the TAC Market Design competition analyzed in [23].…”