Price and Quantity Competition Markets: An Ecological Approach
speaker: Remì Dorat (University of Lille)abstract: Our work deals with market convergence. We study a market for a homogeneous good. At each market period, agents pre-produce the quantity they are going to offer and fix a price. All transactions take place at the posted prices whitch remain fixed during one period. Demand is fully rational (no withdrawing is possible), is down sloping and convex. The agents adapt with regards to the results of their last offers. Several analytical studies have shown that there is no pure strategy Nash equilibrium for this market (Alger (1979), Maskin (1986) et Friedman (1988)). The epistemological void thus created was deepened by several experiments (Kruse (1993), Davis and Holt (1994), Kruse et alii (1994), Wilson (1998), Brandts and Guillen (2007), and Puzzello (2007)). They showed that there is no clear convergence path to any given issue (be it competitive, cartelistic, or Edgeworth-cycle). All these experiments suggest that all possible issues can indeed be observed in experiments. Another result put forward in these experiments is that the mixed-strategy equilibrium that might exist in the analytical game has no predictive power. It seems that behavior undertaken by experimental subjects can explain this diversity. But if evidence is clear that experimental subjects’ behavior departs from neo-classical assumptions of full rationality, the question remains of what would be the market issue if agents were indeed homo oeconomicus? How are sellers supposed to act in a market in which there is no Nash equilibrium? We propose here 12 different types of behavior, all funded on the profit –maximising principle but modelled in order to take into consideration the 3 paradigms of economic behavior available. The first paradigm of behavior is the price-taking behavior. In this perspective an agent believes that there exists indeed a “market price” (the reference of which can be either the mean price observed at the previous period, the competitive price or his own price posted at the previous period) to which she adapts by equalizing her marginal cost. The second paradigm that inspired us was the “strategic behavior” which goes back to Edgeworth. In this perspective agents try to undercut their opponents. However, as soon as they reach the bottom line (corresponding to the competitive equilibrium), agents have a strong incentive to post the monopolistic price and a new “war price” commences. Finally, the last paradigm is the “cooperative” one: agents realize that the monopolistic issue represents a Pareto optimal issue from their point of view. They look forward to the cooperative issue, try to reach it but protect from agressive behavior. The common characteristic of these agents is that they will only produce, at any given period, only the quantity demanded (using the fractional demand). These behaviors are therefore supposed to be very vulnerable to more aggressive attacks from their opponents. The systematic study of all possible tournaments (using either two or three strategies at a time on a market) show that indeed all possible issues can be obtained. The types of behavior agents adopt has a strong impact on the type of issue which result from their interaction. What these tournaments also show is that none of the modelled behavior dominates the other. This is what makes the ecological simulation more interesting.
timetable:
Sat 3 Oct, 17:45 - 18:30, Aula 3
documents:
dorat
<< Go back