Learning in Continuous Double Auction Market

Summary

We start from the fact, that individual behaviour is always mediated by social relations. A heuristic is not good or bad, rational or irrational, but only relative to an institutional environment. Thus for a given environment, the Continuous Double Action (CDA) market, we examine the performance of alternative intelligent agents, in terms of market efficiency and individual surplus.

In CDA markets traders face three non-trivial decisions: How much should they bid or ask for their own tokens? When should they place a bid or an ask? And when should they accept an outstanding order of some other trader?

Artificially intelligent traders have been used to explore the properties of the CDA market. But, in all previous works, agents have a fixed bidding strategy during the auction. In our simulations we allow the soft-agents to learn not only about how much they should bid or ask, but also about possible switching between the alternative strategies. We examine the emergence or not of Nash equilibriums, with a bottom-up approach.

Our results confirm that although market efficiency is an ecological property, an it is robust against intelligence agents, convergence and volatility depend on the learning strategy. Furthermore our results are at odds with the results obtained from a top-down approach, which claimed the existence of Nash equilibriums.