Search

This paper studies the profile of cooperation emerging in a context in which agents may choose to join one of two groups or stay on their own, in a world where similarity across peers matters. In particular, we investigate the role of heterogeneity in individual contributions, of the level of information and of in-group processes of convergence in values (sense making) in fostering higher levels of cooperation, assessed through higher participation rates to groups. Starting from the result that more heterogeneity reduces participation, we show that increasing the level of information available to subjects and activating sense-making dynamics are able to support higher cooperation levels, which, however, come at the cost of increased radicalization of agent types within group.

We model a continuous double auction with heterogenous agents and compute approximate optimal trading strategies using evolution strategies. Agents privately know their values and costs and have a limited time to transact. We focus on equilibrium strategies that are developed taking into account the number of traders that submitted orders previously, as well as the number of who will submit subsequently. We find that it is optimal to place increasingly aggressive orders, according to a roughly linear schedule, and test the resulting equilibrium for robustness and accuracy.

This paper studies the consequences of removing the resampling assumption from the zero-intelligence trading model in Gode and Sunder (1993). We obtain three results. First, individual rationality is no longer sufficient to attain allocative efficiency in a continuous double auction; hence, the rules of the market matter. Second, the allocative efficiency of the continuous double auction is higher than for other sequential protocols both with or without resampling. Third, compared to zero intelligence, the effect of learning on allocative efficiency is sharply positive without resampling and mildly negative with resampling.

The evaluation of an exchange market is a multi-faceted problem. An important criterion is the ability to achieve allocative efficiency. Gode and Sunder (1993) shows that a continuous double auction for singleunit trades leads to an efficient allocation even when the traders exhibit “zero-intelligence”; in other words, market protocols are active contributors in the search for a better outcome. Under reasonable circumstances, most of the commonly used market protocols share the ability to help traders discover an efficient allocation.

In many applications we are interested in finding clusters of data that share the same properties, like linear shape. We propose a hierarchical clustering procedure that merges groups if they are fitted well by the same linear model. The representative orthogonal model of each cluster is estimated robustly using iterated LQS regressions. We apply the method to two artificial datasets, providing a comparison of results against other non-hierarchical methods that can estimate linear clusters.