We present a new multiagent model for the multiperiod portfolio selection p
roblem. A system of cooperative agents divide initial wealth and follow ind
ividual worst-case optimal investment strategies from random portfolios, sh
aring their final profits and losses. The multiagent system achieves better
average-case performance than a single agent with the same initial wealth
in a simple stochastic market. A further increase in performance is achieve
d through communication of hints between agents and probabilistic strategy-
switching. However, this explicit cooperation is redundant in a market that
approximates the Capital Asset pricing Model, a model of equilibrium stock
price dynamics. (C) 2001 Academic Press.