Experimental stock market dynamics: Excess bids, directional learning, and adaptive style-investing in a call-auction with multiple multi-period lived assets

Abstract

We study the behavioral dynamics of limit orders in simultaneous experimental call-auction markets with multiple multi-period lived securities. As analytical decision variable we use excess bids; the number of submitted bids minus the number of offers. The feedback variable is (excess) return. Our results suggest that excess bids are predictive of qualitative asset returns, and that excess bids are formed in an adaptive way. We conclude that the price trend or reversal is reinforced by rejected excess bids and the fundamental laws of demand and supply instigate a regression to the mean. Our analysis of portfolio adjustment dynamics which is based on learning direction theory shows that adaptive value-style investing and path-dependence explain a significant share of individual behavior.