Home Bias or Financial Contagion? Rational Inattention in Experimental Asset Markets

Abstract

This paper studies how rationally inattentive traders allocate information across assets and how those choices shape market outcomes. We design a repeated two-asset call-market experiment in which subjects first choose the variances of asset-specific signals and then trade both assets in a uniform-price market. The treatments compare two information technologies: an entropy cost, which favors deepening information about selected assets, and a linear precision cost, which makes high-variance assets more attractive to learn about. Halfway through the market rounds, Asset B receives a permanent prior-variance shock. Pre-shock outcomes are balanced across treatments. After the shock, treatment-specific capacity shares shift away from Asset B and toward Asset A in both treatments, with a substantially larger and statistically significant reallocation under the linear cost. Signal and posterior variances also rise sharply for Asset B in both treatments, although the linear cost attenuates this deterioration and modestly improves precision for Asset A. Market outcomes point in the same direction: pricing errors, raw-price dispersion, and pricing-error volatility rise mainly for Asset B, while we find little evidence of the predicted price declines or of pricing-error contagion to Asset A. Individual order regressions show that lower information precision is associated with more conservative bids, wider bid-ask spreads, lower bid quantities, and lower final holdings.

Keywords: Rational Inattention; Information Acquisition; Experimental Asset Markets; Financial Contagion; Home Bias

JEL Classification: C92; D83; G14; D53; F30