Ratings and Asset Allocation: An Experimental Analysis

Abstract

Financial product ratings are intended to summarize relevant information in a manner that assists in decision-making. Ratings, however, have the potential to be either helpful or harmful. Ratings are often assigned within categories; ratings across categories then may or may not be comparable. We experimentally assess the effect of ratings in a setting where subjects repeatedly make investment decisions, there is minimal computational burden, and at all times subjects have complete information about the characteristics of the investments. Ratings supply no additional information, but categorized ratings affect investment decisions and harm performance. More knowledgable subjects make better decisions, but are also affected by categorized ratings. The effects are largely reversed when ratings no longer appear.