| Summary: | I study how brokers distort consumer investment decisions. The market for retail convertible bonds offers a unique environment to study consumer investment decisions in a broker-intermediated setting. Using a novel data set, I find that consumers frequently purchase dominated bonds in this market---i.e., cheap and expensive versions of otherwise identical bonds exist in the market at the same time. Moreover, inconsistent with standard search models, consumers purchase more of the expensive bonds. The empirical evidence suggests broker incentives are partially responsible for the inferior investments as brokers earn a 1.12% point higher fee for selling the dominated bond. I rationalize the behavior of brokers and consumers in equilibrium by developing and estimating a search model. Consumer search is endogenously directed according to the incentives of brokers and a broker's ability to price discriminate across consumers based on the consumer's level of sophistication. I use the estimated model to disentangle and quantify the importance of search, consumer sophistication, and broker incentives. Aligning broker incentives with those of consumers' increases consumer risk-adjusted returns by 106bps, but does not resolve the primary friction in this market, consumer search. My estimated model allows for investigation of counterfactual scenarios surrounding the Dodd-Frank Act.
|
|---|
