Testing market efficiency with information on individual investor performance

Abstract

The efficient market hypothesis has been tested under a variety of settings. Several authors have utilized stock market data while others have used betting markets including National Football League games and horse racing. These betting markets are similar to stock markets in terms of their large amounts of information and ease of tracking results. This paper analyzes the efficient market hypothesis for a quite different, nonbetting, racetrack market known as claiming races. The nature of the market and the available data allow individual investor performance to be followed over time. While data limitations have restricted previous authors to ex-post determination of whether particular advantageous decision rules could have been utilized, our access to individual data enables us to analyze whether individuals were able to determine decision paths that resulted in superior performance. Our analysis indicates that the claiming race market is efficient in the sense that active participants do not systematically outperform the market average.