Abstract
The "divergence of opinion" hypothesis implies predictable pricing effects in markets where assumptions of homogeneous investor expectations and unrestricted short selling do not hold. Direct tests of the hypothesis do not exist for traditional financial markets - apparently because of the severity of several requirements, including that measurement of divergent ex-ante expectations be unambiguously paired with associated ex-post results. This and other conditions are met in direct tests of the hypothesis in a pricing arena, such as racetrack betting, where divergence of opinion influences are certainly present. Results do not furnish statistically significant support for the divergence of opinion hypothesis.