How to Alleviate Correlation Neglect

We experimentally study how presentation formats for return distributions affect investors' diversification choices. We find that sampling returns alleviates correlation neglect and constitutes an effective way to improve financial decisions. When participants get a description of probabilities for outcomes of the joint return distribution, we confirm the common finding that investors neglect the correlation between assets in their diversification choices. However, when participants sample from the joint distribution, they incorporate correlation into choices as predicted by normative theory. Results are robust across three experiments with varying expertise and experience of participants (students vs. investors), and varying return distributions (discrete, continuous)