The Diversification Paradox: How Lay Investors Perceive Risk and Covariance Information

With the development of employee savings plans and privatized retirement savings schemes, people with scant knowledge of finance routinely make risky decisions with dramatic implications for their future wealth. Although people can drastically reduce portfolio risk by investing in negatively correlated assets (Markowitz 1952), empirical evidence demonstrates that they fail to do so, instead investing overwhelmingly in positively correlated assets (e.g. French and Poterba 1991; Huberman 2001). Behavioral research suggests that novice investors diversify poorly because they fail to take into account the way assets covary with each other (Hedesstrom, Svedsater, and Garling 2006). For instance, they evaluate the riskiness of assets one at a time (Dorn and Huberman 2010), or they invest in all available assets in the menu evenly (Benartzi and Thaler 2001). Building upon this literature, we examine the interplay between covariance information, risk perception, and financial literacy. We suggest that investing in positively correlated assets can actually be a deliberate choice made by novice investors, in a failed attempt to reduce risk. Indeed, negatively correlated assets may appear incongruent to novice investors. In non-financial decisions, incongruence is usually associated with a notion of risk (Campbell and Goodstein 2001; Meertens and Lion 2008). While retrieving this misleading incongruence-risk association, novice investors may end up making riskier financial decisions (investing in correlated assets and avoiding negatively correlated assets) when trying to minimize risk, compared to when they are encouraged to take risks. In other words, they accomplish exactly the opposite of what they intend to do. In four studies, we show that this erroneous strategy can be applied in sequential (study 1) and in simultaneous (study 2) financial allocation tasks, and that it is the result of a deliberate and reasoned decision-making process (study 3). We also suggest menu design (study 2) and investor education (study 4) solutions to improve novice investors’ investment decisions.