The measurement, understanding and application of risk aversion differences for pension investments.

The purpose of our project is threefold. First, we review and extend the literature on risk elicitation methods. Our aim is to create a more robust method that generates a measure of risk aversion that depends less on the framing of the questions, background risks and behavioral biases (like, for example, probability weighting or loss
aversion). Second, we study how risk preferences are formed over the lifetime of an individual. Risk preferences vary substantially between individuals (Choi et al., 2007), though it remains unclear where this may originate from. We know that past experiences influence our behavior, as Malmendier and Nagel (2011) show that individuals who lived through the Great Depression choose to take significantly less risk throughout their lives. Following neuroscience, we hypothesize that personal characteristics are developed at specific ages, such that our experiences during those specific periods may influence our future behavior more strongly (Bear et al., 2020). For instance, a divorce influences children well into adulthood (Amato and Sobolewski, 2001). We are therefore particularly interested in the role of personal and macro-economic events on the evolution of risk preferences. Third, we translate our insights on risk preferences to life-cycle investment. By including our insights in the optimization, we prevent, or potentially dampen, the sub-optimal result when the strategy is adapted every five years. Our research agenda builds upon a topic that gained a lot of attention recently, both in academia recently and, due to the new pension reform bill, also in the Dutch pension industry.

his project is funded and carried out under the auspices of Instituut Gak.