Tags: insurance demand, portfolio decision, risk preference, wealth, EGRIE
More from: Olivier Armantier, Jérôme Foncel, Nicolas Treich

The decision to insure against the risk of monetary loss and the decision to invest in risky assets reflect the same, albeit opposite, risk retention tradeoff. Namely, an agent reduces his exposure to risk by purchasing insurance, while he increases his risk exposure by investing. Factors that promote risk taking should therefore lower the demand for insurance and increase the demand for risky assets. In particular, because the wealth elasticity of demand for risky assets has been shown to be positive, insurance coverage should decrease with wealth, making insurance an inferior good. The object of this paper is to explore the link between portfolio and insurance decisions, and in particular to test the hypothesis that wealth has an opposite effect on the two decisions.

To do so, we use survey data for a representative sample of U.S. household heads which combines detailed micro level information on wealth composition, portfolio distribution, insurance coverage, and socio-demographic characteristics. The empirical analysis consists of two steps. In step one, we estimate a baseline, easily interpretable, model focusing on auto insurance coverage and investment decisions. Unlike previous literature, the model controls for key covariates such as the value of the good insured, objective and subjective risks and risk attitude. In step two, we conduct a series of robustness checks by considering different specifications and variable definitions, other forms of insurance (homeowner insurance), and a different sample of industry (i.e. not survey) data from a different country (France).

The empirical analysis produces three main results. First, we find strong evidence that insurance is a normal good. That is, all else equal, and in particular after controlling for risks, risk attitude and the value of the good insured, wealthier respondents are found to purchase more insurance coverage. This in itself is a novel and potentially important result that sheds new light on the insurance industry, one of the largest sectors in the world economy. Second, we identify a puzzle, the “insurance-portfolio puzzle”, in the sense that, contrary to economic intuition, risky assets holding and insurance coverage both increase with wealth. Third, we find several joint determinants of investment and insurance behavior. In particular, the two decisions respond to subjective expectations and risk attitude in a way consistent with theory. We also identify several frictions, including liquidity constraints, financial literacy, or information, that impede both the demand for risky assets and the demand for insurance.

To explain the insurance-portfolio puzzle, we first turn to conventional theory. We show that our results are not consistent with the canonical portfolio/insurance model (Pratt 1964, Mossin 1968). We then enrich the model by considering in particular the possibility that insurance and investment decisions are taken jointly, that losses depend on wealth, or that the agents are liquidity constrained. Next, we explore various behavioral factors including prospect theory, context-dependent preferences, and “peace of mind.” We conclude that conventional and behavioral theories are insufficient to explain the puzzle fully.

This paper contributes to the field of household finance by linking two strands of the literature. First, it relates to the extensive empirical literature on households’ portfolio choices (e.g. stock market participation, portfolio diversification, investment mistakes) and their determinants (e.g. risk attitude, wealth, demographics). Second, it relates to the more recent literature that uses micro-level data to explore insurance choices. So far, this literature has mostly focused on testing for the presence of asymmetric information in insurance markets, and whether risk preferences are stable across contexts. Little is known, however, about the link between households’ portfolio and insurance choices. To the best of our knowledge, this paper is the first to identify determinants and frictions that are common to both decisions. More importantly, we identify a puzzle that calls into question standard theory in which portfolio and insurance decisions are modeled as two sides of the same coin.

The complete paper and more about Professor Nicolas Treich are available at:
http://www.nicolastreich.com/.