“Fair go” for all? Wealth and Risk Aversion of Australian Households*

Abstract

A mean‐variance framework is applied to Australian household financial portfolios in order to provide estimates of relative risk aversion in the economy. Controlling for various socio‐economic characteristics, we explore whether risk aversion heterogeneity is a function of wealth heterogeneity. In contrast to most studies, we find evidence of very high risk aversion amongst the majority of households of poor households but vastly lower risk aversion amongst the high percentiles in the wealth distribution. Applying a first differences model across three survey waves spanning 2002 to 2010, we find that risk tolerance increases significantly with wealth. Risk tolerance is positively associated with mortgage payments, but rental payments have no relationship. In addition, we found no evidence that holding a university education has any discernible impact on risk aversion. Lastly, we present some preliminary findings as to the impact of financial advice on observed risk aversion. Financial advice is found to accentuating risk aversion, particularly amongst the wealthiest households. The findings have potential implications for the distribution of wealth in Australia that has received renewed interest recently.

Risk attitudes and demand for insurance: micro evidence from Ghana

“…Existing literature has documented the influence of certain factors such as gender, wealth and residential location on risk attitude (Charness and Gneezy, 2012; Dohmen et al. , 2011; Tsigos and Daly, 2016). The association of these factors with risk attitude could significantly impact the effect of risk attitude on financial choices particularly, insurance decision.…”

Risk Aversion and the Elasticity of Intertemporal Substitution among Australian Households

“…In Australia, Tsigos and Daly (2016) test the relative risk aversion assumption using HILDA wealth data. Using a different measure of relative risk aversion (based on observed portfolio weights for each household), Tsigos and Daly find a negative correlation between risk aversion and wealth.…”

“…While CRRA preferences are a convenient assumption, they impose a restriction on the relationship between risk aversion and wealth, namely that relative risk aversion is constant with respect to a household's level of wealth. In fact, the empirical support for this form of preferences is not overwhelming, in part because there are relatively few studies that have formally tested the CRRA assumption., Chiappori and Paiella (2011) and Tsigos and Daly (2016) Therefore, we undertake an initial test of whether the CRRA assumption is in fact a good one for Australian households, using an approach based on a household's share of risky assets, similar to that in Chiappori and Paiella (2011) and Tsigos and Daly (2016) (the latter being a recent Australian study). Once we account for endogeneity in risk aversion and wealth as well as measurement error (which we show is crucial in such regressions), we cannot reject the CRRA assumption.…”

Compensation Convexity without Utility Restriction

“…More relevantly, Rashad Abdel‐Khalik () develops an index to measure CEO risk tolerance using publicly available databases ExecuComp and Compustat on WRDS. Tsigos and Daly () provide estimates of risk aversion of Australian households. Thus extant studies provide pertinent references and instruments for testing our theory.…”