Abstract
We study the portfolio allocation decisions of Australian households using the relatively new Household Income and Labour Dynamics in Australia (HILDA) survey. We focus on household allocations to risky financial assets. Our empirical analysis considers a range of hypothesised determinants of these allocations. We find background risk factors posed by labour income uncertainty and health risk are important. Credit constraints and observed risk preferences play the expected role. A positive age gradient is identified for risky asset holdings and home-ownership is associated with greater risky asset holdings. A unifying theme for many of our empirical findings is the important role played by financial awareness and knowledge in determining risky asset holdings. Many non-stockholding households appear to lack the experience and financial literacy that might enable them to benefit from direct investment in stocks.
Highlights
Household portfolio allocation is a simple business according to the stylized classical model of portfolio allocation dating back as far as Markowitz (1952)
A growing body of empirical and theoretical research into household financial decisions seeks to develop models that explain and predict observed portfolios or empirically identify factors explaining household portfolio allocations or some combination of the two; see Campbell (2006) for a discussion of this literature. We add to this literature by considering the portfolio allocation decisions of Australian households, using data collected by the Household, Income and Labour Dynamics in Australia (HILDA) survey
Of the various background risk factors that have been considered in the literature, there is (i) a significant negative effect of labour income risk on the risky asset ratio; (ii) a Approximately 90% of employees have retirement account contributions made by employers (ABS, 2006)
Summary
Household portfolio allocation is a simple business according to the stylized classical model of portfolio allocation dating back as far as Markowitz (1952). Neither of the dummy variables for health status is significant in explaining the household’s risky asset ratio in our core specification. This is at odds with findings of previous researchers, such as Guiso et al (1996) and Rosen and Wu (2005). Model 2, where we exclude the risk and time preference variables, a stronger ordering by health status does appear – the coefficient on the ‘poor health’ dummy is still not significant. (which exceeds the mortgage ratio for over 95% of households) implies an increase in the risky asset ratio of 0.004 relative to a household with a mortgage ratio of zero
Sign-in/Register to view highlights & summary 
Submitted Version (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call