Essays on elderly asset management the role of medical expenses and housing.

摘要

With baby-boomers approaching their retirement age, the financial security of elderly Americans has become increasingly crucial for both policy-makers and retirees themselves. Based on the data from the Health and Retirement Survey (HRS), this dissertation examines how healthcare costs and housing affect retirees' saving and investment decisions.;In chapter 1, I investigate the pattern of asset allocation among elderly Americans and test if large out-of-pocket (OOP) medical expenses were preventing older households from holding risky assets. My results show that the elderly are more likely to own risky financial assets while less likely to own housing assets as their OOP medical expenses go up. And conditional on ownership, relatively large OOP medical expenses are significantly correlated with high share of risky financial assets and low share of housing assets. These results imply that large OOP medical expenses may be capturing something other than poor health, which I explain by a "mechanical" effect that older households sell their houses to finance their large OOP medical expenses, driving up the risky financial assets share.;In chapter 2, I develop a life-cycle dynamic programming model that particularly assumes elder homeowners would sell their houses to finance large medical expenses such as nursing home costs. My model finds that retirees with housing are less sensitive to the magnitude of medical expense shocks and thus maintain fewer precautionary savings. Without housing assets, retirees tend to take fewer risks while investing at the early stage of their retirement, but may have to take chances to finance elevated medical expenses as they approach the end of their life cycle.;In chapter 3, I relax the borrowing constraint assumed in the prior two chapters and study the impact of health problems and related expenses on elderly households' financial security. I estimate the probability of medical default among 65 and above age group using various qualitative choice models. Results indicate that high medical debt to income ratio rather than OOP medical expense itself contributes to the high probability of default.