Abstract:
The thesis comprises three studies on household finance and retirement income planning. The
first study with a specific focus on South Korea examines whether individuals should defer
claiming old-age pension benefits, and if so, whether it is beneficial to take out a reverse
mortgage loan (RML) to fund the income gap during deferral years. I use the life-cycle portfolio
choice setting to test these retirement income strategies. The results show that incentivising
RMLs as a tool to delay pension benefits is welfare-enhancing for the average consumer. In
addition, cash-poor investors are found to experience a welfare loss if using the distribution
from individual retirement accounts as a sole source to fund the pension delay. Meanwhile,
individuals who get the most out of RMLs are those who have a low level of required living
standards and an initial money balance that is small but enough to survive the first five years
of retirement.
The second study with a Singapore context sets out to explore the influence of unemploymentinduced
career breaks on retirement income adequacy under fully-funded defined-contribution
pension plans, taking into consideration (i) the timing of the breaks, (ii) the duration of the
breaks, (iii) and the severity of the impact that job losses might have on post-interruption
wages. I find that due to the scarring effect of job loss on post-interruption wage profiles, the
sooner the career break, the larger its impact on pension adequacy. Moreover, the results point
out that median-income workers, who suffer job displacement during the career path, can still
support themselves beyond subsistence in retirement. However, job losses do significantly
decrease the retirement nest egg of displaced workers. I also discover that the duration of career
breaks has a significant impact on the pension adequacy of displaced workers.
The third study with a U.S focus investigates how self-perceived health status interacts with
the portfolio choice of homeowners and renters over their life cycle and how this relationship
differs between homeowners and renters. I first develop a dynamic optimisation life-cycle
model for the joint determination of health investment, housing, and portfolio allocation; and
then perform empirical analysis, using data from the Panel Study of Income Dynamics (PSID),
to examine whether the empirical findings support the life-cycle model predictions. The results
indicate that the life-cycle model can explain key facts about stockholding profiles of
homeowners and renters across different levels of health status in PSID. The proportion of total
wealth in stocks is low overall and is positively related to health for both homeowners and
renters. However, I find that renters’ share of total wealth in stocks responds more strongly to
health status deterioration.