The Pensions, Retirement and Ageing Seminar Series is jointly hosted by CEPAR and the School of Risk & Actuarial Studies at UNSW Sydney.
It takes place on a Monday (or Wednesday) from 12-1pm and provides an excellent opportunity to network with pensions and superannuation experts from Australia and overseas. This is an interdisciplinary group with backgrounds in economics, actuarial studies, finance, psychology, law, accounting, demography, marketing, medicine and related fields. We invite attendance of participants from other universities as well as from industry and government who are interested in both theoretical and applied research on pensions, retirement and ageing.
We also welcome presenters who are early career academics or practitioners. There is no charge to attend these seminars.
In response to the current COVID-19 situation, upcoming seminars are being held online as webinars. All seminar registrants will be notified via email. Please contact Inka Eberhardt if you are interested in presenting, participating or would like to be added to the Pensions, Retirement and Ageing Seminar Series mailing list.
We wish to express our best wishes to the community during this challenging time.
11 March - Geoff Warren (ANU College of Business and Economics) "The ‘Right’ Level for the Superannuation Guarantee: A Straightforward Issue by No Means"
(Location: UNSW Kensington Campus, in Quadrangle Building, Room 2063)
6 April - Katja Hanewald (CEPAR, UNSW School of Risk and Actuarial Studies) "Long-term Care Insurance Financing using Home Equity Release: Evidence from an Experimental Study"
20 April - Ioana Ramia (Centre for Social Impact, UNSW Sydney) and Mălina Voicu (Research Institute for the Quality of Life, Romanina Academy) "Life Satisfaction and Happiness among Older Europeans: The Role of Active Ageing"
4 May - Brooke Brady (UNSW Ageing Futures Institute, CEPAR) "The Development of a New App-based Assessment of Shared Financial Decision Making"
1 June - Xiao Xu (CEPAR, UNSW School of Risk and Actuarial Studies) "Deep Reinforcement Learning for Variable Annuities Hedging"
13 July - George Kudrna (CEPAR, UNSW Sydney) "Wealth and Homeownership in Germany and Australia: The Role of Tax and Retirement Income Policy"
27 July - Craig Sinclair (CEPAR, NeuRA, UNSW Sydney) "The Role of Executive Functions and Reward Sensitivity among Sub-Types of Older Adults with Deliberative and Impulsive Decision-Making Styles under Explicit Risk Conditions"
10 August - Vas Yiengprugsawan (CEPAR, UNSW Sydney) "Pain and its Impact on Functional Health: 7-year Longitudinal Findings among Middle-Aged and Older Adults in Indonesia"
31 August - Daniel Wheadon (CEPAR, UNSW Sydney) "Effects of Means Testing the Age Pension on Populations with Self Control Preferences"
14 September - Elise Payzan-Le Nestour (School of Banking & Finance, UNSW Sydney) "How Our Brains Blind Us to 'Black Swan' Economic Events"
12 October - Arvid Hoffmann (University of Adelaide) "The Financial Vulnerability Trap: Using Latent Transition Analysis to Explore the Dynamics of Consumers’ Financial Vulnerability over Time"
Speaker: Piet de Jong (Department of Applied Finance and Actuarial Studies, Macquarie University)
Topic: Lockboxes and Glide Paths
If you are an average Australian you will immediately spend your super when you retire, go on the age pension, and, if aged care is needed, get the government to pony up. With more old people, government age support spending is likely to rocket. Standing idly by is a bloated and inefficient super "helpers" industry, exploiting a captive, ill-informed customer base, corroding retirement savings with fees, and adding no value. I you want to stand on your own retired feet, you won’t find the one product you’ll likely crave: fairly priced pensions. This submission proposes: 1) employees pay for their future age pension and aged care while employed, saving the government an estimated $30 Bn pa; 2) the winding down of the super industry saving an estimated $1,500 pa per employee and doubling retirement incomes; 3) the super system to be properly choreographed so that retirees can buy fairly priced pensions.
Speaker: Geoff Warren (ANU College of Business and Economics)
Topic: The ‘Right’ Level for the Superannuation Guarantee: A Straightforward Issue by No Means
We deploy a stochastic life-cycle model to examine how differing levels of the superannuation guarantee (SG) impact on the welfare of individual Australians under existing superannuation, tax and pension eligibility rules. Our main focus is the effect of various assumptions on the optimal SG, emphasising the role of income and the retirement objectives of the individual. The analysis supports estimating the gains and losses from changing the SG for various individuals, and associated impacts on net government revenue. We find the optimal SG to vary substantially with income and objectives. While our baseline analysis indicates a SG of below the current level of 9.5%, higher estimates emerge if access to the Age Pension is excluded, and if the SG is used as a mechanism to self-insure against living to a very old age, being forced into early retirement, or incurring lower investment returns. We conclude that the case for raising the SG above 9.5% depends on the underlying assumptions, with the policy objectives that the SG is intended to achieve being critical.
Speaker: Katja Hanewald (CEPAR, UNSW School of Actuarial Studies and Risk)
Topic: Long-term Care Insurance Financing using Home Equity Release: Evidence from an Experimental Study
Abstract: Population ageing is a global trend and many countries including China face increasing pressures to provide long-term care services for the elderly. We explore new mechanisms to fund long-term care using housing wealth. We conduct and analyze an experimental online survey fielded in China that assesses the potential demand for new financial products that allow individuals to access their housing wealth to buy long-term care insurance. We find in our sample of 1,200 Chinese homeowners aged 45-64 that the stated demand for long-term care insurance increases when individuals can use housing wealth in addition to savings to purchase long-term care insurance. Individuals prefer to access housing wealth via reverse mortgage loans rather than via home reversion, which is a partial sale of housing wealth. Our results inform current policy reforms in China which aim at developing the private market for health and long-term care insurance products.
Speakers: Ioana Ramia (Centre for Social Impact, UNSW) and Mălina Voicu (Research Institute for the Quality of Life, Romanina Academy)
Topic: Life satisfaction and happiness among older Europeans: the role of active ageing
Abstract: The older population is growing globally, and more so in some European countries. Aimed at enhancing the quality of life of older people, active ageing has been on the policy agenda in Europe since the beginning of the 21st century. Using a subsample of the European Quality of Life Survey consisting of individuals aged 65 and over living in 27 European countries we explore the effect of active ageing on subjective quality of life. The central argument of the paper is that active ageing is cumulative, consisting of a mix of various interconnected activities. Hence, when assessing the impact of active ageing on quality of life we include the whole collection of activities in which seniors engage and avoid limiting to a single activity. Latent Class Analysis is employed to find the mix of interconnected activities in which older adults engage. We identify three classes: home keepers (mainly engaging in housekeeping activities), carers (mainly engaged in caring, but also some housekeeping activities) and those engaged outside their homes (engaged primarily in paid or unpaid work). Multilevel regression models test the connection between the different strategies to remain active in later life on life satisfaction and happiness, the cognitive and affective indicators of subjective quality of life. Our results show that remaining active in later life does not always lead to improvements in subjective quality of life and that separate strategies to remain active in later life are at work to increase life satisfaction and happiness in later stages of life.
Speaker: Brooke Brady (UNSW Ageing Futures Institute, CEPAR)
Topic: The Development of a New App-based Assessment of Shared Financial Decision Making
Abstract: Older adults are increasingly expected to make complex financial decisions, to promote their current and future wellbeing. Research has shown that older adults have lower financial literacy, are less likely to seek financial advice, and may be less discriminating in the type of advice they use. Very little is currently understood about the family and social contexts in which older adults make financial decisions, including how and when these decisions are shared. This project aims to develop and pilot test a new smartphone application. The shared decision making tool at the heart of this app includes a bank of hypothetical ‘real-world’ financial decision scenarios, designed to enable researchers to measure the dynamics of shared financial decision-making among older adults. This app aims to collect novel data to address existing research gaps, and has the potential to be used for implementing future interventions aimed at promoting financial wellbeing among older adults.
Speaker: Nicolás Salamanca (Melbourne Institute: Applied Economics & Social Research, University of Melbourne)
Topic: How People React to Pension Risk
Abstract: We show that people exposed to greater pension risk are less likely to invest in risky assets. We exploit a reform that links people’s future pension benefits to their pension funds’ funding ratio—a measure of the fund’s financial health—making funding ratios a fund-specific measure of pension risk. The effect of pension risk is stronger for people who are better informed about their pensions, for retirees and pension-age non-retirees, and for wealthier people. The funding ratio does not affect investments in a pre-reform period, nor does it affect bequest intentions, (expected) retirement, or the motivations for saving.
Speaker: Xiao Xu (CEPAR, UNSW School of Risk and Actuarial Studies)
Topic: Deep Reinforcement Learning for Variable Annuities Hedging
Abstract: Variable annuities with guaranteed minimum benefits (VA+GMBs) have become increasingly popular in retirement planning, providing policyholders with life-contingent income and protection against the downside risk of equity markets. The US market for VA products was more than 2 trillion US Dollars by the year-end of 2017. For life insurers offering these products, a challenging issue is developing more efficient pricing and hedging of their VA+GMBs portfolios, something also of interest to insurance regulators. With the artificial intelligence (AI) revolution, deep learning has been applied increasingly applied to classical problems in finance, including portfolio management. This research investigates hedging VA+GMBs contracts with an efficient model-free deep learning algorithm. The deep hedging is performed by implementing a four-layered recurrent neural network (RNN) with long short-term memory (LSTM) units to optimise a convex risk measure – conditional value at risk (CVaR) at 95%. The deep hedging framework is applied to hedge and obtain comparable risk measures for a range of equity model assumptions, including the Black-Scholes, stochastic volatility, Levy and time-changed Levy process models. Compared to the traditional model-based delta hedging strategies, the deep hedging algorithm provides significant improvements in computation and speed. The results also confirm that the resulting solvency measure, value at risk (VaR) at 99.5%, for equity models with jumps and stochastic volatilities are significantly higher than those implied by the Black-Scholes framework. As the 99.5% VaR is the suggested measure adopted in the Solvency II framework, using VaR with commonly used Black-Scholes assumptions will significantly underestimate the required capital in a more volatile market for insurance providers. For both life insurers providing VAs and insurance regulators, equity model assumptions are found to be critical, including in the model-free deep learning environment.
Speaker: George Kudrna (CEPAR, UNSW Sydney)
Topic: Wealth and Homeownership in Germany and Australia: The Role of Tax and Retirement Income Policy
Abstract: Although Germans and Australians have very similar incomes per capita, their wealth holdings and wealth structures are very different. In Germany average household wealth in 2018 amounted to roughly 420.000€, while the respective figure in Australia was more than 50 percent higher. Not surprisingly, only about 44 percent of German households own their place of residence, while the homeownership rate in Australia amounts to almost 70 percent. The question is to what extent these gaps in wealth and homeownership patterns are due to differences in public policies. The (fairly low) taxation of capital income in Germany hardly distinguishes between savings motives, while in Australia old-age savings are taxed at significantly lower rates. Even more important, Germany operates a pay-as-you-go financed public pension system, where benefits are tightly linked to former contributions, which absorbs almost 12 percent of GDP. In contrast, Australia combines tax-financed and means-tested old-age provisions with a funded, so-called superannuation system financed by mandatory contributions. Interestingly, means-tested assets do not include real estate which provides a clear incentive for homeownership in old-age. At the same time about 20 percent of the wealth of Australians has been already accumulated in the still maturing superannuation system.
In order to quantify to what extent the tax and pension design can explain the observed differences in asset levels and homeownership, we apply an overlapping generations model with tenure choice where households face labour income and lifespan uncertainty. The model is calibrated to Germany featuring pension benefits based on individual earnings points accumulated during the working phase and a fairly low taxation of income from capital. Then the Australian tax and pension structures are implemented sequentially in order to distinguish the impact of capital taxation, means-testing and funding. Our simulation results indicate that the Australian tax and pension design has a dramatic impact on asset levels and structures explaining more than two thirds of the above differentials in asset levels and homeownership rates. While capital taxation and means-testing shift the asset structures towards residential properties, the superannuation system increases the overall wealth level.
Speaker: Craig Sinclair (CEPAR, UNSW and NeuRA, UNSW Sydney)
Topic: The Role of Executive Functions and Reward Sensitivity among Sub-Types of Older Adults with Deliberative and Impulsive Decision-Making Styles under Explicit Risk Conditions
Abstract: Older adults are increasingly faced with complex, high-stakes decisions about healthcare, finances and lifestyle. These decisions draw on a constellation of mental processes. Age-related cognitive decline can impact on decision-making, however the trajectories of cognitive decline among older adults are diverse, and can be offset through selective optimization and compensation (Baltes & Baltes, 1990). Early identification of older adults experiencing declines in decision-making may enable supportive interventions.
The ‘Game of Dice’ task (GDT) is a behavioural decision-making task in which participants make repeated gambles under conditions of ‘explicit risk’ with explicit and stable outcome probabilities and reward structures across trials. Performance on this task discriminates between those with clinical conditions (e.g. dementia, Parkinson’s disease, bipolar disorder) and healthy adults across the lifespan. Previous studies have shown the role of executive functions, short-term memory and reward sensitivity, and have observed weak negative associations between age and decision-making performance. The current study is the first large-scale study of older adult performance on the GDT. One thousand and two older adults without cognitive impairment (age 72-79 years) completed the GDT during Wave 4 of the Personality and Total Health (PATH) Through Life study. K-means cluster analysis identified three clusters of older adults, reflecting ‘deliberative’ (more advantageous), ‘heuristic’ and ‘impulsive’ (less advantageous) decision-making tendencies. Learning effects were observed among the deliberative and heuristic groups, however performance among the impulsive group declined over time. Multinomial logistic regression indicated that membership in the ‘deliberative’ group was associated with younger age, male gender and better performance on a measure of short-term memory. Membership in the ‘impulsive’ group was also associated with poorer performance on a test of executive-function. Reward sensitivity showed only inconsistent associations with decision-making performance.
Trajectories of cognition and decision-making performance among older adults without cognitive impairment are complex. It appears that the GDT can identify distinctive sub-types of decision-making tendencies among healthy older adults, and that these sub-types are associated with measures of cognitive function. Future research would benefit from longitudinal studies of GDT performance.
The authors acknowledge funding from the ARC Centre of Excellence in Population Ageing Research.
Speaker: Vas Yiengprugsawan (CEPAR, UNSW Sydney)
Topic: Pain and its Impact on Functional Health: 7-year Longitudinal Findings among Middle-Aged and Older Adults in Indonesia
Abstract: Pain is a growing public health issue worldwide, but there is limited population-based evidence in low- and middle-income country settings. Using nationwide Indonesian Family Life Survey (IFLS) data in 2007 and 2014, this research sets out to investigate the associations between changes in pain status between two time points and its impact on functional health outcomes among middle-aged and older adults in Indonesia. Analyses focused on 7936 adults aged 50 years and older in 2014 who responded to both waves. Functional health was assessed using a composite score of functional limitations (range 20-100), representing difficulty in performing activities of daily living, and grip strength (kilograms). Multivariate linear regression models were used to analyse associations between pain measured in 2007 and 2014 and functional health in 2014. Severe pain in the latest wave of IFLS was associated with older age, female, lower education, having chronic conditions or depressive symptoms. Notably, those who reported 'low-medium' pain in 2007 and 'severe' pain in 2014 belonged to the most vulnerable group with worst functional health outcomes (4.96 points higher limitation scores and 1.17 kg weaker average grip strength). Findings have implications for public health policy in monitoring and management of pain including related co-morbidities as an increasingly critical component of population ageing.
Speaker: Daniel Wheadon (CEPAR, UNSW Sydney)
Topic: Effects of Means Testing the Age Pension on Populations with Self Control Preferences
Abstract: Means testing a public age pension is often justified on the ground that it can target the pension to agents most in need of financial support in old age, reducing the tax burden enough to outweigh the distortions to the agent’s work and savings decisions. This is generally borne out in models where agents choose their allocations between labour and leisure and between saving and consumption each period to optimise utility over their lifetime. However, if agents are myopic then the cost saving from means testing the pension is likely to be less than it would otherwise be, and the distortions may be greater. In this paper I develop an overlapping generations model of a small open economy to consider the effect of means testing the pension on household behaviour when agents have Gul-Pesendorfer style self-control preferences and must exercise costly self-control to resist temptation to consume their savings each period. I consider different pension taper rates and different levels of the cost of self-control, and find that when households have higher self-control costs they will save and consume less over their lifetimes but work more. Higher self-control costs are also associated with higher means-tested pension costs. I also find populations with higher self-control costs prefer lower taper rates on the pension means test, and a universal pension may be preferred if self-control costs are sufficiently high.
Speaker: Elise Payzan-Le Nestour (School of Banking & Finance, UNSW Sydney)
Topic: How our Brains Blind Us to 'Black Swan' Economic Events
Abstract: In this talk, Scientia A/Prof Payzan-LeNestour will explain how neurobiological factors blind people to making sound investment decisions when markets are hit by large macroeconomic shocks such as COVID-19. People tend to initially misinterpret the nature of major unpredictable events (known as black swan events). They tend to grossly underestimate their importance before eventually readjusting their view – though by which time it may be too late. This phenomenon ("outlier blindness") reflects the fact that the human brain is not attuned to properly perceiving unexpected events; rather, it is optimized to focus on things that we expect to frequently encounter in our environment.
Scientia A/Prof Payzan-LeNestour will report on recent experimental evidence for outlier blindness, and will discuss important implications for economic decision-making.
Speaker: Arvid Hoffmann (University of Adelaide)
Topic: The Financial Vulnerability Trap: Using Latent Transition Analysis to Explore the Dynamics of Consumers’ Financial Vulnerability over Time
Abstract: An important question regarding financial vulnerability is not just what makes consumers more or less vulnerable at a particular point in time, but also what drives changes in their vulnerability over time. We explore this question through a Latent Transition Analysis which assesses how individual psychological characteristics predict membership of and transition between states of higher vs. lower financial vulnerability over a three-month period across a nationally representative sample of U.S. consumers. We find that consumers in a state of lower vulnerability are “fragile” in having a relatively high likelihood of moving to a state of higher vulnerability, while those in a state of higher vulnerability are “entrenched” in having a relatively low likelihood of moving to a state of lower vulnerability. We call this pattern of results the “financial vulnerability trap”. We also find that while financial self-efficacy explains state membership, the consideration of future consequences drives state transitions.
Direct enquiries to: Inka Eberhardt