CEPAR is supported by
   
Skip Navigation Links Home RESEARCH Integrating public and private provision of products and services
Search CEPAR
 
RESEARCH PROJECTS
Filter by
ACT strategic plan for positive ageing 2010-2014: Towards an age-friendly city

Kaarin Anstey, Kerry Sergeant-Cox, Trish Jacomb

In June, 2010, Canberra was accepted into the World Health Organisation's (WHO) Global Network of Age-Friendly Cities (AFC), a group of municipalities that seek to improve the living experience of its senior residents.  The WHO Checklist of Essential Features of Age friendly Cities includes  Outdoor spaces and buildings, transportation, housing, social participation, respect and social inclusion, civic participation and employment, communication and information, community and health services.  This project surveyed ACT residents on the features of AFCs to identify positive and negative examples of infrastructure and practices and to elicit suggestions for improvement.  A questionnaire was developed that measures key indicators and will measure long term change.

Commencement:  2011

Completion: 

 

Hazel Bateman, John Piggott

This project documents developments in public sector pensions in Australia, and reports estimated unfunded liabilities associated with benefits promised to public sector employees. Australia's experience with public sector pensions is unusual - currently, the defence forces and the judiciary apart, all new entrants to public sector schemes confront defined contribution (DC) plans. The transition from defined benefit (DB) to DC has taken place over the last 20 years. While legacy costs are considerable, still amounting to some 15% of GDP, they are projected to shrink through time. We argue that one driver of this progressive public sector pension reform has been the introduction of the Superannuation Guarantee, a mandatory funded DC pension plan initially introduced in the late 80s.

Commencement:  2011

Completion:   

 
Developing equity release markets: Risk analysis for reverse mortgage and home reversion products

Daniel Alai, Hua Chen, Katja Hanewald, Michael Sherris, Joshua Teo

Equity release products are sorely needed in an ageing population with high levels of home ownership. We investigate the cash flows and risk profile of the provider for the two most popular equity release products, namely, the traditional reverse mortgage contract and the home reversion contract.

Economic scenarios are produced using a vector-autoregressive model for the house prices, rental yields and interest rates in real terms. A multi-state Markov model is employed to capture contract termination probabilities arising from death, entry into long-term care, and refinancing. The capital requirements and profitability of the two products are compared.

Commencement:  2011

Completion: 

 
Economic analysis of China’s new rural pension scheme

Lu Bei, John Piggott

China is currently implementing the world's largest pension plan, measured in terms of membership. Over the next several years, provinces in China will enrol hundreds of millions of the rural population in a plan that will provide at least some financial relief in retirement. This project, focused initially on Zhejiang Province, one of the richest in China and the first to provide full coverage to rural retirees under the scheme, will analyse the impact of the new scheme by combining several new data sources. An early question will be whether rural workers located in urban areas, and currently providing financial support to their parents, will adjust their behaviour. A second and derivative research topic will analyse the impact of the new pension on rural poverty and income distribution.

Commencement:  2011

Completion: 

 
Individual mortality modelling using longitudinal data

Ramona Meyricke, Michael Sherris

Modelling the mortality of subsets of the population is critical to risk management and pricing for annuity providers. Heterogeneity across the mortality of individuals within a population results in discrepancies between actual mortality experience and that predicted from population life tables. So it is important to understand the relationship between population-level and individual-level mortality, and the impact that heterogeneity has on mortality risk. Using individual-level survival data from the US Health and Retirement Survey, we analyze the impact that aggregating across individuals within an age group has on the model specification of survival models.  Individual and population-averaged mortality models are fit and compared in terms of goodness of fit and parameter stability. The individual-level models allows us to go beyond previous work and to quantify the impact of heterogeneity by i) specifying the distribution of idiosyncratic mortality  ii) decomposing variance of the mortality rate at different ages into systematic and idiosyncratic components and iii) comparing the total mortality predicted by aggregating deaths from the population averaged model versus the individual-level model. In 2012 this research was presented at the 16th International Congress on Insurance Mathematics and Economics.

Commencement:  2012

Completion: December 2012

 
Individual post-retirement longevity risk management under systematic mortality risk

Katja Hanewald, John Piggott, Michael Sherris

This project focuses on an individual's post retirement longevity risk management strategy allowing for systematic longevity risk, recent product innovations, and product loadings. A complete-markets discrete state model and multi-period simulations of portfolio strategies are used to assess individual longevity insurance product portfolios with differing levels of systematic and idiosyncratic longevity risk. Portfolios include: fixed life annuities, deferred annuities, inflation-indexed annuities, phased withdrawals and recently proposed group self-annuitization (GSA) plans. Results so far: GSA plans are found to replace even inflation-indexed annuity products when there are loadings on guaranteed life annuity products. With a bequest motive and loadings, coinsurance portfolio strategies with phased withdrawals and GSA's dominate portfolios with life annuities or deferred annuities.

Commencement:  2011

Completion:   

 
Lifetime dependence modelling using the truncated multivariate gamma distribution

Daniel Alai, Zinoviy Landsman, Michael Sherris

We investigate the valuation of a portfolio, or cohort, of annuitants with similar risk characteristics. We jointly model the lifetimes using a truncated multivariate gamma distribution that induces dependence through a shared gamma component. Model parameter estimation is rooted in the method of moments. Once model calibration is achieved, we obtain joint survival probabilities and quantify the level of dependence within the group of lives.  We value the portfolio of annuities, as well as assess its riskiness. Finally, we compare our results against the assumption of independent lives.

Commencement:  2011

Completion: 

 
Longevity analyses based on Norwegian administrative records

Daniel Alai, Erik Hernaes, Michael Sherris

We perform a longitudinal analysis using administrative records that contain annual data for all individuals resident in Norway at any time during years 1994-2005. The data includes demographic factors and well as covariates related to education, employment, income and wealth. Using event history software, developed at the Frisch Institute, we can analyse the panel data in a flexible way whilst accounting for a large number of covariates. Similar analysis has been successfully performed to study the duration of unemployment, and it is well suited to study longevity as well.

Commencement:  2011

Completion: 

 
Modelling health status and mortality with phase-type markov-ageing models

Michael Sherris, Maria Govorun, Stephane Loisel

This project aims to develop a model for health status and survival using Phase-type distributions. The model is implemented in a stochastic Markov-ageing framework. These models have the advantage that they provide analytical solutions for survival distributions and for the values of health and mortality state dependent cash flows. By defining health states appropriately and calibrating the model to health and mortality data, the model is applied to value life annuities and long term care contracts. Capital requirements for providers of these products are also assessed for a portfolio of life annuity and long term care contracts using the model.

Commencement:  2011

Completion: 

 
Modelling systematic longevity and health risk: A subordinated markov model

Michael Sherris, Qiming Zhou

The research aims to develop a Markov ageing model for mortality using an underlying Markov model subordinated by a time change process in order to incorporate systematic longevity risk and heterogeneity in mortality based on health states. The model allows or closed-form expressions for expected value and variance of future survival rates. Model parameters and time change process are fitted to historical population mortality data and observed data on chronic diseases, cancer, health risk factors, and related population data on health condition heterogeneity. The model is applied to assess the significance of heterogeneity and systematic risk on product pricing for life annuities and health related insurance products based on the health status of the policyholder including expected values and uncertainty in these values. This research project was completed in 2012 and the results were presented to the Australian Actuarial Education and Research Symposium at Monash University.

Commencement:  2012

Completion: 2012

 
Multi-state actuarial models for health status with applications to long-term care

Michael Sherris, John Piggott, Joelle H Fong, Adam Wenqiang Shao

This project considers multi-state models for health status and the application of the models to assessing the impact of morbidity on long term care costs and product development. As individuals live longer the question arises as to the impact this has on an individual's health status requiring community care and long term residential care. By examining trends in health status and care needs from a number of countries along with Australian data, models for health status and transitions are estimated and projected allowing for mortality improvements. To account for uncertainty in transition rates, stochastic models are used along with scenario analysis.

Commencement:  2012

Completion:

 
Multi-state markov models for individual pathways to death (morbidity and mortality) using longitudinal data

Daniel Alai, Ramona Meyricke, Michael Sherris

Long term care products require an understanding of individual morbidity over the period from retirement to death. There is significant heterogeneity across individual pathways to death. We compare different methodologies for extending longevity risk models to include morbidity, and for modelling heterogeneity across individuals' morbidity and mortality risk. This is particularly important for financial products where contract payments depend on ill-health events as well as mortality e.g. lifetime healthcare insurance. A multi-state Markov model is employed to capture (possibly repeated) ill-health events along the pathway to death. The models are fit using longitudinal data and therefore capture variation across individuals' pathways to death. In 2012, the modelling framework was established and preliminary results were obtained using data from the US Health and Retirement Survey. An Australian longitudinal dataset suitable for modelling transition rates along individual  ageing pathways was obtained from the DYNOPTA research group based at the ANU node of CEPAR.

Commencement:  2012

Completion: December 2012

 

 
Price efficiency in the Dutch annuity market

Ralph Stevens, Edmund Cannon (Bristol University), Ian Tonks (University of Bath)

In this project, we provide the first analysis of annuity rates in the Netherlands for the period 2001-2010.  During this period, the number of annuity providers was high and stable and we find that falls in annuity rates can be explained entirely by changes in yields and life expectancy.  We show that annuitants could have increased their annuity income by about 5% by shopping around and purchasing their annuities from alternative providers.    Money's worth calculations show that the market is efficient by international standards, with a money's worth above 0.9 for the whole period and close to unity by the end of the period. The project fits within the broader thrust of research on benefit design being undertaken with CEPAR, and is a building block for a planned research collaboration between CEPAR and Netspar.

Commencement:  2011

Completion:

 
Progressive tax changes to private pensions in a life-cycle framework

George Kudrna, Alan Woodland

Tax concessions are a common feature of private pension pillars around the world. Most countries exempt pension fund earnings from any taxation but tax either benefits (EET regime) or contributions (TEE regime) progressively as regular private income. By contrast, Australia's superannuation taxation features concessional flat tax rates on contributions and fund earnings, with benefits being generally tax free. Concerned with the vertical equity of the current superannuation tax concessions, this paper provides a quantitative analysis of hypothetical replacements of the existing superannuation tax treatment with the EET and TEE regimes commonly found in other countries. Using a general equilibrium OLG model calibrated for Australia, we find that these hypothetical tax reforms to superannuation improve the vertical equity in the short, medium and long run, as indicated by larger relative welfare gains and income improvements experienced by lower income households. This research has been written up in a CEPAR Working Paper and was presented at the 18th International Conference on Computing in Economic and Finance in Prague in 2012.

Commencement : 2011

Completion: March 2012

 
Public sector pension funds in Australia: Longevity selection and liabilities

John Piggott, Michael Sherris, Joelle Fong, Carolyn Njenga

This paper assesses the cost and risk faced by public sector, defined benefit plan providers arising from uncertain mortality, including longevity selection, mortality improvements, and unexpected systematic shocks. Using longitudinal micro data on Australian pensioners, we quantify the extent of longevity selection at both aggregate and scheme level. We also show that as the age-membership structure in a pension scheme matures, scheme-specific longevity selection risk and systematic shocks become quantitatively more important and have larger consequences for plan liabilities than aggregate selection risk or the impact of mortality improvements. This research has been written up as a CEPAR Working Paper and was presented at the American Risk and Insurance Association 2012 Annual Meeting and the 20th Annual Colloquium of Superannuation Researchers.

Commencement:  2011

Completion: June 2012 

 
Public-private strategies to support asset-liability management for retirement insurance products (SLF finance project)

Ramona Meyricke, Michael Sherris, Craig Blackburn, John Piggott

Deploying stochastic optimisation techniques to capture the inherent uncertainty of the financial and economic environment, this project will focus on issues such as appropriate capital reserves for retirement insurance products, and new methodologies for estimating systematic longevity risk, aiming to build capacity to diversify this risk through better calibrated mortality based derivatives. It will investigate how government policy might more precisely support fragile insurance markets through issuing securities which provide the potential for asset-liability management strategies to be successfully pursued by pension funds and related entities.

Commencement:  2013

Completion: 

 
Reform of ill-health retirement benefits for police in England and Wales: The roles of national policy and local finance

Richard Disney, Rowena Crawford (Institute for Fiscal Studies, London)

Around the world, public sector pension schemes have been attracting increasing attention because of their generosity, and their implications for government budgets. Generosity is not confined to the value of the pension when it is paid. Other provisions, related to specific circumstances, such as health status or years of service, can add considerably to the present value of the pension fund liability. This paper examines the ill-health retirement rates of police officers in the forces of England and Wales between 2002-03 and 2009-10.  Differences in ill-health retirement rates across police forces are statistically related to area-specific stresses of policing and to force-specific differences in human resource policies.  Reforms to police ill-health retirement provisions and to pension arrangements that occurred in the mid-2000s - and in particular a shift in the incidence of financing ill-health retirement from central government towards cost-sharing with local police authorities - are described.  We show that these measures significantly impacted on the level of ill-health retirement, especially in forces with statistically higher rates of ill-health retirement pre-reform, once we control for stress factors.  However a degree of 'regression to the mean' cannot be fully rejected by placebo tests.  We also show that residual differences in ill-health retirement rates across forces after these reforms were enacted are statistically correlated with the differential capacities of police authorities to raise revenues from local property taxes, thereby alleviating constraints on central funding of ill-health retirement. The results of this research were written up as an NBER working paper.

Commencement:  2012

Completion: 2012 

 
Risk-based capital requirements of living benefits using a bayesian vector autoregression mortality model

Carolyn Njenga, Michael Sherris

Annuity providers have long term obligations to make payments to their policyholders for as long as they are living. Uncertain future longevity requires insurers to hold capital against these obligations. A new approach to modelling mortality allowing for parameter risk is used for Australian mortality to simulate the capital requirements for longevity in a risk-based capital framework. A Bayesian Vector Autoregression model is used to estimate and simulate an age-period life table for Australian males. The Australian Prudential Regulation Authority proposal for a constant permanent mortality decline of 25% is assessed. This proposal is shown to lead to higher levels of regulatory capital for annuity business than for a 1 in 200 mortality shock. An alternative age dependent simplification is proposed.

Commencement:  2011

Completion: 

 
Systemic risk and the hedging of longevity risk via securitisation

Ramona Meyricke, Michael Sherris

The failure of an annuity provider due to improvements in longevity could create problems within the financial system and the economy at large, and thus presents a systemic risk. Limited availability of reinsurance and the significant impacts of failing to manage longevity risk create a need for longevity derivatives, however they may create new channels of systemic risk. In this paper, first we review the systemic risks posed by the pension and insurance sectors, and by longevity risk securitisation. Second we model the cash flows and capital reserves of an annuity provider, with and without hedging using a fixed term longevity bond. Using this model we optimise the insurer's hedging strategy by minimising the cost of capital (as determined under Solvency II) plus cost of hedging for the insurer. Finally we compare the impact of different strategies on the probability of default of the insurer. In 2012 this research was presented at the Eighth International Longevity Risk and Capital Markets Solutions Conference.

Commencement:  2012

Completion:  December 2012