CEPAR is supported by
Skip Navigation Links Home Publications Working Papers Working Papers 2012
Search CEPAR
Filter by
Ermanno Pitacco
July 2012

Many modern insurance products are designed as packages, whose items may be either included or not in the product actually purchased by the client. For example: the endowment insurance which can include various rider benefits and options, the Universal Life insurance, the Variable Annuities, the presence of possible LTC benefits in pension products.

The benefits provided by these products imply a wide range of "guarantees" and hence risks borne by the insurance company (or the pension fund). Guarantees and inherent risks clearly emerge in recent scenarios, in particular because of volatility in the financial markets and trends in mortality / longevity. Appropriate modeling tools are then needed for pricing and reserving. Hence, a progressive shift from expected present values, and their prominent role in life insurance (and pension) calculations, to more modern and complex approaches, viz the Enterprise Risk Management based approach, is currently updating the actuarial toolkit.

However the implementation of complex mathematical methods often constitutes, on the one hand, an obstacle on the way towards sound pricing and reserving principles. On the other hand, facing the risks by charging very high premiums trivially reduces the insurer's market share. Then, alternative solutions can be suggested by an appropriate product design which aims at sharing risks between the insurer and the policyholders. Interesting examples are provided by the design of life annuities as regards the longevity risk, and by profit participation mechanisms as regards the financial market risks.

Hazel Bateman, Andy Lai, and Ralph Stevens
December 2012

We assessed alternative presentations of investment risk using a discrete choice experiment which asked subjects to rank three investment portfolios for retirement savings across nine risk presentation formats and four underlying risk levels. Using Prospect Theory utility specifications we estimate individual-specific parameters for risk preferences in gains and losses, loss aversion, and error propensity variability. Our results support presentations that describe investment risk using probability tails. Risk preferences and error propensity were found to vary significantly across sociodemographic groups and levels of financial literacy. Our finndings should assist regulatory efforts to disclose risk information to the mass market.

Bei Lu
May 2012

China's new Rural Pension scheme, announced in October 2009, is destined to be the world's largest, at least in terms of membership. By the time it is fully implemented, in 2012, it will comprise some 600 million members, with about 105 million receiving benefits at that time.

The new scheme is motivated by concern about the widening income gap between the urban and the rural sectors, the rich and the poor in China. But it is unclear that the rural elderly will benefit by the full amount of the pension, because many currently receive private transfers from their children, and these may be adjusted after the introduction of pension benefits.

This paper uses the China Health and Retirement Longitudinal Study (CHARLS) data to investigate the net impact on the old age household income inequality when the new rural pension plan is in place. Logit and OLS analysis are used to estimate the changes of the probability and value of family transfers when other variables change. Results indicate that net private transfers are in most cases uncorrelated with household income, suggesting that the current public transfer (the new rural pension) will not crowd out private transfers. Based on these findings, Gini index simulations are employed to compare income inequality with and without rural public pension. The improvement in Gansu rural income inequality is significant while there is only slight improvement in Zhejiang. Transfers to low income regions from migrants are found to   significantly improve income inequality for rural elders as well.

Bei Lu and John Piggott
August 2012

China is currently undergoing the largest regional migration in the world's history. Young rural workers are moving to urban areas, often in a different province, for substantial periods of their working lives. Social security policy inChina, while framed by national protocols and policy guidelines, is administered at more than 2000 lower-level jurisdictions, typically cities and counties, and at present this compromises pension entitlements ofChina's 150 million rural migrant workers.

This paper proposes a Notional Defined Contribution (NDC) mechanism to ensure pension mobility for migrating workers, in both the accumulation and drawdown phases. Plan governance would ensure independence from the three existing pension systems. Although the plan's design would be of the NDC type, requiring no pre-funding, in practice the relatively young migrant demographic has the potential to generate considerable reserves. An appropriately structured NDC plan of this type is shown to be viable by reference to a previously developed model ofZhejiangProvince's social security systems. Such a plan would remove mobility barriers to migrating workers, increase the retirement benefit for mobile workers, and reduce the future government liability for payouts in other pension systems in which migrants currently hold membership. 

Julie R. Agnew, Hazel Bateman, and Susan Thorp.
November 19, 2012

Existing research shows that adjustment to retirement is correlated with pre-retirement planning. This study presents new insights into the retirement preparedness of Australians at the later stages of working life. Recent surveys of those approaching and entering retirement show that the extent of planning around exiting the workforce, financial management, bequest provision and activities during retirement vary greatly. We find that more than half of Australians in their 50s and 60s have not planned key aspects of retirement. A small minority have detailed and advanced plans. In addition, expectations around these issues and actual realisations may not be well matched.

Julie R. Agnew, Hazel Bateman, and Susan Thorp.
November 19, 2012

This paper presents new evidence from a national survey of non-retired individuals with superannuation accounts between the ages of 25 and 65. Fielded in June of 2012, the survey responses suggest that Australians often are ill informed about many important features of the superannuation system. This lack of retirement system knowledge is consistent with findings in other countries. Specifically, the survey suggests that Australian respondents are most challenged by questions associated with the access age to their superannuation savings, the risky composition of balanced funds (a popular default investment option), and the tax treatment of superannuation contributions and investments. Regression analysis suggests a relationship between superannuation knowledge and savings behaviour. The results provide motivation for further research in the area and suggest more can be done to educate individuals about the superannuation system.

Julie R. Agnew, Hazel Bateman, and Susan Thorp.
December 3, 2012

We implement a customized survey to a representative sample of 1,024 Australians to examine the relationship between financial literacy and retirement planning. Overall we find aggregate levels of financial literacy similar to comparable countries with the young, least educated, unemployed and those not in the labor force most at risk. However, unlike the international norm, we find that financial skills increase with age. The role played by the Australia's mandatory private retirement arrangements, system of defaults, and interactions with the means-tested safety net pension at older ages remain open questions.

Joelle H. Fong, Benedict SK. Koh, and Olivia S. Mitchell
September 2012

This paper examines how inability to perform activities of daily living relates to the risk of nursing home admission over older adults' life courses. Using longitudinal data on persons over age 50 from the Health and Retirement Study, we show that aging one year boosts the probability of having two or more disabilities by 9 to 12 percent in a multivariate logistic model. Moreover, at least three-fifths of all 65-year old men and three-quarters of women will experience disability levels during their remaining lifetimes severe enough to trigger nursing home admission. Our analysis also suggests that certain types of disability are more important than others in predicting nursing home admittance and use, which has implications for the design of benefit triggers for long-term care insurance programs.

Erik Hernaes, Simen Markussen, John Piggott and Ola Vestad
June 2012

The relationship between retirement and mortality is studied with a unique administrative data set covering the full population of Norway. A series of retirement policy changes in Norway reduced the retirement age for a group of workers but not for others. By employing a difference-in-differences framework based on monthly birth cohort and treatment group status we first establish that the early retirement program significantly reduced the retirement age - this remains true when we account for program substitution, for example into the disability pension. Using instrumental variables estimation we find that retirement age has no effect on mortality.


Joelle H. Fong, John Piggott, and Michael Sherris
31 May 2012

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.

Elena Capatina

In order to design and understand the effects of health care policy reforms, we need to better understand the different ways in which health affects individuals and their economic decisions. This paper studies four channels through which health affects individuals: (1) productivity, (2) medical expenditures, (3) available time and (4) survival probabilities, and assesses their roles in determining labour supply, asset accumulation and welfare.


John Piggott and Rafal Chomik

Australia's retirement income provision system, comprising the 'three pillars' of a means-tested Age Pension, mandatory occupational superannuation and other, voluntary long term savings, is at the heart of understanding the fiscal implications of ageing. While the Intergenerational Report, an account of long term fiscal sustainability, is celebrating its tenth birthday since the first edition was published, the Superannuation Guarantee (SG), first implemented in 1992, is about to turn a sprightly twenty. This paper considers the intergenerational reports as a prism for studying fiscal, demographic, and policy developments in the Australian retirement income system over the last decade and into the future.

Daniel Alai and Michael Sherris

Longevity risk arising from uncertain mortality improvement is one of the major risks facing annuity providers and pension funds. In this paper we show how applying trend models from non-life claims reserving to age-period-cohort mortality trends provides new insight in estimating mortality improvement and quantifying its uncertainty. Age, period, and cohort trends are modelled with distinct effects for each age, calendar year, and birth year in a generalized linear models framework. The effects are distinct in the sense that they are not conjoined with age coefficients, borrowing from regression terminology, we denote them as main effects. Mortality models in this framework for age-period, age-cohort, and age-period-cohort effects are assessed using national population mortality data from Norway and Australia to show the relative significance of cohort effects as compared to period effects. Results are compared with the traditional Lee-Carter model. The bilinear period effect in the Lee-Carter model is shown to resemble a main cohort effect in these trend models. However the approach avoids the limitations of the Lee-Carter model when forecasting with the age-cohort trend model.

Cagri Kumru and John Piggott

This paper studies the interaction between capital income taxation and a means tested age pension in the context of an overlapping generations model, calibrated to the UK economy. Recent literature has suggested a rehabilitation of capital income taxation (Conesaet al. (2009)), predicated on the idea that capital is a complement with retirement leisure. This leads naturally to the conjecture that a publicly funded age pension contingent upon holdings of capital or capital income may have a similar effect. We formalize this using a stochastic OLG model with multiple individuals differentiated by labour productivity and pension entitlement. Our preliminary findings suggest that a means tested pension has effects similar to capital income taxation in a life-cycle context.

George Kudrna and 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.

Loretti Dobrescu, Dimitris Christelis, Alberto Motta

Using life-history survey data from eleven European countries, we investigate whether childhood conditions, such as socioeconomic status, cognitive abilities and health problems influence portfolio choice and risk attitudes later in life. After controlling for the corresponding conditions in adulthood, we find that superior cognitive skills in childhood (especially mathematical abilities) are positively associated with stock and mutual fund ownership. Childhood socioeconomic status, as indicated by the number of rooms and by having at least some books in the house during childhood, is also positively associated with the ownership of stocks, mutual funds and individual retirement accounts, as well as with the willingness to take financial risks. On the other hand, less risky assets like bonds are not affected by early childhood conditions. We find only weak effects of childhood health problems on portfolio choice in adulthood. Finally, favourable childhood conditions affect the transition in and out of risky asset ownership, both by making divesting less likely and by facilitating investing (i.e., transitioning from non-ownership to ownership).

Daniel Alai, Zinoviy Landsman and Michael Sherris

Systematic improvements in mortality results in dependence in the survival distributions of insured lives. This is not allowed for in standard life tables and actuarial models used for annuity pricing and reserving. Systematic longevity risk also undermines the law of large numbers; a law that is relied on in the risk management of life insurance and annuity portfolios. This paper applies a multivariate gamma distribution to incorporate dependence. Lifetimes are modelled using a truncated multivariate gamma distribution that induces dependence through a shared gamma distributed component. Model parameter estimation is developed based on the method of moments and generalized to allow for truncated observations. The impact of dependence on the valuation of a portfolio, or cohort, of annuitants with similar risk characteristics is demonstrated by applying the model to annuity valuation. The dependence is shown to have a significant impact on the risk of the annuity portfolio as compared with traditional actuarial methods that implicitly assume independent lifetimes.

Dimitris Christelis and Loretti I. Dobrescu

Using micro data from eleven European countries, we investigate the impact of being socially active on cognition in older age. Cognitive abilities are measured through scores on numeracy, fluency and recall tests. We address the endogeneity of social activities through panel data and instrumental variable methods. We find that social activities have an important positive effect on cognition, with the results varying by gender. Fluency is positively affected only in females, while numeracy only in males. Finally, recall is affected in both sexes. We also show that social activities, through their effect on cognition, influence positively households' economic welfare.


Elena Veprauskaite and Michael Sherris

This paper considers optimal reinsurance based on an assessment of the reinsurance arrangements for a large life insurer. The objective is to determine the reinsurance structure, based on actual insurer data, using a modified mean-variance criteria that maximises the retained premiums and minimizes the variance of retained claims while keeping the retained risk exposure constant, assuming a given level of risk appetite. The portfolio of life and disability policies use quota-share, surplus and a combination of both quota-share and surplus reinsurance. Alternative reinsurance arrangements are compared using the modified mean-variance criteria to assess the optimal reinsurance strategy. The analysis takes into account recent claims experience as well as actual premiums paid by insured lives and to the reinsurers. Optimal reinsurance cover depends on many factors including retention levels, premiums and the variance of sum insured values (and therefore claims), as a result an insurer should assess the tradeoff between retained premiums and the variance of retained claims based on its own experience and risk appetite.


Maathumai Nirmalendran, Michael Sherris and Katja Hanewald

This paper provides a detailed quantitative assessment of the impact of solvency capital requirements on product pricing and shareholder value for a life insurer. A multi-period firm value maximization model for a life annuity provider, allowing for stochastic mortality and asset returns, imperfectly elastic product demand, as well as frictional costs, is used to derive optimal capital and pricing strategies for a range of solvency levels reflecting differences in regulatory regimes. The model is calibrated using realistic assumptions and the sensitivity of results assessed. The results show that value-maximizing insurers should target higher solvency levels than the Solvency II regulatory 99.5% under assumptions of reasonable levels of policyholder's aversion to insolvency risk. Even in the case of less restrictive solvency regulation, policyholder price elasticity and solvency preferences are shown to be important factors for a life insurer's profit maximizing strategy.

Dominic Ho and Michael Sherris

The insurance linked securities (ILS) market is an increasingly important alternative asset class for which risk and return analysis differs from other asset classes. Measures of portfolio risk and return for an ILS portfolio are based on the expected losses and expected excess returns over the risk free rate. Multiple criteria decision making (MCDM) has found successful applications to many real world decision problems. This paper examines the application of two popular MCDM methods, Analytical Hierarchy Process (AHP) and ELECTRE III, to ILS portfolios. These methods are used to screen the securities before constructing portfolios using linear optimisation with constraints. The objective function is to minimise the portfolio expected loss for a given level of expected excess return. Upper and lower bounds are also placed on the investment in each individual ILS. The results demonstrate the benefits from applying MCDM to ILS portfolio selection.


Elisabetta Magnani, Garima Verma, Anu Rammohan

We model the allocation of time resources by adult children between competing caring activities - those towards coresiding elderly and those towards coresiding children. We test the implications of our model for children's school performance by focusing on Indonesia, a country characterized by heterogeneity in social norms, population ageing and reliance on the family for elderly support. Specifically, we exploit the unique richness of the Indonesian Family Life Survey (IFLS) (Wave 2 to Wave 4) to find robust evidence of a negative impact on children's school achievement of social norms regulating elderly bequests to coresiding adult carers.

Jonathan Ziveyi, Craig Blackburn, Michael Sherris

This paper considers the pricing of European call options written on pure endowment and deferred life annuity contracts, also known as guaranteed annuity options. These contracts provide a guarantee value at maturity of the option. The contract valuation is dependent on stochastic interest rate and mortality processes. We assume single-factor stochastic squareroot processes for both interest rate and mortality intensity, with mortality being a timeinhomogeneous process. We then derive the pricing partial differential equation (PDE) and the corresponding transition density PDE for options written on deferred contracts. The general solution of the pricing PDE is derived as a function of the transition density function.

We solve the transition density PDE by first transforming it to a system of characteristic PDEs using Laplace transform techniques and then applying the method of characteristics. Once an explicit expression for the density function is found, we then use sparse grid quadrature techniques to generate European call option prices on deferred insurance products. This approach can easily be generalised to other contracts which are driven by similar stochastic processes presented in this paper. We test the sensitivity of the option prices by varying independent parameters in our model. As option maturity increases, the corresponding option prices significantly increase. The effect of miss-pricing the guaranteed annuity value is analysed, as is the benefit of replacing the whole-life annuity with a term annuity to remove volatility of the old age population.


George Kudrna and Alan Woodland
2012 (update)

In this paper we investigate the macroeconomic and welfare effects of the major changes of the mandatory superannuation reform proposed in the 2010-11 Australian federal budget. These changes include gradual increases in the mandatory employer contributions from 9 to 12 percent of gross earnings and a policy that effectively removes the concessional 15 percent tax on mandatory contributions for workers with annual taxable income of up to $37,000. Using a computable overlapping generations model that incorporates main aspects of mandatory superannuation, the means tested age pension and progressive personal income taxation, we find significantly larger superannuation asset accumulations as a result of the reform, which generate increases in domestic total assets and household saving. The reform improves self-funding in retirement, with government expenditures on the age pension falling by almost 4.6 percent in the long run. The reform also has positive impacts on households' long run welfare, with higher income households solely benefiting from the increased superannuation contributions while lower income households from the contribution tax removal. The aggregate efficiency calculations indicate that the superannuation reform improves efficiency, generating a gain of almost 0.8 percent or $11,753 in initial resources for each future generation.

Elisabetta Magnani

Training (for workers) and innovation (for workplaces) are not free lunches. From the viewpoint of the firm, training is also highly risky, because there is uncertainty over the size of any future returns from employer-provided training. Stylized facts stress that constraints in achieving preferred working hours have major impacts on job satisfaction. Consequently hour constraints may lead to workers' job mobility and older workers' retirement. Firms internalize the risk of workers' mobility by reducing their training investments in these workers. I contrast this model with a signalling model of hour constraints where, in the face of asymmetric information over workers' quality and reliability, and so over profitability of training, workers may trade present hour constraints (at the current wage), for training (and future wage) opportunities. This set of reasoning implies that, empirically, we should observe a positive correlation between training and hour constraints at the individual level. I use two matched employer-employee datasets, for Australia and Canada respectively, to test the competing empirical implications of these two models for the link between hour constraints and training. The main result of this study is that there is little support for hour constraints as a signal of future reliability and productivity. Rather, hour constrained individuals appear to have less chance to receiving training. This result survives a number of robustness exercises that attempt to control for selection on observables and selection on unobservables that determine the hour constraint outcome. Institutional differences in the retirement funding system, and the differential appeal of outside option (the option of exiting the labour force) in Australia and Canada in the two survey years contribute to explain the different patterns of training and hour constraints older workers face in these two countries.

Meliyanni Johar and Shiko Maruyama
February 2012

When siblings wish for the well-being of their elderly parents, the cost of caregiving and long-term commitment creates a free-rider problem among siblings. We estimate a sequential game to investigate externality and strategic interaction among adult siblings regarding their location choice relative to their elderly parents. Using the US Health and Retirement Survey, we Önd a positive externality and strategic interaction. The Örst-mover advantage of eldest children and the prisonerís dilemma are likely to exist but their magnitudes are negligible compared with ine¢ ciency in joint utility. Ine¢ ciency is large in a family with an educated, widowed mother and with educated siblings who are younger (relative to parents), married, and similar to each other. Had siblings fully internalized externality and jointly maximized utility sum in 2010, 17% more parents with multiple children would have had a child nearby. Public policies that reduce childrenís private costs may enhance social welfare.