16 August 2016
New mortality research initiatives and update on the fall in UK (period) life expectancies UK Mortality and Longevity Update- May 2016. Bennett and Ezatti (see sec 1.4) think previous forecasts are pessimistic. The CMI (see sec 1.2) thinks previous forecasts are optimistic. The ONS (see sec 1.3) publishes its 2014-based projections on 29 Oct 2015.
Human Mortality Database is in danger (see sec 2)
Note: the referenced articles come from a variety of sources in the UK and elsewhere, and I have included them because I think they are interesting. Thanks to the many people who have contributed. If you have any comments or further news, please email me.
Some papers are firewalled, but members of the Institute and Faculty of Actuaries (IFoA) can access many, including those in the North American Actuarial Journal and the Annals of Actuarial Science, via the Athens information portal. Details here.
Contents
- UK papers, presentations and articles
- International papers, presentations and articles
- Upcoming events: International Symposium Sept 2016
1. UK papers, presentations and articles
1.1 Institute and Faculty of Actuaries
IFoA Longevity Bulletin No 6, Pandemics July 2015
Discusses pandemic risks with articles from a multi-disciplinary group on subjects such as risk modelling, epidemiology, health policy and technology.
Contains a good section on further reading on Pandemics and Antimicrobial resistance.
Call for Research Project Proposals on the Key Challenges for Actuarial Science, IFoA Sept 2015
The IFoA has re-constituted a Research and Thought Leadership Committee with responsibility for coordinating research across the organisation. A KEY QUESTIONS document has been produced to summarise some of the major research challenges of interest to the actuarial community.
Question no 2 is: “What are the actuarial consequences of changes in how people age?”
The call for (funded) research ends on 5 Oct 2015.
The Longevity Catalysts Working Party added to their section on Postponement of Ageing. It covers the Longevity Panel’s paper “What is Ageing”, and the tests on Metformin, a FDA approved treatment for type 2 diabetes.
Further comment on the trials on Metformin here: “Anti-ageing pill pushed as bona fide drug - Regulators asked to consider ageing a treatable condition”, Erika Check Hayden, Nature June 2015
The Nature of Longevity Risk, Sacha Dhamani, SIAS June 2015
Quote: “The IAA breakdown of longevity risk (Trend, Level, Catastrophe and Volatility) is still perceived to be the standard definition used despite many practitioners considering them insufficient as a foundation for modelling longevity risk…. it is time for a revised conceptual framework for longevity risk which can provide a stronger foundation to take into account how different types of provider are exposed to longevity risk and the differences in their approach to longevity risk management. Sacha Dhamani’s paper and presentation is intended to trigger a debate in this area by proposing such a framework.”
1.2 Continuous Mortality Investigation; full news here
CMI_2015 Mortality Projections Model Sept 2015 was published, accompanied by CMI Working paper 84 which describes changes to the calibration of initial rates and illustrates the impact of incorporating the latest England and Wales population mortality dataset into the Model. To accompany the Model the CMI has also published CMI Working Paper 83. Recent mortality in England & Wales which contains analysis of recent mortality in England & Wales, particularly the exceptional experience of the last four years, to provide background information for CMI_2015. Publically-available. This led to the announcement:
Increases in life expectancy between 2011 and 2015 much lower than in the past: CMI Sept 2015
Increases in life expectancy between 2011 and 2015 have been much lower than in the past. And life expectancy at age 75 has shown no improvement at all between 2011 and 2015.
Between 2000 and 2011 life expectancy increased by over three months per year on average. If that trend had continued then life expectancy in 2015 would have been thirteen months higher than in 2011. However, the CMI’s analysis shows an increase of only four months, meaning that nine months of potential life expectancy have been lost.
The picture is potentially even worse at high ages: life expectancy at age 75 shows no improvement at all.
Tim Gordon, Chairman of the CMI, said, ‘Insurers and pension funds will need to consider whether this recent experience indicates a fundamental change in mortality improvement trends, or whether it is a short term variation due to influences such as influenza and cold winters – the financial implications are material.’
Hugh Nolan, Chief Actuary at JLT Employee Benefits, commented:
“This reduces the liabilities of UK private sector pension schemes by some £15bn”
Working paper 82. The CMI format for heatmaps of mortality improvements has an accompanying macro to enable others to use the same format. Publically-available.
The future of the CMI Mortality Projections Model
The CMI will now consult on possible revisions to the model over 2016, and aims to publish the first version of the next iteration of the model in March 2017.
In view of the widespread use of the CMI model, the CMI will hold public consultation meetings in Edinburgh and London in October 2015 to discuss issues and plans including:
- recent national mortality experience, in particular the higher than expected mortality in 2015
- responsiveness v stability, i.e. how do we distinguish short-term volatility from longer-term trends
- adjustments to exposure data
- possible modifications to existing deterministic models
- state space models
- coherent modelling of multiple populations (males v females, UK v developed world)
- incorporation of user-assumptions (e.g. long term improvement rates).
1.3 Office for National Statistics
Population Estimates for UK, England and Wales, Scotland and Northern Ireland - Mid-2014
Released: 25 June 2015
The latest national life tables 2012-14: Sept 2015
Based on this period, life expectancy at birth is M: 79.1, F: 82.8 at age 65 M: 18.4, F 20.9.
Decennial Life Tables, English Life Tables, No.17, 2010-12 (England and Wales): Sept 2015
These are fully graduated life tables that are prepared every 10 years based on the 3 years of data around a census year
- Over the last 100 years life expectancy at birth has increased by nearly 3 years per decade.
- For males, life expectancy at birth increased from 51 years in 1910-1912 to 79 years in 2010-12, while for females it increased from 55 to 83 years.
- People aged 60 could expect to live around 9 years longer in 2010-2012 than 100 years earlier.
Estimates of the Very Old (including Centenarians), 2002 to 2014
Death registrations summary tables, England and Wales - 2014
Death registration statistics including cause of death figures by age and sex, mortality rates, death registrations by area of residence and single year of age.
Avoidable Mortality in England and Wales, 2013
Statistics on avoidable mortality - deaths caused by certain conditions which should not occur in the presence of timely and effective health care or through wider public health interventions.
The European Standard Population was updated in 2013 (I missed it earlier)
The new version is now incorporated into all UK official statistics which feature age-standardised mortality rates.
Upcoming ONS releases/publications include:
- Trend in life expectancy by socioeconomic position by the National Statistics Socioeconomic Classification, England and Wales: 21 Oct 2015
- National Population Projections - 2014-based projections: 29 Oct 2015
1.4 Other
Are our population mortality forecasts too low?
The future of life expectancy and life expectancy inequalities in England and Wales: Bayesian spatiotemporal forecasting. Bennett JE, Li G, Foreman K, et al. Lancet 2015; 386: 163–70.
Forecasts mortality and life expectancy for England and Wales' districts. Combining the results, concludes that present forecasts underestimate the expected rise in life expectancy, especially for men, and calls for improved health and social policies to needed to avoid a so-called grand divergence in health and longevity by district and socio-economic status.
Professor Majid Ezzati said: “The bigger gains in life expectancy we predict will mean pensions will have larger payouts, and health and social services will have to serve an older population than currently planned. We also forecast rising inequalities, with bigger increases in lifespan for people in affluent areas than those in disadvantaged areas”. “Our methods better reflect how longevity is changing than those currently used, and our forecasts are more accurate”
Includes an interesting data visualisation tool.
Future inequalities in life expectancy in England and Wales John N Newton, Chief Knowledge Officer for Public Health England. Lancet Vol 386 July 2015
Quote: “In The Lancet, James Bennett and colleagues (above) have used elaborate Bayesian models to analyse present mortality patterns in England and Wales. They then forecasted life expectancy to 2030 for 375 districts. Their models take separate account of age, cohort, period, and geography; the one that performed best emphasised the effect of cohort. One of the many benefits of this approach is that it makes few, if any, assumptions about trends in mortality. It allows for non-linear trends of the sort that might be expected to follow known patterns of, for example, smoking in past decades.
The model output is more optimistic than official figures.”
Inequalities in life expectancy - Changes over time and implications for policy: D Buck and D Maguire, King’s Fund. (See also Bosworth for USA analysis)
“The social gradient in life expectancy improved between 1999–2003 and 2006–10. In short, income-related inequalities in life expectancy improved. Marmot’s goal – ‘to shift the gradient’ – happened. Some factors are shown to be consistently important in explaining life expectancy differences between areas. “
Unemployment and older people’s deprivation play a particularly important role in determining differences between areas in life expectancy.
Changes in health in England, with analysis by English regions and areas of deprivation, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. John N Newton et al. Lancet Sept 2015
“We use the Global Burden of Disease Study 2013 data on mortality and causes of death, and disease and injury incidence and prevalence to analyse the burden of disease and injury in England as a whole, in English regions, and within each English region by deprivation quintile.”
Concludes that health in England is improving although substantial opportunities exist for further reductions in the burden of preventable disease if policies are designed to reduce inequalities associated with the socioeconomic determinants of health. Declines in mortality have not been matched by similar declines in morbidity, resulting in people living longer with diseases. Health policies must therefore address the causes of ill health as well as those of premature mortality. Systematic action locally and nationally is needed to reduce risk exposures, support healthy behaviour
Office of Science - Foresight Project on Ageing
The UK Government’s ‘Future of an ageing population’ project has published nineteen reports in the past six months. In particular see:
Future of ageing: life expectancy and healthy life expectancy trends, Carol Jagger, March 2015
Compares UK life expectancy and health expectancy trends with other EU and OECD countries, and discusses the main influences on trends.
Concludes (inter alia) “Increases in heath expectancies in the UK are not keeping pace with gains in life expectancy, particularly at older ages. This expansion of ill health and disability is also observed in some EU countries, Japan and the USA, although other European countries (Belgium, Sweden, Switzerland) appear to be experiencing compression of disability, possibly due to smaller gains in life expectancy.”
FCA Retirement income market study - Another “mis-selling” situation, this time for annuities?
The UK Financial Conduct Authority says that “eight out of ten consumers who purchase their annuity from their existing provider could get a better deal on the open market”. It is focusing on encouraging better selling methods, but the Daily Telegraph “understands that the Financial Conduct Authority is working on the details of a programme in which up to one in four savers who retired in the past six years could claim redress”. Could this also apply to people who bought regular annuities when they could have had an enhanced annuity?
Common conditions can make a big difference to retirement income – Partnership Group
The most common condition that people declare when taking out an enhanced annuity is high blood pressure followed by obesity, diabetes and high cholesterol. All of these conditions can be relatively minor and treated with medication but mean that people can receive almost £21,000 more income assuming a 20-year retirement.
2. International papers, presentations, articles and websites
Human Mortality Database in danger
The Human Mortality Database (HMD) collects and provides open, international access to detailed mortality and population data for 37 countries or areas. It is the work of two teams of researchers in the USA and Germany, with the help of financial backers and scientific collaborators from around the world. It is widely used by researchers to compare consistent multi-country data. Funding for the HMD is now at risk and the team are keen to find new sources of funding. If you can help, contact Magali Barbieri at hmd@mortality.org
UN Sustainable development goals Sept 2015
The UN General Assembly met in September to sign off on a new global development agenda designed to ensure a better and more just future for all the world’s people. The 2030 Agenda for Sustainable Development, if taken seriously, should impact population mortality and morbidity in all the countries that signed up. In particular there is the commitment to “leave no one behind”.
Predictors of Exceptional Longevity: Effects of Early-Life and Midlife Conditions, and Familial Longevity, Gavrilov and Gavrilova, July 2015
“Knowledge of strong predictors of mortality and longevity is very important for actuarial science and practice.” Comparing centenarians born in the United States in 1890–1891 with peers born in the same years who died at age 65, investigates how interactions between parental characteristics, early-life conditions and midlife environment play a significant role in survival to advanced ages.
Japan population mortality July 2015
The Health Ministry released Abridged Life Tables for Japan 2014. Life expectancy at birth is at a record high for males and females. Males 80.50 world no 3, Females 86.83 world no 1. (Source: Hiroshi Yamazaki)
Sources of Increasing Differential Mortality Among the Aged by Socioeconomic Status: B P Bosworth et al, June 2015
Uses data from the Health and Retirement Study (HRS) to explore the extent and causes of widening differences in life expectancy by socioeconomic status (SES) for older persons in the USA.
- Finds: “There is strong statistical evidence of a growing inequality of mortality risk by SES among more recent birth cohorts compared with cohorts born before 1930.
- Both educational attainment and career earnings as constructed from Social Security records are equally useful indicators of SES, although the distinction in mortality risk by education is greatest for those with and without a college degree.
- There has been a significant decline in the risk of dying from cancer or heart conditions for older Americans in the top half of the income distribution, but we find no such reduction of mortality risk in the bottom half of the distribution.”
Estimates of individual life expectancies, R Cumpston et al
Investigates how “to quantify the relationships between socio-economic variables, diseases and mortality. This may allow better estimates of individual life expectancies, helping individuals make investment and lifestyle choices”
The paper proposes the use of macro-models to approximately replicate past changes in causes of death, and the use of microsimulation to validate assumed relationships between socio-economic variables and diseases.
Valuing annuities based on alternative mortality projections Tang, Browne and Bruhn
“We describe the recent developments in mortality projections that have arisen from the UK Institute and Faculty of Actuaries’ Continuous Mortality Investigation (CMI), and apply this method to Australian mortality data. Projected mortality under this approach is used to calculate values for both immediate and deferred life annuities, and is compared with values arising from other mortality projections for Australia”.
The eleventh international Longevity Risk and Capital Markets Solutions Conference Lyon, Sept 2015
(Programme hyperlinked above) Some papers will be published in the journal Insurance: Mathematics and Economics.
AncestryDNA and Calico to Research the Genetics of Human Lifespan July 2015
Calico, a Google-backed company focused on longevity research and therapeutics, and AncestryDNA, a consumer genetics company, announced a partnership to investigate human heredity of lifespan. They will work together to analyse and investigate the role of genetics and its influences in families experiencing unusual longevity. Calico will then focus its efforts to develop and commercialize any potential therapeutics that emerge from the analysis.
International Mortality and Longevity Symposium 2016
7-9 September 2016, Royal Holloway, University of London, UK
This conference will provide a multi-disciplinary forum for the exchange of information on the latest relevant research, and also an opportunity to learn about established knowledge from a range of different disciplines, all with the aim of better understanding and managing this complex yet critical subject.
The themes for this event are:
- How will population longevity develop in the future in your defined country or internationally?
- New evidence or analyses of historical morbidity and mortality patterns.
- What would disrupt current mortality trends?
- How will Big Data and 'internet of things' contribute to mortality trends and analyses?
- New techniques for mortality and longevity analyses and forecasting.
- Implication of mortality and morbidity trends.
To submit a proposal for a presentation or a poster please complete the call for speakers survey by Friday 26 February 2016
This is a note for the UK actuarial profession and others, and for the International Actuarial Association (IAA). The last six-monthly report is here
The most recent UK version of the report from the IAAMWG is here
The web page for the Institute and Faculty of Actuaries Mortality Research Steering Committee is here.
The web page for the IAA Mortality Working Group is here.
See below for a Reading list of articles on mortality provided by the IFoA Library Services
Brian Ridsdale
September 2015
br@ridsdales.com
INSTITUTE AND FACULTY OF ACTUARIES
LIBRARY SERVICES
Reading list on Articles on Mortality Sept 2015
ANNALS OF ACTUARIAL SCIENCE
Brazauskas, Vytaras; Jones, Bruce L; Zitikis, Ricardas (2015). Trends in disguise. [RKN: 47181]
Annals of Actuarial Science (2015) 9(1) : 58-71.
Shelved at: Per: AAS (Lon) Shelved at: EDI Journal Store.
Human longevity is changing, but at what rate? Insurance claims are increasing, but at what rate? Are the trends that we glean from data true or illusionary? The shocking fact is that true trends might be quite different from those that we actually see from visualised data. Indeed, in some situations the upward trends (e.g. inflation) may even look decreasing (e.g. deflation). In this paper, we discuss this “trends in disguise” phenomenon in detail and offer a way for estimating true trends.
DOI: http://dx.doi.org/10.1017/S1748499514000232 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Alai, Daniel H; Arnold-Gaille, Séverine; Sherris, Michael (2015). Modelling cause-of-death mortality and the impact of cause-elimination. [RKN: 47186]
Annals of Actuarial Science (2015) 9(1) : 167-186.
Shelved at: Per: AAS (Lon) Shelved at: EDI Journal Store.
The analysis of causal mortality provides rich insight into changes in mortality trends that are hidden in population-level data. Therefore, we develop and apply a multinomial logistic framework to model causal mortality. We use internationally classified cause-of-death categories and data obtained from the World Health Organization. Inherent dependence amongst the competing causes is accounted for in the framework, which also allows us to investigate the effects of improvements in, or the elimination of, cause-specific mortality. This has applications to scenario-based forecasting often used to assess the impact of changes in mortality. The multinomial model is shown to be more conservative than commonly used approaches based on the force of mortality. We use the model to demonstrate the impact of cause-elimination on aggregate mortality using residual life expectancy and apply the model to a French case study.
DOI: http://dx.doi.org/10.1017/S174849951400027X (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Sanders, David (2015). Book review: Portfolio Theory and Risk Management, by Maciej J. Capinski, Ekkehard Kopp, Cambridge University Press, 2014. [RKN: 47187]
Annals of Actuarial Science (2015) 9(1) : 187-188.
Shelved at: Per: AAS (Lon) Shelved at: EDI Journal Store.
Review of Portfolio Theory and Risk Management, by Maciej J. Capinski, Ekkehard Kopp, Cambridge University Press, 2014, 169pp. (hardback), £50. ISBN: 9781107003675
DOI: http://dx.doi.org/10.1017/S1748499514000311 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
AUSTRALIAN JOURNAL OF ACTUARIAL PRACTICE
Tang, Chen; Browne, Bridget; Bruhn, Aaron (2015). Valuing annuities based on alternative mortality projections. [RKN: 47249]
Australian Journal of Actuarial Practice (2015) 3 : 23-33.
Available online
Shelved at: EDI Journal Store.
Future mortality is a key component when pricing longevity-based products such as annuities. Estimating future mortality is, however, a significant challenge, with a range of approaches adopted by various users to date. We describe the recent developments in mortality projections that have arisen from the UK Institute and Faculty of Actuaries’ Continuous Mortality Investigation (CMI), and apply this method to Australian mortality data. Projected mortality under this approach is used to calculate values for both immediate and deferred life annuities, and is compared with values arising from other mortality projections for Australia. Our results are noticeably similar to other projection approaches, in particular those of Tickle and Booth (2014) [Tickle, L, & Booth, H, The longevity prospects of Australian seniors: an evaluation of forecast method and outcome, Asia Pacific Journal of Risk and Insurance(2014) 8(2): 259-92] and the Productivity Commission (2013) [Productivity Commission (Australia), An ageing Australia: preparing for the future, Government of Australia, 2013]. Sensitivity analyses are also presented. It is apparent that an embedded and ongoing practice of effective risk management, rather than the mere pursuit of a more “accurate” picture of future mortality, is key to managing longevity exposure.
Internet URL: http://www.actuaries.asn.au/knowledge-bank/australian-journal-of-actuarial-practice
Cumpston, Richard; Sarjeant, Hugh; Service, David (2015). Estimates of individual life expectancies. [RKN: 47250]
Australian Journal of Actuarial Practice (2015) 3 : 35-46.
Available online
Shelved at: EDI Journal Store.
Estimates of individual life expectancies require assumptions about future causes of death, and allowances for socio-economic factors. Deaths from circulatory diseases, cancers, respiratory diseases and external causes were 75% of deaths in Australia in 2012. All have shown continuing downwards trends since 1986. Reported deaths from the next three major causes of death have all increased. Expert advice is needed to understand the reasons for past changes, and to make reasonable assumptions about future deaths from each cause. Unmarried persons have higher mortality rates than the married. Recent Danish data show that lone persons are more likely to die from cardiovascular disease, respiratory disease and suicide. Australian data show that lone persons are more likely to have mental disorders and nervous system diseases, and disabilities arising from external causes. UK data show that persons in unskilled occupations have higher mortality rates. The paper proposes the use of macro-models to approximately replicate past changes in causes of death, and the use of microsimulation to validate assumed relationships between socio-economic variables and diseases.
Internet URL: http://www.actuaries.asn.au/knowledge-bank/australian-journal-of-actuarial-practice
Adamic, Peter (2015). Life expectancy improvement with a cure distribution for a cause of death. [RKN: 47252]
Australian Journal of Actuarial Practice (2015) 3 : 59-61.
Available online
Shelved at: EDI Journal Store.
In many circumstances, the increase in life expectancy when a certain cause of death is eliminated is sought. Traditionally, these calculations have been based on the assumption that the cause in question is simply omitted, which is equivalent to the cause being taken out of consideration, from the outset, with certainty. In this paper, we propose continuous and discrete models whereby a probability distribution for the cure of a specific cause of death over time can be incorporated so as to more accurately predict the increase in life expectancy. The theoretical results are applied to a real data set involving HIV-related deaths from the State of Colorado, United States of America, between the years 2000 and 2012.
Internet URL: http://www.actuaries.asn.au/knowledge-bank/australian-journal-of-actuarial-practice
Browne, Bridget (2015). Recent Australian insured lives mortality: a review of the Actuaries Institute - Financial Services Council 2004-2008 lump sum graduation. [RKN: 47255]
Australian Journal of Actuarial Practice (2015) 3 : 99-106.
Available online
Shelved at: EDI Journal Store.
In 2012 the Graduation Taskforce of the Actuaries Institute published a report “Graduation of the 2004–2008 Lump Sum Investigation Data”. This is now the most recent publicly available report on Australian insured lives’ mortality experience. The tables in that report are compared with, on the one hand, contemporary Australian population mortality, and on the other, the preceding Australian insured lives tables, IA95-97. Thus the paper reports and comments on the level of insured lives’ mortality compared with that of the total population as well as changes in both of these over the period from 1996 to 2006. It appears that insured lives mortality has improved less than population mortality over the 10 years.
Internet URL: http://www.actuaries.asn.au/knowledge-bank/australian-journal-of-actuarial-practice
BAJ
Winston, Robert (2015). Genes, genomics, genetics: human or hubris? : Abstract of the London Discussion. [RKN: 47308]
BAJ (2015) 20(2) : 404-422.
Spring Lecture by Professor Robert Winston on 15 May 2014 to the Institute and Faculty of Actuaries and discussion
Shelved at: online.
DOI: http://dx.doi.org/10.1017/S1357321714000294 (access via Athens login http://www.openathens.net/)
Jagger, Carol (2015). Presentation of Autumn Lecture by Professor Carol Jagger : Abstract of the Edinburgh discussion. [RKN: 47329]
BAJ (2015) 20(2) : 423-440.
Autumn Lecture by Professor Carol Jagger on 1 October 2014 to the Institute and Faculty of Actuaries, Edinburgh and discussion
Shelved at: online.
DOI: http://dx.doi.org/10.1017/S1357321714000057 (access via Athens login http://www.openathens.net/)
French, Declan; O'Hare, Colin (2015). From atheoretical to informed mortality modelling. [RKN: 47307]
BAJ (2015) 20(3) : 1-2 tbc.
Shelved at: online.
A recent period of low inflation and low interest rates has highlighted the importance of mortality risk to actuarial practice. In order to better understand changing patterns of mortality, the profession has recognised the need to gain insights from other disciplines and established the Mortality Research Steering Group. Our recent research funded by an Institute and Faculty of Actuaries (IFoA) Pump-priming grant follows this research agenda by incorporating insights from demography, health economics, epidemiology and financial econometrics to mortality modelling.
DOI: http://dx.doi.org/10.1017/S135732171500001X (access via Athens login http://www.openathens.net/)
INSURANCE: MATHEMATICS & ECONOMICS
Mammen, Enno; Martínez Miranda, María Dolores; Nielsen, Jens Perch (2015). In-sample forecasting applied to reserving and mesothelioma mortality. [RKN: 47225]
Insurance: Mathematics & Economics (2015) 61 : 76-86.
This paper shows that recent published mortality projections with unobserved exposure can be understood as structured density estimation. The structured density is only observed on a sub-sample corresponding to historical calendar times. The mortality forecast is obtained by extrapolating the structured density to future calendar times using that the components of the density are identified within sample. The new method is illustrated on the important practical problem of forecasting mesothelioma for the UK population. Full asymptotic theory is provided. The theory is given in such generality that it also introduces mathematical statistical theory for the recent continuous chain ladder model. This allows a modern approach to classical reserving techniques used every day in any non-life insurance company around the globe. Applications to mortality data and non-life insurance data are provided along with relevant small sample simulation studies.
DOI: http://dx.doi.org/10.1016/j.insmatheco.2014.12.001 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Lin, Tzuling; Wang, Chou-Wen; Tsai, Cary Chi-Liang (2015). Age-specific copula-AR-GARCH mortality models. [RKN: 47228]
Insurance: Mathematics & Economics (2015) 61 : 110-124.
See online abstract for correct notation used.
In this paper, we propose AR-GARCH (autoregression-generalized autoregressive conditional heteroskedasticity) models to fit and forecast mortality rates for a given age by two alternative approaches. Specifically, one approach is to fit a time series of mortality rates for some age to an AR(nn)-GARCH(1, 1) model, and project the mortality rate for that age in the next nnth year; the other is to fit an AR(1)-GARCH(1, 1) model, and project the mortality rates recursively for the age in the next consecutive years. Further, we employ the copula method to capture the inter-age mortality dependence. Adopting mortality data of Japan, the UK, and the USA, we demonstrate that it is indispensable to consider the conditional heteroskedasticity in our mortality models which provide better performances in out-of-sample projection and prediction intervals with a higher degree of coverage than the Lee-Carter model. Finally, we numerically illustrate with mortality data of Japan that VaR (Value at Risk) values for longevity risk, regarded as additional reserves for annuity or pension providers, will be overestimated if the conditional heteroskedasticity or/and the inter-age mortality dependence structure are ignored.
DOI: http://dx.doi.org/10.1016/j.insmatheco.2014.12.007 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Li, Hao; O'Hare, Colin; Zhang, Xibin (2015). A semiparametric panel approach to mortality modeling. [RKN: 47241]
Insurance: Mathematics & Economics (2015) 61 : 264-270.
During the past twenty years, there has been a rapid growth in life expectancy and an increased attention on funding for old age. Attempts to forecast improving life expectancy have been boosted by the development of stochastic mortality modeling, for example the Cairns-Blake-Dowd (CBD) 2006 model [A J G Cairns, D Blake, K Dowd, A two-factor model for stochastic mortality with parameter uncertainty: theory and calibration, Journal of Risk and Insurance (2006) 73: 687-718]. The most common optimization method for these models is maximum likelihood estimation (MLE) which relies on the assumption that the number of deaths follows a Poisson distribution. However, several recent studies have found that the true underlying distribution of death data is overdispersed in nature (see Cairns et al. 2009 [A J G Cairns et al., A quantitative comparison of stochastic mortality models using data from England && Wales and the United States, North American Actuarial Journal (2009) 13(1): 1-35] and Dowd et al. 2010 [K. Dowd et al., Evaluating the goodness of fit of stochastic mortality models, Insurance Mathematics and Economics (2010) 47: 255-265]). Semiparametric models have been applied to many areas in economics but there are very few applications of such models in mortality modeling. In this paper we propose a local linear panel fitting methodology to the CBD model which would free the Poisson assumption on number of deaths. The parameters in the CBD model will be considered as smooth functions of time instead of being treated as a bivariate random walk with drift process in the current literature. Using the mortality data of several developed countries, we find that the proposed estimation methods provide comparable fitting results with the MLE method but without the need of additional assumptions on number of deaths. Further, the 5-year-ahead forecasting results show that our method significantly improves the accuracy of the forecast.
DOI: http://dx.doi.org/10.1016/j.insmatheco.2015.02.002 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Yang, Bowen; Li, Jackie; Balasooriya, Uditha (2015). Using bootstrapping to incorporate model error for risk-neutral pricing of longevity risk. [RKN: 47264]
Insurance: Mathematics & Economics (2015) 62 : 16-27.
Where mortality projection is concerned, it is essential to quantify the extent of the prediction error. This is especially important in light of the aggravating risk of longevity and as a result the increasing demand for longevity-linked products. In the literature so far, only parameter error and process error have been considered jointly while the issue of model error has yet been systematically studied. In this paper, we propose a method to account for process error, parameter error and model error in an integrated manner by modifying the semi-parametric bootstrapping technique. We apply the method to two data sets from the Continuous Mortality Investigation (CMI) and use the simulated scenarios to price the q-forward contracts via the maximum entropy approach. We find that model selection has a significant impact on the risk-neutral valuation results and thus it is crucial to incorporate model error in mortality projection.
DOI: http://dx.doi.org/10.1016/j.insmatheco.2015.02.004 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Danesi, Ivan Luciano; Haberman, Steven; Millossovich, Pietro (2015). Forecasting mortality in subpopulations using Lee–Carter type models: a comparison. [RKN: 47275]
Insurance: Mathematics & Economics (2015) 62 : 151-161.
The relative performance of multipopulation stochastic mortality models is investigated. When targeting mortality rates, we consider five extensions of the well known Lee-Carter single population extrapolative approach. As an alternative, we consider similar structures when mortality improvement rates are targeted. We use a dataset of deaths and exposures of Italian regions for the years 1974–2008 to conduct a comparison of the models, running a battery of tests to assess the relative goodness of fit and forecasting capability of different approaches. Results show that the preferable models are those striking a balance between complexity and flexibility.
DOI: http://dx.doi.org/10.1016/j.insmatheco.2015.03.010 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
NORTH AMERICAN ACTUARIAL JOURNAL
Tsai, Cary Chi-Liang; Yang, Shuai (2015). A linear regression approach to modeling mortality rates of different forms. - 23 pages. [RKN: 75004]
North American Actuarial Journal (2015) 19 (1) : 1-23.
Shelved at: Per NAAJ (Lon)
In this article, we propose a linear regression approach to modeling mortality rates of different forms. First, we repeat to fit a mortality sequence for each of K years (called the fitting years) with another mortality sequence of equal length for some year (called the base year) differing by tj years (j = 1, …, K) using a simple linear regression. Then we fit the sequences of the estimated slope and intercept parameters of length K, respectively, with the sequence of {tj} by each of the simple linear regression and random walk with drift models. The sequences of the fitted slope and intercept parameters can be used for forecasting deterministic and stochastic mortality rates. Forecasting performances are compared among these two approaches and the Lee-Carter model. The CBD model is also included for comparisons for an elderly age group. Moreover, we give a central-death-rate–linked security to hedge mortality/longevity risks. Optimal units, purchased from the special purpose vehicle, which maximize the hedge effectiveness for life insurers and annuity providers, respectively, are derived and can be expressed in terms of the cumulative distribution function of the standard normal random variable. A measure with hedge cost involved, called hedge effectiveness rate, for comparing risk reduction amount per dollar spent among mortality models is proposed. Finally, numerical examples are presented for illustrations.
Internet URL: http://www.openathens.net/
Arnold-Gaille, Séverine; Sherris, Michael (2015). Causes-of-Death mortality: What do we know on their dependence?. - 13 pages. [RKN: 75134]
North American Actuarial Journal (2015) 19 (2) : 116-128.
Shelved at: Per NAAJ (Lon)
Over the last century, the assumption usually made was that causes of death are independent, although it is well-known that dependancies exist. Recent developments in econometrics allow, through Vector Error Correction Models (VECMs), to model multivariate dynamic systems including time dependency between economic variables. Common trends that exist between the variables may then be highlighted, the relation between these variables being represented by a long-run equilibrium relationship. In this work, VECMs are developed for causes-of-death mortality. We analyze the five main causes of death across 10 major countries representing a diversity of developed economies. The World Health Organization website provides cause-of-death information for about the last 60 years. Our analysis reveals that long-run equilibrium relationships exist between the five main causes of death, improving our understanding of the nature of dependence between these competing risks over recent years. It also highlights that countries usually had different past experience in regard to cause-of-death mortality trends, and, thus, applying results from one country to another may be misleading.
Internet URL: http://www.openathens.net/
Russo, Vincenzo; Giacometti, Rosella; Rachev, Svetlozar; Fabozzi, Frank J (2015). A three-factor model for mortality modeling. - 13 pages. [RKN: 75135]
North American Actuarial Journal (2015) 19 (2) : 129-141.
Shelved at: Per NAAJ (Lon)
In this article, we propose a three-factor model for mortality modeling in which the dynamic of the entire term structure of mortality rates can be expressed in closed form as a function of three variables x, t, and y. Due to this feature, we are able to project mortality rates across age (x), across time (t), and for y years (y 1) after t. Our proposal differs from most existing models where only the one-year mortality rate is considered (y = 1). The model is characterized by three parameters that are calibrated yearly. We describe the stochastic dynamic of the three factors with correlated autoregressive processes. We generate stochastic scenarios accounting for the historical mortality trend in a consistent manner with the Gompertz law. Using population mortality data for Italy, the U.S., and the U.K., the model’s forecasting capability is assessed, and a comparative analysis with other models is provided.
Internet URL: http://www.openathens.net/
POPULATION STUDIES
Hanson, Heidi A; Smith, Ken R; Stroup, Antoinette M; Harrell, C Janna (2015). An age-period-cohort analysis of cancer incidence among the oldest old, Utah 1973-2002. [RKN: 47165]
Population Studies (2015) 69(1) : 7-22.
Available via Athens: Taylor & Francis Online
Shelved at: Per. (Lon)
We used age-period-cohort (APC) analyses to describe the simultaneous effects of age, period, and cohort on cancer incidence rates in an attempt to understand the population dynamics underlying their patterns among those aged 85+. Data from the Utah Cancer Registry (UCR), the US Census, the National Center for Health Statistics (NCHS), and the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) programme were used to generate age-specific estimates of cancer incidence at ages 65-99 from 1973 to 2002 for Utah. Our results showed increasing cancer incidence rates up to the 85-89 age group followed by declines at ages 90-99 when not confounded by the separate influences of period and cohort effects. We found significant period and cohort effects, suggesting the role of environmental mechanisms in cancer incidence trends between the ages of 85 and 100.
DOI: http://dx.doi.org/10.1080/00324728.2014.958192 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Babiarz, Kimberley Singer; Eggleston, Karen; Miller, Grant; Zhang, Qiong (2015). An exploration of China's mortality decline under Mao: a provincial analysis, 1950-80. [RKN: 47166]
Population Studies (2015) 69(1) : 39-56.
Available via Athens: Taylor & Francis Online
Shelved at: Per. (Lon)
Between 1950 and 1980, China experienced the most rapid sustained increase in life expectancy of any population in documented global history. We know of no study that has quantitatively assessed the relative importance of the various explanations proposed for this gain in survival. We have created and analysed a new, province-level panel data set spanning the decades between 1950 and 1980 by combining historical information from China's public health archives, official provincial yearbooks, and infant and child mortality records contained in the 1988 National Survey of Fertility and Contraception. Although exploratory, our results suggest that gains in school enrolment and public health campaigns together are associated with 55-70 per cent of China's dramatic reductions in infant and under-5 mortality during our study period. These results underscore the importance of non-medical determinants of population health, and suggest that, in some circumstances, general education of the population may amplify the effectiveness of public health interventions.
DOI: http://dx.doi.org/10.1080/00324728.2014.972432 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
Mayhew, Les; Smith, David (2015). On the decomposition of life expectancy and limits to life. [RKN: 47167]
Population Studies (2015) 69(1) : 73-89.
Available via Athens: Taylor & Francis Online
Shelved at: Per. (Lon)
Life expectancy is a measure of how long people are expected to live and is widely used as a measure of human development. Variations in the measure reflect not only the process of ageing but also the impacts of such events as epidemics, wars, and economic recessions. Since 1950, the influence of these events in the most developed countries has waned and life expectancy continues to lengthen unabated. As a result, it has become more difficult to forecast long-run trends accurately, or identify possible upper limits. We present new methods for comparing past improvements in life expectancy and also future prospects, using data from five developed, low-mortality countries. We consider life expectancy in 10-year age intervals rather than over the remaining lifetime, and show how natural limits to life expectancy can be used to extrapolate trends. We discuss the implications and compare our approach with other commonly used methods.
DOI: http://dx.doi.org/10.1080/00324728.2014.972433 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net
SCANDINAVIAN ACTUARIAL JOURNAL
Itskovich, Igor; Roudebush, Bradley T (2015). A new parametric model for converting excess mortality from clinical studies to insured population. [RKN: 47135]
Scandinavian Actuarial Journal (2015) 2 : 184-199.
Available via Athens: Taylor & Francis Online -- See online abstract for mathematical notation used
Shelved at: EDI Journal Store.
We propose a new parametric model – the generalized excess mortality (GEM) model – for converting excess mortality from clinical to insured population. The GEM model has been formulated as a generalization of the excess death rate (EDR) model in terms of a single adjustment parameter (m) that accounts for a partial elimination of a clinical study’s EDR due to the underwriting selection process. The suggested value of the parameter m depends only on the ratio of the impairment’s prevalence rate in the insured population to that in the clinical population. The model’s development has been implemented in two phases: the design phase and the validation phase. In the design phase, the data from the National Health and Nutrition Examination Survey I pertaining to three broad impairments (diabetes, coronary artery disease, and asthma) have been used. As a result, the following equation for the parameter m has been proposed: mk = (Pi,k/Pc,k)n, where Pi,k, Pc,k are the prevalence rates of impairment k under study in the insured and the clinical populations, respectively, and n a single universal parameter with its value best approximated as n = 0.5 (95% confidence interval 0.5-0.6). In the validation phase, several independent clinical studies of three other impairments (Crohn’s disease, epilepsy, and chronic obstructive pulmonary disease) were used. As it has been demonstrated in the validation phase, for a number of impairments, the GEM model can provide a better fit for observed insured population mortality than either one of the conventional EDR or mortality ratio models.
DOI: http://dx.doi.org/10.1080/03461238.2013.807299 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net/
Fernández-Durán, J J; Gregorio-Domínguez, M M (2015). Seasonal mortality for fractional ages in short term life insurance. [RKN: 47172]
Scandinavian Actuarial Journal (2015) 3 : 266-277.
Available via Athens: Taylor & Francis Online
Shelved at: EDI Journal Store.
A uniform distribution of deaths between integral ages is a widely used assumption for estimating future-lifetimes; however, this assumption does not necessarily reflect the true distribution of deaths throughout the year. We propose the use of a seasonal mortality assumption for estimating the distribution of future-lifetimes between integral ages: this assumption accounts for the number of deaths that occurs in given months of the year, including the excess mortality that is observed in winter months. The impact of this seasonal mortality assumption on short-term life insurance premium calculations is then examined by applying the proposed assumption to Mexican mortality data.
DOI: http://dx.doi.org/10.1080/03461238.2013.819028 (access via Athens login http://www.openathens.net/)
Internet URL: http://www.openathens.net/
e-mail: libraries@actuaries.org.uk
Contact Details
Brian Ridsdale