[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries
.

Current research

Projects already underway
Robust and efficient fitting of the generalised Pareto distribution with actuarial applications in view Led by Dr Vytaras Brazauskas of the University of Wisconsin-Milwaukee
The aim of the project is to develop modern methodological tools for the fitting of the generalised Pareto distribution (GPD), a widely applicable model, and to investigate the impact of the fits on risk measuring and risk modelling applications. In particular, the effects on estimation of such risk measures as value at risk and conditional tail expectations will be examined theoretically, via Monte Carlo simulations, and via realdata investigations. It is hoped to present the results at forthcoming conferences as well as submitting them for publication.

Exploration of a novel bootstrap technique for estimating the distribution of outstanding claim reserves in general insurance
Led by Dr Robert Cowell of Cass Business School
This project aims to:
>> Develop the theoretical background of the novel bootstrap method
>> Implement the variations of the bootstrap method in simple computer programs, and with these programs, compare the predicted distribution of reserves to known reserves in a set of simulated claims triangles, following appropriate (realistic) models of the claim development
>> Compare the predictions of the method to predictions arising from other stochastic claims reserving techniques that have been published in the literature. This project is due to report in October 2009.

Assessing longevity risk and pricing annuities with the Lee-Carter model
Led by Dr Iain Currie of Heriot-Watt University
This project, due to be presented in August 2009 as a sessional meeting paper, aims to develop an easily used tool for assessing longevity risk in the pricing of annuities and reserving for pensions.

IFS Retirement Saving Consortium
Led by Carl Emmerson of the Institute for Fiscal Studies (IFS)
A consortium of organisations including the Profession funded a programme of work on pensions at the IFS over the period January 2004 to December 2006. Since then the policy agenda has moved on considerably but there are still gaps in what is known about levels of saving and retirement provision across the life cycle. More recently, data resources have become available which, taken together, will allow us to answer important questions about the nature of both pension and non-pension saving across all stages of the life cycle.

Example questions to be researched are:
>> How do saving rates and wealth accumulation change over the lifecycle, and how have these patterns changed in recent years?
>> What is the relationship between people’s lifetime earnings and their retirement incomes, and what factors are associated with high ratios of the latter to the former?
>> How many people might be affected by the move to Personal Accounts, and what are the characteristics of these individuals and their employers?
>> How have recent trends in the overall value of pension provision differed between public and private sectors and how might such differences be expected to impact on retirement and savings outcomes?

In addition to the ongoing benefits of being part of this consortium and having the chance to influence leading research and reflect on policy implications with other senior industry and government officials, output from the project is expected in 2010.

Claims inflation and data validation in non-life reserving. Three scientific papers
Led by Professor Jens Nielsen of Cass Business School
Professor Nielsen writes, “Our work on claims inflation is based on or inspired by a... paper that... gives the theoretical foundation allowing a combination of modern econometric time series theory and classical statistical regression methods. [Our” first paper... will adjust the theory of the paper mentioned above such that actuarial reserving techniques of aggregated data can also be considered in combination with a latent time series. The second paper... will develop a validation technique of aggregated actuarial validation data. The third paper extends the second paper to the claims inflation case. Here we will probably not be able to find closed form solutions, but will take advantage of generalised linear model methods.” The project is due to report in September 2009.

Dependent competing risk survival: Impact on annuities and life expectancy
Led by Dr Vladimir Kaishev of Cass Business School
This project aims to provide further extensive investigations into the sensitivity of the methodology proposed in Kaishev et al. (2007); to consider model uncertainty with respect to two alternative formal definitions of elimination of a cause of death, see Elandt-Johnson (1976); to extend the methodology in order to incorporate different methods of closing the life table and extrapolating to a limiting age (of, for example, 120 years); and to apply the methodology proposed in Kaishev et al. (2007) to country-specific mortality data, provided in cause-specific format and compare the results for different countries. The project is due to report in October 2009.

Optimal capital and premium allocations in non-life insurance
Led by Dr Andreas Tsanakas of Cass Business School
It is proposed that capital allocation methods are developed that address practical problems. In particular, flexible capital allocation methods, based on optimisation criteria, such as the ones proposed in Dhaene et al. (2003), will be formulated that can control the effect of the dependence structure on the allocation.

A two-stage allocation algorithm will be developed that treats differently capital allocation at a strategic (for example, line of business) and policy level. Criteria that allocate capital fairly to members of an insurance group, without producing perverse incentives, will be used to derive capital allocations. A capital allocation method will be developed based on the use of demand functions proposed in the work of Zaks et al. (2007), that addresses the problem of fairly allocating capital within a portfolio while it grows and its final composition is uncertain. The due date for this project is December 2009.

____________________________________________________________

New projects
The Profession is delighted to announce the funding of two new research projects.

Systemic risk in financial services
Led by Dr Alistair Milne and a team from Cass Business School
The research will concentrate on exposure to common risks (one risk affecting many institutions in a similar way) and contagion effects (negative movements in values triggering further falls), and will cover:
>> A review of relevant literature n Analysis and commentary on past events
>> Description of the range of risks to which financial institutions are exposed and an analysis of the extent of those exposures
>> Analysis of the impact on institutions should a risk exposure materialise as an adverse risk event
>> Analysis of the extent to which regulation or market disclosures may have adverse feedback effects to amplify the impact of risk events n Comparison of proposed, existing and historical solvency regimes, including Solvency II and Basel II
>> Recommendations on changes to regulation or market disclosure requirements that may reduce the likelihood of adverse systemic impacts
>> Recommendations on actions that individual firms can take to mitigate the risk of adverse systemic impacts.

The team aims to present its interim findings at a seminar organised by the Profession on 9 June 2009 with the final report to be delivered shortly after. Results will be notified on the Profession’s website and in The Actuary.

International comparative study of mortality tables used in pensions
Led by Professor Richard Verrall and a team from Cass Business School
The aim of the project is not to move away from country-specific tables but to understand the extent to which there are systemic differences between standard European tables that could be accounted for by methodological differences in their construction or differences in the underlying data/assumptions/projections differences. The team is due to report in September 2009.

____________________________________________________________

More details on research, including the research process, applying for funding, contacts and prizes are available at: http://www.actuaries.org.uk/knowledge/research