Mortality Assumption On Funding Of Pension Schemes Finance Essay - Free Essay Example

Benjamin Franklin wrote, In this world nothing can be said to be certain, except death and taxes, but while death might be a certainty, its timing is far from certain. It is generally accepted that there exists a relationship between an individuals socio -economic status and their mortality, but that this relationship is difficult to measure. The problem of how to fund a defined benefit pension scheme arises from the problem of how to find cost of pension schemes. This study examines the relationship between mortality for individuals in a pension scheme and its impact on pension funding methods. We consider a very general setting where there exists basic funding methods and we try to find explicit solution how to coup up with future improvements and decreasing mortality. Introduction: In this project we have tried to examine effect of change in mortality assumption on funding of pension schemes. The aims of project are: To assess the appropriateness of current mortality assumptions i.e. use of up to date mortality tables or using projected mortality with future improvements. To make broad comparison between different funding methods using different mortality rates. To consider whether recommendation should be made about appropriate mortality assumption and funding method. To consider whether a single statistic could be used as a measure of the strength of the mortality assumptions used in funding methods. The population of many countries might undergo dramatic changes in the coming years due to continuous increase in life expectancy. The fact that people seems to live longer and the low mortality rates contribute to an increasing share of elderly people in total population in future. These project is written with the intension that it will be r ead by actuaries as well as by non actuaries. To merely give list of the actuarial methods and assumptions used in valuation of pension liability will be just misleading. This project considers the impact on different actuarial funding methods for using different mortality tables for pensioners. Chapter 1 discuss about basic type of pension benefits, which are defined contribution and defined benefits. But, for our study we only discuss about defined benefits and calculation of past service benefits and future service benefits using actuarial assumptions i.e. demographic and financial assumptions. It also gives basic idea of basic three actuarial funding methods i.e. entry age method, attained age method and projected unit method. In Chapter 2 we will be discussing on basis for selecting mortality assumptions and selection of mortality tables which are available and effect of future improvements on life expectancies and how population structure has changed from past and what will be expected population structure inclusive of increase in retirees and pensioners. In Chapter 3 after discussion of available mortality tables in chapter 2 we now see the impact of using old and new mortality tables on one pension scheme and assumed data set for the same scheme and analyze which is most suitable funding method for the same pension scheme. We will also look at the financial impact of changing mortality table. Finally in Chapter 4 we will conclude for using up to date mortality tables and funding methods which can be used. We also suggest the available alternatives to compensate for decreasing mortality and future improvements. And finally we provide some direction for further research for using mortality investigation. Chapter 1: Types of Occupational Pension Schemes There are two common types of Occupational Pension Schemes Defined Contribution Schemes Defined contribution schemes are much simpler and can be summarized as follows. Here the level of contributions may be defined in absolute term or percentage of salary or earning of employee. Defined Benefit Schemes Defined benefit is benefit payable on retirement and it is specified monthly or yearly benefits which are defined in advance on employeesÂÂ  earnings or salaryÂÂ  history, tenure of service andÂÂ  age, rather than depending onÂÂ  investmentÂÂ  returns. 1.2 Funding of Defined Benefit Schemes These benefits can be either Unfunded (Pay-As-You-Go) or Funded. In anÂÂ  unfundedÂÂ  defined benefit pension, no assets are set aside and the benefits are paid for by the employer or scheme as and when they are paid. In aÂÂ  fundedÂÂ  scheme, contributions from the employer, and sometimes also from scheme members, are invested in a fund towards meeting the benefits. Typically, the contributions to be paid are regularly reviewed in a valuation of the schemes assets and liabilities, carried out by anÂÂ  actuaryÂÂ  to ensure that the pension fund will meet future payment obligations considering set of assumptions. 1.3 Valuation and Liability 1.3.1 Liability The promise to pay certain defined benefit depends on future timing and durations which are not fixed or certain, but dependent on beneficiary. Also the amount of benefit is uncertain (e.g. if it depends on final salary or average salary which is unknown). There will be actuarial involvement from the time of benefit promised till they are actually paid i.e. when benefits are to be paid (demographic assumption) the level of benefit to be paid (financial or economic assumptions) Our objective is to calculate amount of money we need as at present to fulfill promised benefits of past pension liability and to calculate amount of money required to contribute toward fund for the benefits to be paid in future considering certain actuarial assumptions. These actuarial assumptions are discussed later in the chapter. 1.3.2 Purpose of Valuation There are several assumptions and methods of valuation which actuaries can use to value pension liability for funding purpose discussed later in the chapter. The need to calculate and make provision for these benefits in advance requires actuarial involvement, to find out the present value of future payments i.e. projecting the amount which is to be held now in order to meet the uncertain commitments in future; this is done by assuming some future mortality and expected future cash flows. 1.3.3 Actuarial Calculations and Decrements There are several factors which affect the actuarial valuation, we will discuss in particular the probabilities of following factors which are called decrements and will try to make service table. We consider an active member under DB scheme. There are four cases which can affect the pace of funding. Age (i.e. normal) Retirement Early retirement because of ill health Death in service Withdrawal (leaving the employer or scheme) These can be explained in multiple decrement model, in form of multiple state model shown Appendix II Fig: 1.1. 1.3.4 Actuarial Assumptions The Actuarial assumptions are divided in two main parts: Demographic assumptions: These assumptions are required to project when the benefit will be payable. Mortality: Mortality assumptions are required both before and after retirement. Generally, the assumption of lighter mortality implies that more pensions will be paid and will be paid for longer, this increases pace of funding. Withdrawal Rates: Schemes expects to profit from members leaving service. This situation might occur where the benefit to the person who leaves were subject to price inflation while normal retirement benefits were subject to (higher) salary inflation. The inclusion of withdrawal rates normally reduces the pace of funding. Ill Health Retirement: Where enhanced benefits are payable on Ill Health Early Retirement, an assumption will be required to assess the amount of Ill Health Early Retirement benefit to be paid. Higher assumed Ill Health Early Retirement rates will normally increase s pace of funding, although the impact is offset through the assumption of heavier mortality. Proportions married: Where Spouses benefits are provided, an assumption is normally made regarding the proportion of members who will be married at retirement, leaving, or death. Higher proportions imply faster funding but the impact of this assumption is not significant. Financial / Economic assumptions: These assumptions are required to project amount of the benefit will be payable. Investment Return/Interest Rate: This is the level of return expected to be achieved on fund before members retire. As the rate of investment is used to discount future benefits, a higher rate will lead to lower value being placed on future benefits, hence slower pace of funding through lower contribution rates Price Inflation: Benefits are often linked to price inflation (both pre and post retirement), so projected benefits will depend on level of inflation assumed for the future. Future price inflation can be obtained from the difference between yields from index linked bonds and government bonds. Salary Inflation: This assumption will determine projected benefits, where they are linked to final salary either on retirement or on exit from scheme or on death. The assumption is likely to be based on the used for price inflation, with an addition of 1% or 2% to reflect historical experience. 1.3.5 The Service Table Based on above model, we can construct and use the service table. We define a radix, as expected number of actives at age and we also define to be the expected number of who will retire at NRA, retire with ill health, die and withdraw, respectively before. Probability of that active member at age x should suffer any one of the decrement before age are: 1.3.6 The Salary Scale It is the increases in pay an employee gets when they spend a certain length of time at a particular level over a period of time due to inflation and promotion. We define salary scale inclusive of salary inflation and promotion for a person aged after years: 1.3.7 Valuations for Past Service Benefits (PSB) and Future Service Benefits (FSB) Let pension scheme provides a pension of 1/100 of future final pensionable salary, guaranteed for 5 years and if person died his widow will get 50% pension, a person joins scheme at age, with past service of and his current salary is Sal and it is assumed 90% married proportion. Let be EPV of an age retirement pension of 1 p.a. commencing at age. Future pensionable salary at age Probability of retirement at age We now define some functions: where we define: Similarly we can define functions for ill health retirement (), withdrawal and death. We have to take care of survival probability for person withdrawing from the pension scheme before retirement till his normal retirement. Now we look at: , accumulated for each future year of service till retirement age. We will accumulate each year accrual separately till retirement so total FSB is: where we define: Example 1.1: Suppose 1.4 Actuarial Funding Methods 1.4.1 A good funding method should meet the following four factors (Lee E M, 1986) Stability: It is generally accepted that an employer will look for Stability. Fluctuating cash flows are considered unacceptable. Liquidity: Only expected present value of future income is greater than expected present value of future outgo is not sufficient condition. Any funding method must ensure there is enough money available for the fund to pay the promised benefits. Security: Any funding scheme is secured if sufficient fund is built up to make future liabilities payments. However, as Lee (Lee E M, 1986) says the mere existence of fund separate from the assets of the employer obviously does not in itself guarantee pension rights. The size of the fund in relation to its liabilities is crucial. Durability: Basically it is special case of stability; it is required to deal with changing structure, without becoming unstable. However, the relative importance of these and other issues affecting the rate at which funds are put aside to meet future liabilities varies from case to case. 1.4.2 Actuarial Funding Methods After calculation of liabilities considering actuarial assumptions, we will look now the available methods of funding to fund required to meet future liabilities. The term funding method is used to refer to the way of determining the amount and timing of contributions made to meet the future liabilities. Whichever the method we use, the main objective is always the same, the contributions made to the fund, have to be sufficient to ensure that promised benefits are paid on time. Under the funding methods we need to define the following: Standard Fund (SF) This is the amount of liabilities to be recognized (theoretical value of fund that should be held) as on valuation date. Standard Contribution Rate (SCR) This is the amount of contribution required as derived by the method used. It assumes that fund held equals the standard fund. In this context the fund may be taken to be the value of asset held in pension fund. The contribution derived may be defined as an absol ute amount or percentage of salary. Contributions are only normally payable in respect of members accruing benefit i.e. active members of pension scheme. Recommended Contribution Rate (RCR) This is the contribution rate which is required to maintain the accrued liability and accumulated fund equal. Therefore RCR is SCR plus correction adjustment (i.e. any imbalance between the accumulated assets and accrued liabilities). We will discuss the following three methods of funding: Entry Age Method Aim To establish the level of contribution rate that, when payable over the active lifetime of the employees, is sufficient to meet the benefits being provided. Description Attained Age Method Aim To establish for active members of the pension scheme the level of future contribution rate such that, the future contributions will sufficient to provide future accruals of benefits. Description Projected Unit Credit Method (Projected Unit Method) Aim To main tain a fund equal to the value of accrued benefits, by taking in to consideration projected amount to date of payment. Description We will look in later chapters at the available mortality tables, the effect of changing mortality table and impact of future improvements on funding methods discussed above. Chapter 2: 2.1 Morality Assumption and Standard Mortality Tables Benjamin Franklin wrote, In this world nothing can be said to be certain, except death and taxes, but while death might be a certainty, its timing is far from certain. As with any actuarial calculation, technical provisions require assumptions to be made about the future course of all those factors affecting the cost of providing the benefits. These assumptions must beÂÂ  chosen prudently. Key assumptions will include inflation, investment return and how long scheme beneficiaries are expected to live (longevity). A mortality rate refers to the assumed probability of dying within a year whereas longevity usually refers to the future expected lifetime derived from any particular set of mortality rates. There have been significant new developments in the field of mortality relevant to pension scheme funding. Evidence has shown for many years that mortality is steadily reducing, so that the expectation of life (longevity) is increasing. Evidence also shows that there is s ignificant variation in pensioner mortality by amount of pension. There is evidence to suggest that socio economic circumstances and lifestyle choices such as smoking, drinking and exercise habits have an impact on mortality. However, it is not likely that these factors can be accessed directly. We should bear in mind that mortality has the following features: wide variability is observed between individuals; there is variability year on year in the whole population; long term trends can be observed in age specific mortality of whole populations; and historically, experts have usually underestimated the rate at which mortality will reduce (longevity increase). An analysis of a schemes own mortality experience will usually provide relevant evidence. However, with a small membership, random fluctuations could make this unreliable as a sample from which to draw inferences about the future. In the case of small schemes, it may not be worthwhile to undertake this an alysis, instead relying on more general factors such as industry, occupation or pension size or use of the standard tables. There are two basic decisions need to be taken on mortality assumptions while valuation: the base table (including any adjustment) to reflect the schemes current mortality experience; and the allowance for future improvement. We are available with many pensioners mortality tables some older ones are: Based on 1967 1970 experience collected from UK insurance companies: Life Office Pensioners, Female, Amounts (PA) 90f. Life Office Pensioners, Male, Amounts (PA) 90m. Based on 1991 1994 experience collected from UK insurance companies, published in CMIR 17, 1999: Pensioners, Female, Amounts (PFA) 92. Pensioners, Male, Amounts (PMA) 92. Latest are: Based on 1999 2002 experience and around 90,000 deaths, collected from UK insurance companies, published in CMI Working Papers 21 22, 2006 and CMIR 23, 2009. Pensioners, males, N ormal, Amounts (PNMA00) Pensioners, females, Normal, Amounts (PNFA00) Based on 2000 2006 experience called SAPS tables Series 1 S1, collected by 30 June 2007 from UK self administered pension schemes (SAPS), published in CMI Working Papers 34 and 35, 2008. All pensioners (excluding dependants), Female, Lives (S1PFL) All pensioners (excluding dependants), Female, Amounts (S1PFA) All pensioners (excluding dependants), Female, Amounts, Light (S1PFA_L) All pensioners (excluding dependants), Female, Amounts, Heavy (S1PFA_H) All pensioners (excluding dependants), Male, Lives (S1PML) All pensioners (excluding dependants), Male, Amounts (S1PMA) All pensioners (excluding dependants), Male, Amounts, Light (S1PMA_L) All pensioners (excluding dependants), Male, Amounts, Heavy (S1PMA_H) The data collected by CMI which covers 7 years period, which includes almost 10 million lives and around 380,000 deaths and collected from 350 separate pension schemes (Punter S outhall, 2008). (See Appendix I Table 2.1 for the extracts of some of the tables mentioned above.) 2.2 Selection of Mortality Table It is normal practice to use standard mortality tables, unlike when choosing other demographic assumptions. However, you may choose to adjust those standard mortality tables to reflect various characteristic of covered group, or to provide for expectation of future mortality improvement. If the scheme population have sufficient credibility to justify its own mortality table, then the use of such table could be appropriate. Unlike other decrements, mortality rates have consistently improved in the past. Past experience indicates mortality rates have continued to improve and that we can see in the available old and latest mortality tables which can be shown in Appendix II Fig: 2.1. In 2008, the Office for National Statistics published the 2008 based UK national population projections. Using these new projections results in a significant change from our previous assumptions, which has reflected the 2006 based population projections. Appendix I Table 2.2 shows comparison from 2006 to 2008 data of expected future lifetime values for male aged 65 years with and without future improvement in mortality. As we can see from 2006 to 2008 with or without improvement in mortality expectation of life have increased around 13.1%. The effect of a change in mortality basis on schemes liabilities depends on individual factor such as the age and gender profile, and the balance between current and former members and the benefits payable. A strengthening of longevity as explained above would be expected to increase the cost of accruing benefits in pension schemes by around 0.5% to 1% of pensionable pay and significant effect on PSB. But the actuarial profession publishes standard tables of mortality rates derived from extensive analysis of information it collects from insurance companies. So we will only focus on available standard tables. 2.3 Probability of survival and expected no. of deaths Appendix II Table: 2.3 and Appendix II Fig: 2.2 shows conditional survival probabilities for an individual aged 65. We can note that as mortality is improving probability of survival increases to higher age and people live longer and more benefits will be payable as they live longer. We can see from Appendix I Table 2.4 and Appendix II Fig: 2.3 how no. of deaths which have reduced over past years. We can also note that expected no. of death of person aged 55 in 1980 and in 2008 have dropped around 50% and expected no. of death of person aged 75 in 1980 and in 2008 have dropped around 40%. This definitely shows that people live longer and these will have effects on pension scheme and its pace of funding. 2.4 Expectation of Life Particular attention should be paid to assumptions about future mortality. Here the experience is of a sustained trend in one direction, that of longer life expectancy (i.e. decreasing mortality rates). We must use the latest available relevant data on likely future mortality rates. We can see the expectation of life of a male life aged 65 has increase from around 15 years in 1980 to 19 years in 2008. (See Appendix II Fig: 2.4) In the UK, where pensions are index linked (and for which the impact of changes in mortality are therefore more significant*), a pension liability calculated using a base (i.e. un projected) mortality table might typically increase by 30 percent when switching to the same table projected forwards to allow for future improvements, and by 35 percent when switching to a generational table based on the same base mortality table. (Figures are based on the approximate increases at age 65 for a UK male born in 1950. For older plan members and for current pens ioners, the impact will be less.) (See Appendix II Fig: 2.5) (* The impact of mortality improvement on index linked pensions is greater than on flat pensions because the amounts paid to those who live long are much greater.) Over the last 25 years the no. of people aged 65 and over in the UK has increased by 16 per cent, from 5.5 million to 9.8 million. In 1982 the population aged 65 and over accounted for 15 per cent of the population; by 2007 this had reached 16 per cent. There are far older people in the population than ever before. In addition, the older population itself is ageing. The fastest increases in population were seen for the population aged 85 and above (sometimes called oldest). Since, 1982 the no. of oldest have risen by nearly 680,000, to reach 1.3 million in 2007. Oldest represent 1.1 per cent of total population in 1982 and, nearly double in 2007, represent 2.1 per cent of total population. National population projections indicate that population ageing w ill continue for next few decades. By 2032 the no. in oldest population is projected to more than double, reaching 3.1 million representing 4 per cent of total population. The no. of people aged 65 and above is projected to increase by two third to 16.1 million, while the no. of people between ages 16 to 64 will only increase by 2.90 million. Based on projections, the population aged 65 and above will account for nearly 23 per cent of the total population in 2032, while the proportion of the population aged between 16 and 64 is due to fall from 65 per cent to 60 per cent. (See Appendix II Fig: 2.6) 2.5 Main drivers of past and future mortality Medical Advances: Medical advances have been responsible for a large element of current improvements in mortality rates. Smoking Trend: The prevalence of smoking in the UK fell from 51 per cent for males aged 16 and above in 1974 to 28 per cent in 1994, since when declined has slowed, with 25 per cent of males smoking in 2005, with similar trend in female smoking prevalence. Infectious diseases: Whilst recent medical advances and other factors have continued to lead a regime of increasing life expectancy, other factors work in opposite direction. These include the threat from new infectious disease and the re emergence of gold ones, such as tuberculosis, which may prove resistant to existing antibacterial agents. Uncertainty at young ages: Mortality rates in 1980s and 1990s increased for young ages as death related to AIDS, drug and alcohol abuse and violence more than offset improvement in health related causes of death at these ages. The cohort effect: Various explanations for cohort effects have been put forward like: Differences in smoking patterns between generations Better diet and environmental conditions during and after the Second World War Differing birth rate, with those born in periods of low birth rate facing less competition for resources as they age Benefits from the introduction in the late 1940s of the Welfare State These generations have benefited from medical advances which have increasingly affected older people. 2.6 Assumptions of Future Improvement in mortality and life expectancies Current annual improvements in mortality rates vary considerably by age and sex. The mortality projections assume that these rates will gradually improve. We will see how these improvements affects expectation of life considering period and cohort effect. Period life expectancy is the average no. of years a person would live based on age specific mortality rates throughout his life. It does not allow for any later actual or projected change in mortality. (E.g. life expectancies are worked out assuming pensioner experiences the projected mortality rates for age 65 in 2008, age 66 in 2008 and age 67 in 2008 and so on.) Cohort life expectancy are calculated using age specific mortality rates which allow for know or projected change in mortality in later years and it measures how long a person of a given age would be expected to live. (E.g. life expectancies are worked out assuming pensioner experiences the projected mortality rates for age 65 in 2008, age 66 in 2009 and age 67 in 2010 and so on.) Period life expectancies are useful measure of mortality experienced over given period and for past years. Cohort life expectancies, even for past years usually require projected mortality rates. Above mentioned period and cohort life expectancies are illustrated in Appendix II Table 2.5 and Fig: 2.7. It is noted that in 1981 the period life expectancy was around 13 years and cohort life expectancy was around 14 and in 2051 it has increases to around 23.7 years and 25.3 years respectively, which shows that pensioners live longer and that affects pace of funding. For example a person aged 65 now his expectancy is compared in Appendix I Table 2.2, which shows just in 2 years time how these life expectancies have changed. It might be appropriate to adjust standard mortality tables to reflect scheme, employer or geographic factors where evidence exists to support this. Consideration should be given as to the likely persistence of these differences in the f uture. Illustrations can be of variety of forms, such as: mortality rates at certain sample ages; risk sensitivity on funding methods by showing the effect of a 10% reduction in all mortality rates; life expectancy at certain sample ages; annuity factors at certain sample ages on period and cohort mortality rates; showing the discount rate shift which is equivalent to the mortality improvements adopted; for large schemes, showing charts of future funding levels based on initial funding, future contributions; small schemes might wish to illustrate concentration risk by, for example, scenario testing (e.g. by supposing those with the highest liability survive 5 years (say) longer than assumed.) Chapter 3: 3.1 Impact of different mortality tables on Funding Methods As discussed in chapter 2 of available mortality tables, we now analyze the impact of using different set of mortality on funding methods (i.e. on SCR, SF and RCR). While not wanting to increase employers costs needlessly, we might use up to date mortality when calculating pension scheme solvency or contribution requirements. Thus, there may be specific evidence (for example, a study of actual plan mortality experience) to suggest that plan members will experience heavier than average mortality. So how significant are the differences between the various types of mortality tables? Or more important, what is the impact on pension liabilities and costs of changing the assumed level of future mortality? The answer, of course, depends on the current mortality assumption and other assumptions. 3.1.1 Actual pension calculation We can assume one pension scheme which pays pension of 1/80th of final pensionable salary and 50% widows pension with assumed male members given in Appendix I Table 3.1 and we will analyze the effect of changing mortality table on SCR, SF and RCR. Other assumptions are. We can see from Appendix II Fig: 3.1 and 3.2, because of improvements the mortality has decreased from 1980 to 2008. There is increase in SCR and RCR due to increase in future accruals and benefits because of decreased mortality rates. If we compare these rates using similar mortality S1PMA, we can note that these rates are decreasing as we compare light to heavy mortality respectively. Similarly we can see in Appendix II Fig: 3.3 where SF has same effect, but as under attained age and projected unit method definition of SF is same so it will have same value but under entry age method it will always be higher. 3.2 Future Improvement in mortality For funding valuations where tables have not already been moved to a projected basis, companies are likely to face pressure from management boards, employee representatives and plan trustees to move to more realistic mortality assumptions, leading to increased contributions to the plan. If companies assume heavier mortality than their scheme members actually experience, shortfall in fund and actuarial losses will emerge as pensioners live longer than expected and it increases the pace of funding. Companies typically use the same mortality assumption for their funding valuations as for their accounting valuations. But there is an argument that the prudence assumption has to be built into a funding valuation to calculate contribution requirements and plan solvency which is not appropriate for an accounting valuation. The materiality of the mortality improvement assumptions increases when: The group being valued is predominately active lives. The scheme provides cost of li ving increases. If we underestimate future improvements the cost of the pensions to be paid in coming decades will also be underestimated. And the contributions made as financing will be too low, all other things being equal. On other hand, if people dont live as long as assumed we may have locked money unnecessarily. The developments over past few years lead us to choose mortality by two steps. First we decide upon base or current mortality for the pension scheme and then if required we apply some set of improvements to obtain assumptions for future years. There is tremendous change in population structure age wise as shown in Appendix I Table 3.2 and Appendix II Fig: 3.4 in 1981 around 20.2% of population was over age 60 where as in 2010 around 22.5% of population is over age 60 and it is expected that this will increase to 30.8% of population by 2051 which definitely shows how people are living longer which increases pressure on pension schemes and its pace of funding. For people from over pension age group, in 1981 it was around 10.0 million from total population and in 2010 it has reached to around 12 million due to improvements and these trend is expected to continue and no. of people in over pension group in 2051 are expected to be around 16 million (See Appendix I Table 3.3 and Appendix II Fig: 3.5). These growing populations will have large impact on pension schemes and its funding. 3.3 Suitable funding method After discussing effect of different mortality rates on funding methods we need to see which methods will be most suitable for funding pension scheme: Entry Age Method Under this method for contribution, only prospective members future benefits are considered so there might be accrued liability of present members which is unfunded. Future funding contributions are the entry age future benefits and adjustment for difference between accumulated assets and accrued liabilities. Attained Age Method As these calculations ignore accrued benefit, another calculation is required to compare the accrued liabilities with accumulated fund assets. The accrued liabilities are calculated in same manner under PUM and adjustment also made in exactly same way as under PUM. The important factor under this method is that it focuses on stability of future. Projected Credit Unit Method (Projected Unit Method) This is arguably the most important actuarial funding method, so it is importa nt to understand its fundamental objectives. The goal is to maintain the pension fund assets at such a level that, with future contributions the fund will be able to pay all accrued benefits until the last beneficiary dies. Each method has its own strengths weaknesses. The SCR and RCR as percentage of earning are obviously more under the attained age method. However the SCR and RCR under the entry age and projected unit methods are satisfactorily level as well. However if we compare SF, it is more under entry age method than attained age and projected unit method. And the use of any of these methods should not cause concern to pension schemes but the crucial thing is mortality assumption to use for these funding methods. Chapter 4: 4.1 Conclusions It is very difficult to justify mandating a single actuarial funding method with single mortality assumption. Employers in different industries or at different stages of their development (from start up to developed) will have correspondingly different funding objectives and experiences different mortalities. All funding methods described and discussed in chapter 1 are sound and systematic, but it is important to select ideal mortality assumption considering improving mortality. This leads to the following conclusions on the observations: Periodically updating the mortality table assumption to reflect current mortality levels with or without mortality projections or using mortality projected beyond the valuation date may accumulate assets closer to accrued liabilities. The stronger the mortality change, the grater the difference between funding methods if mortality assumptions are not updated on regularly. Defined benefit pension schemes that experiences consistent mort ality improvements and do not adjust or update the mortality tables may develop inadequate assets. This project suggests updating the mortality table will help maintain the assets level close to accruing liabilities. Mortality rates have decreased significantly over the last few decades, and the improvement continues. But there is considerable uncertainty over future trends in mortality. Companies that sponsor DB plans are left with difficult decisions about what allowance to make for future age improvements when determining costs. It concludes that there is considerable evidence suggesting that socio economic variables affect mortality rates, in particular there exists an inverse relationship between mortality rates and pace of funding. The study showed a strong inverse relationship between funding methods and post retirement mortality, and that this relationship diminished with age. To make allowance for future improvements in mortality, an adjustment can be made to the discount rate (investment return assumption). Additionally, as a matter of good practice, an adjustment to one factor to allow for prudence in another factor can be made. However, the interest rate deduction which is equivalent in its effect to the improvement factors adopted could be useful. 4.2 Recommendations: There are many ways in which we can tackle the problem of future improvement and decreasing mortality. One way of approximate allowance for uncertainty in future improvements and decrease in morality is to adjust other assumptions (for example, using a latest available mortality table, but reducing the discount rate by 0.5% to compensate for improvements). Another change can be made in pension scheme by capping pensionable salary growth to some fixed level to control the impact of pay rises on DB pension costs. Another can be longevity swaps which means derivative contract that offsets the risk of pension scheme members living longer than expected. This is a scheme that makes regular payments based on agreed mortality assumptions to an investment bank or insurer and, in return the bank or insurer pays out amounts based on the schemes actual mortality rates. Pension schemes keep assets and so retain investment and inflation risks. The swap counterparty is usually an investm ent bank, which then lay off all mortality risks. Pension scheme pay a fixed regular premium to offload the risks that they will have to keep paying out pensions to retirees for longer than planned for In the UK, the use of generalized linear models (GLMs) to help and understand longevity exposure in pensions schemes. Some suggestions that companies can do to manage the risk of decreasing mortality and future improvements: Review current processes: Many companies can manage the process by reviewing their selection of funding assumptions on regular basis. But, the important factor is that companies should make an effort to understand their schemes assumptions and how they relate to their specific experience. And they should thoroughly review their scheme and assumptions, possibly analyzing industry wide experience or segmentation of their scheme membership in more detail. Analyzing the risks: Carrying out a detailed analysis to identify and quantify the main risks in valuation and funding, including mortality, in providing benefits. This will help companies explore how these risks can be managed or mitigated. Buy out the risk: One of the important possibilities can be that to buy out the pension liabilities using group annuity insurance policies, where insurance markets are developed. In theory, this is one way to settle all aspects of the pension obligation (not only mortality risk, but also expense and investment risk) and may be the only way of terminating a plan. Insure against volatility: Reinsurance of longevity risk is one option, although this is very challenging to set up and has not yet been used extensively but, innovation is leading to other ways of carving up the mortality risks of pension funding more cost effectively. 4.3 Direction to further research In the process of preparing the project, several issues have come up which are to be considered for additional research. Is there a relationship between the concentration of retiree lives and appropriate projection of mortality table? It seems that choosing an appropriate post retirement mortality table is very important with regards to ensuring adequate assets, but how important are pre retirement mortality tables? There can be many mortality adjustments which can be analyzed based on following factors: Collar (White (Skilled), Blue (Unskilled), etc.) Income Occupation Country of Residence or other geographic location Presence of medical coverage It would be interesting to know how often valuing actuaries are changing mortality assumptions. One suggestion to find out: When the mortality assumptions were changed What was the new mortality table What is the size of pension scheme Is there any correlation when mortality assumptions are changed with the release of new mortality tables and when a mortality table is prescribed by law? These shows improvements in longevity come at a cost. Although there are wide ranging, socio economic costs associated with improved longevity, it is the pressure on pension schemes that has frequently caught the attention of governments, companies, and the general public. Employers have come under increasing strain as with limited amount of available capital is used to provide pensions for longer periods than expected when the schemes were designed.