Long term care insurance in the UK being a new product, means insurers generally look abroad or turn...
Long term care insurance in the UK being a new product, means insurers generally look abroad or turn to data based on the whole of the population in order to calculate their premiums. It is only recently that actual claims experience is letting actuaries assess how accurate their pricing has been. This article describes the current approach taken by insurers to calculating LTCI premiums and at early claims experience, before looking forward and considering what might happen to premium rates in the future.
Pricing without data
With no UK data, the ideal approach would be to use data from abroad. Unfortunately, the products sold in the most likely candidate the US are too different from UK products to be of much use. Insurers need to look at UK population data.
The primary data source used is the 1985 Office of Population Census and Surveys (OPCS) disability survey The Prevalence of Disability Among Adults. This survey was commissioned by the Department of Health and Social Security to provide information which the Government could use to guide policy on disability benefits.
Almost 100,000 people were surveyed. Questions were asked on the levels of disability in 13 distinct areas of day-to-day functioning. These included mobility, reaching and stretching, dexterity, personal care, continence, intellectual functioning and so on. Individuals were scored on their level of disability in each of these areas, and the three worst scores were combined to produce a final summary measure of severity of disability. The survey estimated that there were 5.7 million disabled adults, although most of them experience relatively minor disabilities.
Unfortunately, the data from the survey is not in the right format for insurers it needs to be re-analysed before it can be used. The main problem is that the survey did not explicitly ask about the six insurance activities of daily living, or ADLs, (mobility, washing, dressing, feeding, toileting and transferring). A considerable amount of judgement is required to decide how the 13 disability areas relate to the six ADLs. As an example of this process, the box above shows examples of people who would and would not be counted as valid LTC claimants.
Constructing a model
After this re-analysis has been done, a number of further adjustments are still required. Firstly, underwriting removes high risk lives or those already disabled. Secondly, an allowance has to made for trends between the date of the survey (1985) and the present. Finally, and most importantly, the effect of social class is included. In general, people who purchase LTCI products come from the wealthier part of the population, and these people are less likely to experience disability than the population as a whole. All of these factors act to reduce the level of disability assumed.
Once this 'theoretical' model of disability has been constructed, premiums can be calculated. It should be clear that this pricing approach requires considerably more judgement than for other types of insurance. Where there is uncertainty of this kind, insurers need to take a cautious approach to pricing, which translates into higher premiums for the client. This is particularly the case where guaranteed premiums are being offered.
Theory vs reality
How does the early claims experience compare with the theoretical approach? Does it suggest that premium rates are likely to go down or up?
There have only been a small number of claims to date and pricing actuaries have to take care when drawing conclusions from the limited data. However, the claims are showing strong patterns and the good news is that these early signs suggest that premiums may come down, perhaps by as much as 10%. This should not come as a great surprise. Firstly, these reductions would move prices in the UK closer to those charged in the US. Secondly, as the US market developed, premiums came down as claims experience came in.
Can we really expect prices to come down? To answer this question, pricing actuaries need to consider more than just the current claims costs. They need to consider the future.
Falling term assurance rates, lower annuity rates, the guaranteed annuity option problem everyone is aware that mortality rates are falling. However, recent research has identified that while mortality in general has been falling, one group has done particularly well. This is the group of people who were born in the 1930s and are retiring now. They are a 'cohort' who have experienced much lower mortality rates than their predecessors. For example, the mortality rates of 70 year olds today are 20% less than the mortality experienced by 70 year olds a decade ago. This means the cohort will live much longer on average. By coincidence they also make up a large proportion of LTCI purchasers. Possible explanations for the cohort effect include better childhood health, the effect of active service during the war, the introduction of State education and the NHS, as well as changing smoking habits.
Is this increasing longevity a good thing? The main issue for LTC insurers is a simple one; policyholders may be likely to live for longer, but will these extra years of life will be spent free of disability or will they be spent in poor health?
The potential impact on premiums, and profits, is huge. Consider the simple example in the graph above. In the example, life expectancy is currently six years, of which four and a half are spent healthily and one and a half are spent disabled. Imagine that life expectancy increases by one year from six to seven years. It shows scenarios of how these extra years of life might be spent, and what the implication would be for the single premium of an LTCI product.
Changing disability
The worst case scenario is where all the extra life is spent disabled, where single premiums would need to increase by more than 60%. Where the extra life is healthy, there would be little change to premiums.
Information is beginning to emerge on how disability rates have changed in the UK over recent years and these are the best guide we have to the future. It seems that for severe disability, the time spent healthy is rising and the time spent disabled has stayed the same. However, this is not the case for more moderate levels of disability. It seems that much of the extra life is spent suffering from mild, but not serious disability. This pattern is being seen not only in the UK, but across the developed world. Once again recent experience seems favourable. Extra years of life appear to be healthy ones.
We are beginning to see evidence for both current and future rates of disability which support reductions in LTCI premiums, perhaps in the region of 10%. While the levels of reduction are not likely to make LTCI an affordable, mass market product, they would give it a useful boost.
Reductions would also call into the question the value of guaranteed, rather than reviewable, premiums. This good experience should also benefit policyholders with reviewable products. They may see premium reductions when their policies come up for review this is the type of message the long term care insurance industry would be happy to be able to give.
Ross Ainslie is product actuary at General & Cologne Re
