Pricing of critical illness insurance - art or science?

The starting point for pricing any insurance product is to estimate the likelihood of an event taking place and the expected payment to the customer if that event occurs. Fortunately for critical illness insurance, the amount of the payment is agreed upon at the outset, so the role of the pricing actuary is to estimate the likelihood of an event in each future year.

To do this, the actuary will construct a table of risk premiums, by age and gender, that predicts the probability that an individual will make a claim in the following year. To estimate these premiums, some data must be accumulated, which, in practice, may be more difficult than it sounds. Early critical illness rates in the UK were based on population hospital episodic statistics. Even today, when accurate data are available, the actuary must be wary of poor-quality, biased data, and then adjust for a numbers of factors including:

- The illnesses to be covered - for some conditions, the data may not be available or there may be illness where the experience is included in the data but the policy is not covering the condition, for example, for minor cancers.

- The fact that insurance is underwritten - how much should the expected claim experience be reduced to reflect the fact that insurers have a healthy subset of the general population?

- Allowance for the fact that those buying insurance will generally be from a higher socio-economic group and, consequently, are likely to be healthier.

- Multiple events - most policies only pay out once, even if there is more than one event. The data would treat this as multiple events, but the insurer would only pay one claim.

- Sudden deaths - an example of this would be someone having a heart attack and dying before reaching a hospital. These cases are not included in hospital data but must be charged for in the insurance premium.

- Any other rating factors - for example, in the UK, insurers quote separate rates for smokers and non-smokers. This is not split out within the population statistics, so a view must be taken as to the relative experience of the two classes.

- The impact of anti-selection - this reflects the fact that applicants may be in a higher risk group and know something that the insurer does not, for example, family history or diagnosis of a medical problem in another country.

When offering guaranteed rates, the anticipated trend in future claim experience must be reviewed. Past population trends are available but, as can be seen in figures 1-3, the world is moving and so the past needs to be interpreted carefully when considering what might happen in future.

Additionally, some policies do not pay out in the case of death from the critical illness, while others do. It is important to allow correctly if someone experiences the critical event and then dies shortly afterwards.

A further problem is that of 'incurred but not reported' (IBNR), well-known in the world of general insurance but not usually a problem for life insurance actuaries. The longest delay seen in the UK was a stroke claim, seven years after the event. This is less of a problem if IBNR is within the data and reporting is stable. If neither of these applies, care must be taken to load the rates accordingly.

Against this backdrop, it is perhaps not surprising that early products sold more than 20 years ago had premiums that could be reviewed, high margins and covered a relatively small number of conditions. As the market has become more mature and data has become available, actuaries have more confidence in their pricing, so margins are falling, rates are guaranteed for longer periods and more illnesses are covered. The UK is fortunate in that not only are there a number of reinsurers with good data, but there is a centralised bureau to collect industry statistics.

The process described previously is very much a case of actuaries looking at data and analysing data, and adjusting that data to make it relevant to a specific situation. Throughout the process, a number of judgement calls and a huge dose of common sense are required. In addition, there are a number of less obvious problem areas that may have an impact on future claims experience.

The medical world

Changes in diagnostic techniques will mean that some conditions, particularly cancer, are detected sooner rather than later. This may mean more claims for the insurer but may also reduce the number of claims if the detected cancer is not quite serious enough to meet the definition. In turn, this can lead to disputes and customer dissatisfaction where customers would have received payouts if they had not sought treatment. This could lead to adverse publicity, with consequential damage to reputation for the industry. The problem would be made much greater if the government-sponsored nationwide screening programmes were not allowed for in the pricing.

A changing world may mean that there are alternative treatments available. Suppose that the policy covers open heart surgery and that a simpler treatment is developed. Would the insurer deny the claim, just because the policyholder has a less painful but equally effective treatment?

In another example, a doctor may have diagnosed a condition - for example, a heart attack - which is not serious enough to meet the insurance company's definition. Again, the insurer is in a difficult position of disagreeing with a medical professional.

The legal world

Given the complexities of the definitions and some of the issues outlined above, it is not surprising that there is potential for dispute between insurers and their customers.

This is exacerbated with claims also being declined due to customers not declaring material facts when applying for cover. This has resulted in relatively high numbers of claims being declined, and arguments with the consumer as to what a 'material fact' actually is.

In the US, the tight legal framework, the cost of legal advice and the attitude towards punitive damages are among the primary reasons insurance companies are reluctant to offer this product.

In the UK, there is an Ombudsman - an arbitrator appointed by the Financial Services Authority - to adjudicate on disputed claims, which provides a relatively quick and inexpensive way to resolve disputes. However, there is a problem in that the Ombudsman is increasingly prepared to 'err' on the side of the consumer, which may not be surprising but must be allowed for when setting premium rates. Recognising the problem that the legal world can change quickly and unpredictably, the pricing actuary must question the appropriate allowance to make.

At the end of the day

Critical illness insurance has been a major success in the UK market, in terms of sales and in terms of providing extremely valuable cover for our customers. UK insurers pay out more than £400 million in claims per year. Uncertainty in data reliability, coupled with ever-changing medical and legal environments, continue to make critical illness pricing extremely difficult. However, actuaries can better price policies, resulting in more predictable experience by artfully applying sound judgement to accepted theory.

Contact

Jason Hurley, Head of Sales and Marketing

RGA UK Services Limited

T: + 44 (0)20 7448 8224

email: jhurley@rgare.com

www.rgare.com.

Sponsored content