Reviewing Disability Income Experience in New Zealand
Gen Re has been collecting data on disability income (DI) policies individually written in New Zealand since 1994. The latest study covers the calendar years 2008 to 2012, examining claim incidence, termination and payment rates, and this article shares high-level findings of that experience. The results of the investigation were also presented individually to the participating companies, and to the industry as a whole, at the New Zealand Society of Actuaries Biennial Conference in November 2014.
Eleven companies representing approximately 90% of the DI market by in-force premium contributed data to the study. The total number of claims over the four-year period was 7,755 and there was over 922,621 life years’ exposure. Findings of the study included:
- Incidence rates have been stable and improving for females.
- Termination rates have been weaker for longer benefit periods.
- Lower benefit offsets exist in accident claims.
- Experience worsens by policy duration.
- Larger policies and longer benefit periods have worse experience.
The New Zealand market
The New Zealand life and health insurance market is characterised by sophisticated products and a high level of competition between insurers. There is a significant reliance on independent financial advisers for distribution of products, as evidenced by high levels of upfront commission. For lump sum insurance, these levels can exceed 200% of first year premium; for DI it is slightly lower, at around 180%.
Traditionally, DI is available in two types, each with a distinct benefit calculation at time of claim. Agreed Value (AV) policies pay the sum insured irrespective of income at time of claim. These policies make up 29% of the total in-force data. Indemnity (IN) policies pay the lower of either sum insured or 75% of income at time of claim. These policies make up 33% of the total in-force data. Therefore, IN policies pay a lower benefit amount where the insured’s income has fallen between policy inception and claim, and are therefore less expensive than equivalent AV policies. In both types, other sources of continuing income, such as statutory benefits and other insurance policies, are offset from the claim amounts paid.
Mortgage cover has been gaining popularity in recent years and accounts for 25% of the total in-force data over the period. The sum insured and financial underwriting, in this case, are linked to a home loan repayment. The monthly benefit is usually of an AV nature and other income is typically not offset against this benefit at time of claim. In the last five years, it has become increasingly common for the mortgage cover products to be delinked from an actual loan. For example, the sum insured usually does not relate to the mortgage repayment and the policy can be purchased at any time.
Other DI benefit types make up about 12% of the total in-force data over the period. The definitions of total disability are generous compared to other markets and usually allow the insured to work for up to 10 hours a week or perform one important income-producing duty.
Experience by time
The investigation analyses the actual claim experience and compares the results to a well-known industry table called IAD89-93. Published in 1996 and drawn from Australian experience, the IAD89-93 table varies by gender, age, cause, occupation and waiting period. The actual experience is expressed as a ratio of the expected result based on the IAD89-93 table. In practice, actuaries apply significant adjustments to the table, as it no longer reflects the actual experience well enough, either in terms of the overall level or by risk factors.
Sections below discuss first the experience of the traditional DI products – that is, AV and IN combined – followed by a comparison to the Mortgage Cover experience.
The incidence rate over the investigation period is 68% of expected when measured by policy count. The experience is higher at 78% when measured by policy amount, indicating a higher incidence rate amongst the large policies. We are seeing a downward trend in incidence rates over the investigation period, especially among females (see Figure 1). However, when we look at the incidence rate over a longer period, since our first study in 1994, the downward trend is obvious for females only. The incidence rate for males has remained similar for the last 10 years.
The termination rate over the investigation period remained roughly steady, with 2011 termination results higher than other years (see Figure 2). The overall termination rate is 61% of expected when measured by policy count. The experience is slightly lower when measured by policy amount, indicating a lower termination rate amongst the large policies.
A more detailed look at the termination rate discloses that people who have been on claim for longer than two years are terminating at a much slower rate than seen in previous study periods, even though the overall termination rate remained steady. There has been a downward trend since the 2000-2002 study (see Figure 3).
A unique feature of the market is the role of the Accident Compensation Corporation (ACC). The ACC provides comprehensive, no-fault personal injury cover for all New Zealand residents and visitors to the country. It offers a wide range of services, from payment towards treatment to help around the home, and pays up to 80% of an employee’s pre-accident earnings until age 65 through a feature called CoverPlus Extra. Self-employed people can select a lower level of cover.
Depending on the provisions of the insurance policy, any insurance claim lodged may be offset by the amount of ACC benefit received. Offsets from the ACC are often the biggest source of claim offsets for accident claims. A payment rate below 100% indicates the claim is receiving an offset from one or more sources.
Offsets from other sources are also possible, including:
- Partial disability payments, whereby a benefit is paid in proportion to the reduction in pre-disablement earnings
- Payments under indemnity policies, whereby the total disability benefit due is reduced in line with pre-disablement earnings
- Payments the claimant receives or is entitled to receive for illness or accident from the insured’s employer, a superannuation plan, other disability policies or other government-authorised income
There is an upward trend in the payment rate for accident claims, while the payment rate for sickness claims shows no obvious trend (see Figure 4). It is reasonable to assume that the level of claim offsets from the ACC has lowered over time, increasing the claim cost of accident claims.
An industry experience investigation on Mortgage Cover was conducted for the first time in this study. Both the incidence and termination rates for Mortgage Cover policies have been better than the traditional DI products. The overall incidence rate for Mortgage Cover over the investigation period is about 15% lower than the traditional DI products, while the overall termination rate for Mortgage Cover over the investigation period is about 25% higher than the traditional DI products.
The payment rate has also been higher for Mortgage Cover, almost 20% higher than the traditional DI products; this is expected, as most Mortgage Cover policies do not offset the payments from the ACC.
When comparing the claims experience between Mortgage Cover and the traditional DI products, it is important to note these differences:
- Mortgage Cover policies are, on average, smaller in size and shorter in benefit periods.
- At the same time, they tend to be younger in policy durations, reflecting the nature of a newer product.
- Mortgage Cover claims are younger in claims durations.
Experience by policy duration
We are seeing lower incidence rates in the first few policy years supporting a selection effect, a selective lapsation effect, or both. Selection effect reflects the positive effect from underwriting at policy application, which wears off over time. Selective lapsation reflects higher lapses among healthier policyholders who can more easily switch to another insurer that offers a better product, or a lower price, leaving behind the less healthy, who are naturally more likely to claim.
The incidence rate for policies older than five years is almost 20% higher than the incidence rate for policies less than five years old. Some older policies may have different product terms and conditions and underwriting to the new policies, possibly dampening the effect of any potential selective lapsation effect (see Figure 5).
Experience by benefit size
It is a common market practice to offer discounts for large sums insured. However, experience does not support such a practice. Generally, people with larger sums insured are more likely to claim and less likely to terminate once they are on claim (see Figure 6). The standard errors at the 95% confidence intervals are also shown, indicating sufficient numbers of claims in each sum insured band.
One of the potential drivers of this experience could be the difference in portfolio mix among the different sum insured bands. However, the pattern of worsening experience with increasing policy size is still evident when analysed further by gender and by occupation.
There are many other potential drivers behind this pattern. It could be argued that a 25% drop in income post disablement has a greater impact on the living standard of someone earning less than NZ$2000 per month than those earning more NZ$6000 per month, reflecting the diminishing value of money.
Experience by benefit period
Experience also shows significantly higher claim cost for longer benefit periods. Policyholders of longer benefit periods are more likely to claim, and once on claim, less likely to terminate (see Figure 7).
Typically, claims with long-term benefit periods receive more attention from claims managers as they could potentially become the more expensive claims. For that reason, long-term benefit period claims could arguably produce better termination rates than short benefit period claims. However, experience does not show such a pattern.
We analysed the termination experience further by duration since claim and benefit period (see Figure 8). Claims have higher termination rates nearing the end of their benefit periods, suggesting that claimants become self-motivated to return to work as insurance benefits are ending. Claims with two-year benefit periods outperform claims with five-year and to-age-65 benefit periods in the second year (H3 and H4). Similarly, claims with five-year benefit periods outperform claims with to-age-65 benefit periods in the fourth and fifth year (Y4 and Y5). These findings highlight that (dis)ability is best understood in terms of a combination of biological, psychological and social factors rather than purely in biological terms.
Better insurance product design is paramount in managing the difference in experience between benefit periods and responding to the changing landscape of the New Zealand DI market. Some effective changes include the following:
- Objective and clear definitions to improve claims assessors’ capability to interpret and apply policy definitions
- Mandatory compliance with the treatment and rehabilitation program from the insurer
- Ability to offset deemed income for those able to return to work on a partial basis but choose not to do so
An industry study on disability income provides valuable insights to all concerned. This is especially true where no individual company has significant experience on its own, such as termination rates for long duration claims.
Gen Re provides detailed feedback to participating companies, providing valuable insights into the evolving experience in New Zealand and thus helping the industry to better manage this risk. This survey not only helps to price the product more accurately going forward but also to channel claims management resources where it is needed most.