Home Field Notes Operations Computers Regulation Expense Control Email Web Sites Phones Security Out Sourcing Sarbanes Oxley
Product Critical Illness Lapse Support Universal Life Burial Insurance Finite Insurance Leads Systems How To Stories Underwriting/ Claims   Using Consultants 
                    Last update March 9, 2005

FIELD NOTES

Critical Illness Insurance

Critical illness insurance is probably the newest product available from life insurance companies, having been introduced in the United States only a little over 5 years ago. It has a longer history in other countries, originating in South Africa in 1983 and spreading to the UK in 1986 and Canada in 1993. In the US it is usually presented as a rider on a life insurance policy, and provides for the "acceleration", or payment, of the face value of the base life plan upon the occurrence of specified major health "events". These events were originally the 4 "dread diseases", cancer, coronary artery disease requiring surgery, heart attack, and stroke. Early on in the UK competition among carriers took the form of adding to the list of qualifying events, spiraling over time to more than 25 events. In the US the list is usually shorter, usually the basic 4 plus kidney failure, organ transplant, paralysis and angioplasty, and less commonly, blindness and disability. Approximately 50% of the claims are from diagnosis of invasive cancer, 25% heart and stroke, 15% death, and 10% from the remainder of the events. The product thus gets its dominant characteristics from the familiar lump sum cancer policy and the old heart-stroke policy. All the other stuff is pretty much just for the sizzle.

Companies introducing CI need to be careful who they copy, since it was not adequately priced when first introduced in the U.S. Many of the original issuers have increased premiums about 40% in the last 2 years, but not all. There is also a trap in the tendency to consider CI "life insurance" when sold as a rider on a life policy. Since 80% of the claims will result from cancer, heart disease, and the other A&H type coverages, the product pays out money like A&H and needs to be judged accordingly. Assuming that CI is "life insurance" might be harmless if the only effect was permit sales management to report higher sales of "life insurance" as opposed to the less desirable "A&H". Unfortunately, other things can follow from getting the product into the wrong pigeonhole. For example, some companies are not tracking loss ratios on CI. While that is not useful for life insurance, it would be done without question with cancer and heart and stroke policies. There is also the issue of commissions, which are significantly higher for life than for A&H, which ultimately has to satisfy state loss ratio tests, sometimes retroactively. It is also evident that some companies that would not consider issuing a cancer policy with a benefit greater than $50,000 have no such concerns with CI.

Notes

Discussion

While Critical Illness policies have two decades of history in South Africa and nearly that in the UK, they have only become popular in the US in the last three years or so. The Accelerated Death Benefit rider, which pays part or all of the face of the life policy when the insured is determined to be terminal, was introduced in 1989 in the U.S., but popularity was limited, and it is doubtful that it will continue to be sold in any market where CI is available.

The surface similarity between the accelerated death benefit rider and the critical illness rider has led to some confusion. Both riders provide for the payment of part of the face amount of the underlying life insurance policy in certain events, but the events are entirely different and have entirely different consequences for the insurer and the insured. The exposure on the accelerated death benefit rider is measured by mortality plus the probability of paying the proceeds up to 2 years before death in the case of terminal illness (with some extra in case the diagnosis proves wrong). The exposure on CI is morbidity, primarily the incidence of cancer and heart disease. It is equivalent to combining the familiar lump sum cancer policy with a heart/stroke policy, and throwing in some sizzle, like kidneys, organ transplant, and disability.

The introduction of CI in South Africa was largely motivated by the absence of an adequate health care system, and the non availability of private medical insurance. Globally CI has been most popular in countries with poor health insurance systems and slow to grow in countries with good health insurance.

In the U.S. the appeal is probably more like that of cancer insurance, its half sibling. Most company sales materials and web sites emphasize a disability need rather than a medical cost need, whether or not disability is specifically a covered event. The marketing message, protecting one's lifestyle when unable to work, and providing extra income for living expenses, fits the disability approach.

In the UK mortgage sales, primarily by banks in the mortgage market, have outstripped non mortgage sales. That should be an attractive market for the risk adverse insurance company, since the CI is a rider on decreasing term sold primarily to younger buyers, minimizing exposure in the ages where most events occur.

Critical illness policies are available as life policies or A&H policies in most states. They are considered "life" when a rider on a life product. The main difference seems to be the commissions, not the claims. The benefits should be the same, with the exception of the term life insurance which might make up only about 10% to 20% of the claims. Many of the CI policies filed as A&H include a death benefit, which would make them indistinguishable on the benefit side.

Since 80% of the claims are A&H claims, watching the cash claim ratio (something not done on "life insurance") would seem mandatory.

For analysis it seems better to define a product by the benefit it provides rather than how it is filed. Critical illness products, rider or stand alone, pay about 50% of claims for cancer, 25% for heart and stroke, and the remaining 25% for the assorted other maladies covered, including death, of which death is usually about half of that 25% of claims. You would therefore expect the premium for CI to be about twice that of a lump sum cancer plan (but see the next paragraph). Some companies appear to be offering CI premiums that are substantially less than that. While there is not a lot of United States experience available for critical illness coverage, there is plenty on cancer policies, and on heart and stroke policies. When dealing with CI as a "life" product, consider that the first year commissions are 30 to 50 points higher than those on cancer products (usually offset by lower renewals). Claims will increase with age and duration. If the underlying life plan is the usual 10 year R&C, the term premiums will increase every 10 years, and on most plans the CI premium will increase also on renewal of the base plan. Both premiums usually increase to that for a new issue at attained age.

A 2002 ALUCA conference reported Australia CI claims from cancer, heart/stroke, other, and death, respectively for males: 36%, 33%, 12%, 19%, and for females: 71%, 8%, 10%, 11%.

The premiums on CI policies filed as A&H can be increased by the company if the loss ratios become too high, but will otherwise be level. The premiums on CI policies filed as life will increase in a predetermined amount, usually to attained age, on the renewal of the base plan, usually every 10 years. This makes an interesting difference in the premiums and the risks. The initial premiums on "life" filed CI should be lower, but the ultimate downside risk is greater.

The different premium patterns on CI filed as life from that on A&H complicate the comparison of CI and cancer insurance. While CI is clearly half lump sum cancer insurance, CI premium will increase every 10 years while cancer premiums will be level for life unless rate increases are approved, which will occur only if the loss ratios exceed a percentage established by the regulator, i.e. the company is losing money, probably serious money. Therefore both the risk of loss and the chance of gain is greater on the CI than on cancer insurance (and CI filed as A&H).

It is probably more accurate to compare the present value of future premiums of "life" CI to twice the present value of lump sum cancer premiums than to compare the initial premiums, but then persistency is so important to the calculation you just have another chance to go wrong. The fact that premiums will increase does reduce the lapse support on "life" CI, for better or for worse.

Is a critical illness policy a life insurance policy or an accident and health policy? That question is an example of the dangers of pigeonhole thinking. Labeling something a life policy implies one set of expectations and controls, while labeling it A&H results in an entirely different set. Either label, without thinking further, is going to cause miscalculations. A prime example is the cash loss ratio on CI. Calling it life or health doesn't change the claim pattern.

Apparently some companies that have filed CI as life insurance are not internally reporting the loss ratios on the product, something they would never fail to do if they considered it an A&H policy. Life insurance is such a long run proposition that loss ratios have no value, and even the familiar "actual to expected" mortality reports do not tell you much that is helpful. You never know whether you are seeing a pricing or underwriting problem or just a mortality fluctuation. Most A&H products generate claims much earlier, and the results become somewhat credible within several years. For example, underwriting has little effect on cancer claims, and after the first year, where selection against the company has a big influence, claim frequency is more dependent on age than duration. It is the aging of the block that causes loss ratios to rise over time, so analysis can show you whether you have a problem. An exception is the long term care policy. It is A&H, but like life insurance, you are going to be in deep trouble before loss ratios tell you anything.

Once you focus on the morbidity of the various elements of CI, there is no reason the product should not be profitable if priced correctly. It seems clear that the product was under priced by the early introducers, possibly due to confusion with the earlier Accelerated Death Benefit offered on a life insurance products.

With the latest round of price increases, implemented mostly in late 2002, CI looks like it should be profitable. Most companies are paying life commissions (higher first year) rather than A&H commissions (higher renewals), so profitability will require persistency to be high enough to allow the lower renewals to offset the higher first year commissions.

However, the industry will have to watch that other "mix" of qualifying events that is assumed to make up only 10% to 15% of the claims. CI is like a baked potato. What it does to you depends upon what you do to it. Including disability paid as a lump sum seems unsound, at least in the payroll market, and alzheimer's could certainly become a joker, since it is hard to define, much less diagnose, while a claimant is alive.

Critical Illness. Do the mathematics of the effective level commission apply to the critical illness product filed as a "life" plan? The idea is that the levelized commission plus the loss ratio and other expenses shouldn't exceed 100% of the premium. It doesn't make sense to talk about the loss ratio of a true life policy, but here we have a policy where about 85% of the claims are on the A&H benefits. A state requirement makes the calculation easier, but just because there isn't one on CI filed as life doesn't mean that the loss ratio can be ignored.

The first year commission on a life plan is higher, usually well over 100%, but the renewals are usually less than on A&H. Logically persistency on CI should be about the same as on cancer or heart coverage sold in the same market, but the effect of dressing it up as a life policy is unknown. If that improves persistency, it will reduce the effective level commission. If your CI premium is about double the cancer plan you are using as a comparison, the CI loss ratio should be about the same as the cancer plan. Since half the claims cost will be cancer claims, there should be a pretty good correlation.

So how does it come out? If first year is 120% and renewals 10%, it takes about 4 years to come back to a level commission of 40% with 100% persistency and no adjustment to present value. In most markets the level commission will be higher. Add 40% to a 50% loss ratio and you don't have much room.

 

I believe the company with the most Critical Illness experience in the United States is Colorado Bankers. I asked Dave Pavletich, the CEO, to discuss his recent increase in the premiums at which CI is offered. I haven't compared the old and new rates, but an agent indicated the increase was about 40%. That mirrors what has been happening recently in the UK, where CI has been offered for longer. Those increases are about the same, in the 40% range, and include restrictions on renewability.

There may be an issue of whether morbidity in the work site market might be different from that in more affluent markets, for example, due to the frequency of smoking. I personally doubt this, as the major appeal of CI would logically mimic that of cancer insurance, which is has historically been to the less affluent buyer.

COMMENT, Dave Pavletich, CEO of Colorado Bankers Life: We started issuing critical illness in July of 1998. I think Canada Life may have been out in the U.S. before us , but I'm pretty sure that we were the first to sell CI in the less affluent markets , and the first in work site. We implemented a rate increase effective 10/1/02 for our Critical Illness product. We received very little resistance from our distribution. Since our product is sold pretty much on a money purchase basis it actually looks like a benefit decrease to the field. The results risk wise have been great . The face amounts for the new product have decreased 31% compared to the old version. This is determined by running a program against our in-force policy data base , so it's not hypothetical , it's real. And the average annualized premium ,old vs. new , has gone from $400 to $440. This was very surprising and we attribute it to our agents selling a higher premium to get the face amounts up.

Even after this change we're still competitive with National Travelers ( a work site competitor) and Guaranteed Trust Life , a new player in the CI market.

How clear are your Critical Illness definitions? Is it reasonable to believe the policyholder understands them? If you and your lawyer are satisfied with the clarity, the next step is to find out if the sales brochures and the presentation used by the field are as clear as the policy language. You might also ask some of your employees, since people like them, not the lawyers, will be the ones on the jury.

There is plenty of good material on definitions on the internet. Try "critical illness"+definitions in your search engine. Example: http://www.cis.co.uk/Pages/critical_illness/media/PY181crill.pdf. Many issuers have sites on the subject. I think one lightning rod of potential lawsuits is the definition of cancer, which excludes non-invasive cancer or cancer in situ. Do YOU know what that means? The agent told the applicant, or the applicant heard, that the policy covered cancer. I really don't know how you protect yourself here, but your agent training materials could be important here. It goes without saying that the Q. & A. on the sales brochure better have the question such as "Am I covered if I get cancer" and the answer better be "maybe".

The variations in critical illness plans in the U.S. is currently more limited than in the UK, where competition took the form of adding more and more covered diseases. There is some indication the US companies will follow that trend, as the list is getting longer, "severe" burns and "loss of independent living" auditioning. More definition problems.

You can almost hear the marketing meetings churning out CI choices. According to CheapLifeInsurance.ca, you can have basic coverage, which is merely heart, stroke and cancer, or the comprehensive which covers up to 20 major medical events. You can have it with or without return of premium at death or maturity, and even after 10 years. And of course, you can have it on a variety of life plans or just by itself. The site is worth reading. It claims that if you are unable to get disability insurance, CI is the "next best thing". And you can buy it with waiver of premium, so "you do not lose your critical illness insurance on account of disability."

You will probably get the best results from your CI plan if you limit the illnesses covered to the basic four, cancer, coronary artery disease requiring surgery, heart attack, and stroke. It is reasonable to add kidney failure and organ transplant, simply because it has become almost customary and claims should be rare. Steer clear of problems that are impossible to diagnose with certainty such as alzheimer's and disability, and procedures that have become routine (and really not critical any more) such as angioplasty.

Throwing in diseases that hardly every occur and nobody is concerned about doesn't get you anywhere, and may hurt. Agents and consumers know that is just fluff, and the agents probably never mention the irrelevant extras for fear of discrediting the sale, or making the prospect laugh. Also dangerous is jumping on the latest trend, where there is no credible experience, just to get the agents excited. Some company somewhere is going to add West Nile Fever. That will sell all right, but they may not be around in a couple of years.

Also dangerous, but rare, is omitting coverage of one or more of the basic diseases the consumer could reasonably expect to be covered by anything la bled "critical illness". For example, if you omitted cancer because you already have a cancer plan and you want to avoid cannibalizing those sales, you are probably going to be involved in a lot of litigation, and my guess is you will lose.

 

CI coverage is generally offered as a rider on a life insurance plan, usually on a base plan of 10 year R&C term. It is less frequently offered on whole life and the type of decreasing term used a mortgage insurance. The type of base plan should make a big difference in the pricing. The highest premium would be on the whole life rider, when the face and premium on the rider are level for the duration, and the duration is the longest. The lowest premium would be for the rider added to decreasing term, when the premium stays level while the benefit drops along with the face of the base. Somewhere in the middle is the rider on 10 year term, which will deliver a premium increase for the rider at renewal.

Marketing pressures, real or imagined, seem to mitigate against fully recognizing the necessary pricing differences created by variations in the base plan. If the consumer, and more problematically, the agent, do not appreciate the difference, then the CI offered on the longer duration base plans is going to look unduly expensive, and the rider on mortgage plans, very cheap. It appears that some companies have responded to this by using one rider premium set for all the base plans they offer. Companies that distribute through exclusive agents may be able to survive this if the mix of business is fortunate, but those distributing through agents that write for several companies will have problems.

CI sold on a base of UL will probably have to abandon either the practice of defining the CI coverage as a percentage of the face of the base plan, or the flexibility of the UL. If the CI benefit is equal to the cash value, it is going to be a tough guess for the actuary as to whether the benefit will increase or decrease over time.

When a critical illness rider is added to a base plan, most commonly the 10 year R&C term, a form of lapse support is introduced which will either increase the profit on the base plan, or permit a lower premium. This is easiest to see with the 100% CI rider. When the policyholder incurs a payable event (such as cancer), an amount equal to 100% of the face of the base plan is paid, and the coverage on the term plan is reduced by that amount, in this case, to zero. There is no further exposure on the part of the company, and yet it has received premium on the base plan (in addition to that on the CI) to the date of the event.

Based on the conversations I have had regarding this concept, it is not all that obvious. Consider it this way. Compare the morbidity on a 100% CI rider with a stand alone CI plan with the same event design and face, and, in theory (as I know of no example), a CI rider that does not reduce the amounts payable on the base plan with a subsequent death. It is apparent that the claims paid on the CI will be identical on all three versions. Yet on the carve out rider the company has received life premium for an exposure that was eliminated upon payment on the CI rider. You could also argue that the base plan exposure is eliminated on the highest risk individuals, ones that have cancer, heart attacks, and other payable events, but I am not quite sure how that works out and will have to check with my actuary friends.

The company is likely to capture this extra profit, since the CI is an optional rider in most plans, so the term plan has to be priced as a stand alone product. You might expect stand alone CI to be more expensive than rider CI, but I can't see that in the premium comparisons. That may be disguised because the prices charged for CI are still all over the map

 

Maybe it is worth restating the above paragraph set, not only because it is not intuitive, but also because it offers an opportunity to survive the price wars on 20 year level term, the internet favorite. Start with this question: you have 3 different CI products, a stand alone, a rider that reduces the face of the base plan by whatever it pays (a "carve out"), and a rider that doesn't reduce the face of the base plan (an "in addition to"). If you ignore the profitability of the base plan and for the moment consider only the CI, which CI is the most profitable? Right, they are all the same. The benefits and the premiums are identical, so also is the profit.

Now let's consider what happens to the base life plan when a payment is made on the CI carve out rider. Let's say, for clarity, that the CI is 100% of the base plan face. When the CI pays the 100%, the base plan has zero left, and that is the end of it.

Now suppose you could issue a 20 year life term policy with a provision that if the policyholder was diagnosed with cancer, or had a heart attack, the policy was immediately canceled, and paid nothing to the policyholder, with you keeping the premium paid to the date of the event. Sounds more profitable than the term you are issuing now, doesn't it? Too good to be true? Nope, that is exactly what you are issuing when you issue a base plan with a 100% CI rider.

That cheap term offered on the internet is of doubtful profitability, but if you feel you have to compete with it, you might offer your product with a CI rider built in. That ought to increase your odds of surviving the current pricing. Note that the competitive pressures on term are not present on CI pricing. Comparisons are not so easy, and all you have to match is the pricing on the stand alone CI, which has to carry its own weight.

It seems reasonable to expect a lot of replacement activity, CI replacing cancer insurance already in force, and to a lesser extent, other A&H products, particularly in the worksite market where it is easy for the agents to return to offer the "upgrade". The negative effect on profitability caused by the payment of another first year commission on a similar product will be exacerbated by the higher first year commission on "life", about 110% or more instead of the 50% - 70% common on cancer plans.

Replacements are of course agent driven, so the relative merits of CI and cancer insurance will not be as important as the effect of company replacement rules on commissions. Since most companies are thinking of CI as a "life insurance" policy, and cancer is A&H, most replacement rules will not by their present terms apply. The full first year commission is likely to be paid on the replacement, motivating the agent, and the company is likely to record the replacement as new life insurance sales, motivating company management. Even management that understands the profit impact will have a tough time against the argument that "if we don't replace it, some other company will".

In addition, most companies that are heavy A&H writers have tried for years to get their agents to write more "life insurance", with incentives like bonus, recognition, and convention qualifications. For this CI is wonderful, a product with which the agent who likes to sell A&H is right at home and which the agency management can report as life sales. It doesn't get any better than this!

There are some CI problems introduced by state filing restrictions. Possible the most serious is the waiting period, which many appear to be limiting to 30 days, and a few not allowing any at all. Cancer is very anti selective, and a 90 day waiting period is recommended. Another danger area are future loss ratio requirements.

CI was first introduced in Japan without a waiting period. Severe anti selection was demonstrated in first year claims triple the second year claims. The policies were quickly changed.

Several states have loss ratio requirements that apply whether the policy is filed as life insurance or as A&H. In other states CI is treated like AD&D: if filed as a stand alone policy it is A&H, with loss ratio requirements, if filed as a rider on a life policy it is life, without loss ratio requirements. This really doesn't make much sense, and there is the risk that the regulators will tumble to the true nature of the coverage, defined by the claims it pays, and apply loss ratio requirements retroactively. With some cancer and hospital indemnity plans that has resulted in mandatory refunds.

 

CI, like cancer insurance, long term care, and most other level premium A&H coverage is to a significant degree lapse supported. If later duration lapses are lower than expected, more of the population will get to the higher claim periods, and margins will suffer.

How about lapse rates? Are lapse rates important? I see several people nodding their heads up and down. The ultimate lapse rates could be quite low in the individual market. We don't know what it's going to be like, but in other countries, the ultimate lapse rate is quite low. Remember, this product is lapse-supported. There are no cash values and it's level premium and the benefits go up rapidly with age. It's a very highly lapse-supported product, and for worksite and direct response business, there should be high initial lapse rates, then lower ultimate rates are possible. The lapse rates for group life should again follow the group term life policy pattern.
Abraham Gootzeit, http://www.soa.org/library/record/2000-09/rsa02v28n37pd.pdf
This is an excellent report from the Boston SOA annual meeting October 2002, moderated by Susan Kimball, with Panelists Abraham Gootzeit and Tom Ming.

The amounts of CI coverage that can be purchased, upwards of $1 million from some companies, are far greater than the limits have been on lump sum cancer policies, and lump sum heart-stroke policies (which together make up about 75% of the CI claims), which were usually around $50 thousand. Lump sum policies carry a great anti selection risk, as do larger size policies.

There are other reasons most companies have chosen to limit the amount of specified disease policies they will issue. There is some thinking that the amount of premium a household devotes to such less worthy coverages, should not rise to a significant percentage of what a comprehensive health policy would cost. In others there is the fear that easier to sell (you don't have to die to collect) A&H will cannibalize life sales, or at least divert limited paying capacity from needed life insurance protection.

There is also a persistency problem. While large life sales tend to have better persistency, larger A&H sales, at least of specified disease cover, tend to have much worse. If the thought is "Oh good, because CI is a life sale", consider that while persistency can be influenced by many factors, the label under which something is filed is not one of them.

Reinsurance on CI can be a real problem if substantial amounts of coverage are offered. See page 29 of this 2001 discussion of reinsurance by Munich Re. What you would like to do, of course, is keep your normal retention on the base life policy, but reinsure a higher portion of the unfamiliar CI risk. While there is talk of excess coverage being available, it appears that the only practical quotes are for quota share in equal portions of the base and the CI. If you give away a substantial portion of the CI risk, you may have to give away enough of your normal production to nullify the extra premium production.

COMMENT, Cary Goggin: Reinsurance on CI is a key element. I called virtually every reinsurer at one point and found 5 that seem to be 'CI active'. ERC, Munich American, Optimum Re, General Cologne, ING. For a CI rider attached to life, the only thing I found to be available was coinsurance on the base and rider together. Rate guarantees run from 1 to a max of 5 years.

The Munich Re discussion also suggests that insurers limit their rate guarantees to a maximum of 5 years. That is not happening in the US market, as CI riders are attached to 10 year term, and policies of longer duration.