You can search for articles in back issues of Contingencies from July/August 2000 to March/April 2009 using the search box to the right. Simply type in subject words, author's name, or article title and click search. To search for articles from May/June 2009 to the present, go to the current digital issue of the magazine and use the search function on the left of the top navigation bar.
Exploring Underwriting Ramifications
By Douglas A. Ingle
The internet and financial services industry's focus on instantaneous transaction processing has influenced consumers’ expectations about the life insurance industry. The need for speed drives the financial institutions. This philosophy, when applied to life insurance, has created a growing market for instantaneous, or quick-issue, delivery of life insurance policies.
If only to suit customers’ new demands for fast answers, set in motion by everyone from mortgage lenders to the nation’s biggest retail banks, there’s a real role for instant-issue policies in today’s life insurance industry. However, meeting this growing demand implies a need to reassess the mortality implications created by these products.
For the life insurance industry to be able to offer instant issue, it must be willing to give up underwriting requirements that slow the process down. The name “simplified issue” refers to the concept of issuing a policy quickly due to less underwriting. The question that needs answering is, does the industry have any insured lives data that describes the mortality associated with simplified issue practices?
An Underwriter’s Perspective
The time between sale and policy delivery generally has been identified as time spent in the “underwriting department.” Although it may take a considerable amount of time to get the policy into the insured’s hands, even in the old days the actual time the underwriter spent on a case may have involved as little as 30 seconds to assess the risk and approve the application. Yet mail time, data entry, processing bottlenecks, policy printing, and more mail time made the time in underwriting seem longer.
Years ago, small face amount policies that required no additional hurdles beyond an application and Medical Information Bureau (MIB) check were often referred to as nonmedical. These policies were easy to process and fairly quick to issue. Some companies actually separated these applications for fast-track handling to speed up policy delivery to the agent’s hands.
Now, as the industry embraces instant issue and streamlines the underwriting process, it may actually be streamlining areas other than the underwriting function. This is good news and has several favorable ramifications.
First, quick issue or simplified issue has a historical foundation for both pricing and mortality assumptions; the underwriting function, past and present, may be nearly identical from a risk assessment standpoint. Although the methods used to gather and input the data might have changed, the basic information and sources have not.
Second, from that foundation, further enhancements are possible: for example, implementing some of the latest technologies to gather even more underwriting evidence without hindering time to issue.
Finally, although the time in underwriting isn’t truly the issue, even the 30 seconds to underwrite can be reduced to milliseconds using computerized electronic efficiencies.
Nonmedical, Paramedical, Medical
To draw upon industry experience, the Society of Actuaries (SOA) routinely tracks and reports on individually underwritten life insurance mortality through its basic mortality tables and annual ordinary experience study updates. The mortality studies stratify risk by underwriting type: nonmedical, paramedical, and medically underwritten.
Underwriting involves the assessment of insurability and one component of the process includes reviewing an individual’s medical health history. By definition, a nonmedical application requires no exam, although medical history information is still gathered, usually as a section included in the application. A paramedical exam is conducted by a health care professional other than a physician, generally a nurse or nurse practitioner. A medical application asks individuals to undergo examination by a licensed physician.
Simplified issue may mirror nonmedical life insurance in many ways, and in reality, there’s a considerable blurring of the lines between nonmedical business and simplified-issue business. Simplified issue doesn’t describe a unique process; rather, it merely relates to quick turnaround due to point-of-sale information gathering. Numerous carriers currently use their regular application along with an MIB check to actually define their company’s simplified-issue business.
To simplify the underwriting process, every requirement is fair game for streamlining. The definition of simplified issue is in the eye of the beholder. When is an application deemed complete versus simplified? Some suggest a finite number of questions—six, 10, or whatever—to differentiate full from simple. Yet by combining questions, the number may decrease, but the content stays the same.
Carefully drafted applications make sure to include "catch all" questions, open-ended questions that cover the holes in the application process and help identify any other medical conditions not yet covered. These questions are intended to complete an application, while gathering relevant data. This allows both the proposed insured and the insurance company to home in on the relevant history in a timely manner. Therefore, the true difference between a full nonmedical application and a simplified application, from an underwriting perspective, may be smaller than originally anticipated.
Although one might think the SOA's nonmedical results would have some bearing on mortality for simplified issue, these actuarial study differentiators date back to a time before the introduction of broadly used requirement testing protocols that have had a profound influence on mortality results. The primary change relates to the widely employed use of blood and urine specimens.
The underwriting community also uses many other tools to stratify risk, and all have value in discerning mortality. It's not unusual for a company to employ 10 or 12 such items. Some examples include the following: application, MIB, paramedical exam, medical exam, fasting blood draw, urinalysis, resting EKG, treadmill EKG, timed vital capacity test, chest x-ray inspection report, telephone insurance increases.
Even though cost-benefit studies and competition tend to justify more underwriting for larger face accounts, perhaps another reason for procuring additional testing has to do with a desire to reduce volatility associated with large-account claim results. Often, underwriting has been scrutinized postmortem by management that asks how the large face-amount claim could have been avoided, and sometimes the answer is that additional underwriting requirements would have uncovered the problem during underwriting.
The actuarial committees that produced the aforementioned mortality studies understood the pitfalls associated with the no longer relevant nonmedical, paramedical, and medical risk class differentiators. They recognized the need to further elucidate the value of different underwriting packages. Yet with so many different potential combinations of underwriting requirements, it quickly became clear the committees needed a different potential combinations of underwriting requirements, it quickly became clear the committees needed a different technique to overcome this obstable.
As more underwriting requirements were collected and the amount of life insurance applied for increased, it became apparent a surrogate for any specific underwriting package could be derived from banding mortality by amount of insurance. The 1990 to 1995 Basic Select and Ultimate mortality table did that by creating three bands, all referenced to the aggregate results. Band 1 studied policies up to $99,999 of life insurance; Band 2 included $100,000-to-$249,999 face amounts; and Band 3 studied policies with face amounts of $250,000 or higher.
These pieces, taken together, show that the life insurance industry has studied simplified-issue mortality and has a foundation for crafting some starting-point mortality assumptions. But there are still some additional pitfalls in interpreting industry data.
SOA Mortality Data
The 1990-1995 SOA Select and Ultimate Mortality Table broke experience out by face-amount bands. By reviewing under-writing requirements necessary by age and face amount for life insurance applied for from the early 1990's, one can see that most companies were issuing life insurance on a nonmedical basis for face amounts up to $250,000 or $350,000 for ages 0 through 40 or 45. Starting at ages 18 and up, blood and urine specimens were obtained at $100,000 by most companies. At around age 46 to 50, for face amounts of $50,000 or greater, many companies were collecting paramedical exams that included urine and sometimes blood samples.
Viewing mortality by band works as a surrogate for different underwriting packages so long as the underwriting dynamics are kept in mind. Because of the blurring of lines between a simplified-issue application and a nonmedical application, business that was issued using an application and MIB alone may mimic some examples of simplified issue.
With this background, let's look at male, Band 1, Duration 1 mortality as a percent of the Aggregate Duration 1 mortality rate by age. Band 1e is largely composed of policies undergoing minimal underwriting, similar to simplified issue. Duration 1 is used because this describes policies underwritten and issued in the 1990's. Notice how mortality ramps up at age 47 where policies under $100,000 are 254 percent of Aggregate Duration 1. This number by itself if striking and worthy of note as it reflects how divergent the mortality is. Also notice, from age 47 and forward, that the mortality rates between the two converge.
Understanding that there are more underwriting requirements obtained at the older ages coincides with the point in time where the differentials created by underwriting packages between Band 1 and aggregate start to disappear. Notice how the peak appears to occur in the upper 40s by age, which is consistent with the timing when the underwriting community is procuring additional requirements for policies under $100,000. These requirements tend to be paramedical exams with urinalysis and sometimes blood testing, thus no longer simplified issue.
It's important to note that other mortality study differences between Band 1 and aggregate reflect compositional variations between smokers and nonsmokers, different companies contributing different volumes of exposure to different bands, and target market and underwriting philosophy differentials by insurance companies contributing to the study.
Graph 2 describes Duration 1 male age nearest birthday mortality rates by age for Bands 1, 3, aggregate, and ultimate. Because bands are surrogates for underwritten packages, it's worthwhile to compare how less underwriting Duration 1 mortality might compare with fully underwritten first-year mortality rates. It's hard to say exactly how much of the mortality differential is due to underwriting versus other factors, but it would not be incorrect to assume the lion's share pertains to the underwriting tools used to find mortality, which is our y-axis dependent variable.
As would be expected, the lowest mortality rate is awarded to policies whose face amount is $250,000 or more; the aggregate, which is all policies issued no matter what face amount it applied for, it next. As expected, Band 1 mortality is closest to ultimate mortality, and the mortality rates all increase with increasing age. Also, there's noticeable spread on an absolute basis commensurate with the dynamics of more disease/impairments and greater and/or lesser ability to detect these diseases, depending on the underwriting package.
Amount of Underwriting
Does degree of underwriting matter? Table 1 reveals the scalar reductions in mortality associated with a Duration 1 risk compared with a similarly aged insured whose mortality is that of the ultimate table after underwriting has worn off. Actuarially, this is best described as q[x]/q[x-t]+t where t=26 and q[x-t]+t describes ultimate mortality and q[x] describes select period Duration 1 mortality. This table makes intuitive sense because it describes mortality rates by face amount for under $100,000, over $250,000, and all face amounts combined (the aggregate table). Comparing Duration 1 with ultimate is used to reflect current underwriting practices employed in the 1990-to-1995 era compared with mortality associated with individuals whose effect of underwriting occured at least 25 years before 1990 and has now worn off.
For smaller face amounts where only a medical declaration occurs in conjunction with the application, the majority of underwriting finds diseases that have already been diagnosed and reported. For larger face amounts, where blood, urine testing, resting EKGs, and other medical studies occur, additional undiagnosed diseases may be uncovered. These tools also have the ability to help predict future problems—for example, high cholesterol manifesting itself in coronary artery disease mortality.
Viewed another way, Table 2 shows how much mortality would have to increase were there no underwriting and assuming no antiselection. The reciprocal of the mortality discount in Year 1 describes the increase in mortality associated with no underwriting.
To further elucidate the mortality differentials, Table 3 looks at selected mortality rates for the Duration 1 low-band (<$100,000) face-amount policies compared with high-band ($250,000+) duration 1 mortality rates. This is done by the ages previously listed to produce the table.
The striking feature of interest is to see how much higher Duration 1 mortality is for low-band business compared with high band. There's a continuous upward grading of mortality rates, which would be consistent with less underwriting identifying fewer impairments, as the number of impairments increases with age.
This data is helpful in describing how much higher true nonmedical mortality is compared with fully underwritten by age. It could be argued the $250,000 and up band includes moderately to fully underwritten policies. Fully underwritten, in terms of underwriting requirements obtained, only truly occurs at very large face amounts of perhaps $5 million or higher. For policies that large, more underwriting should reduce mortality even more. If we truly compared nonmedical mortality with fully underwritten mortality, the ratios described here would be even greater still.
Price has gone down significantly since the implementation of additional underwriting in the late 1980s. Major drivers of cheaper life insurance have been the impact of additional testing associated with the underwriting process, plus the advent of preferred risk underwriting. It therefore follows that simplified underwriting—which doesn't include the source of these mortality reductions—can't assume fully underwritten mortality discounts.
As the industry moves forward there will be concerted effort to discover new underwriting tools to replace today's tools and simplify the process. It will be important to remember that these tools will drive mortality down from a nonmedical starting point. And, as shown here, that point is very different from fully underwritten mortality assumptions.
This is not to suggest there isn't a place in the industry for simplified-issue life insurance. In fact, underwriters everywhere would welcome simplified-issue risk classification because a win-win scenario amoung agent, applicant, and underwriter involves the agent and applicant wanting the life insurance at the applied-for risk class, and the underwriter gladly accepting and placing the risk on the insurance company books. As an industry, it behooves us to do this at the right mortality assumption to create a win-win for all.
Douglas A. Ingle is vice president, mortality and underwriting research, for ING Re’s individual life and health operation in Denver. The views expressed in this article are solely those of the author and do not necessarily represent the views of Security Life of Denver Insurance Co. or its affiliated companies.
Contingencies (ISSN 1048-9851) is published by the American Academy of Actuaries, 1100 17th St. NW, 7th floor, Washington, DC 20036. The basic annual subscription rate is included in Academy dues. The nonmember rate is $24. Periodicals postage paid at Washington, DC, and at additional mailing offices. BPA circulation audited.
This article may not be reproduced in whole or in part without written permission of the publisher. Opinions expressed in signed articles are those of the author and do not necessarily reflect official policy of the American Academy of Actuaries.
Inside Track: Anecdotal Evidence