Reprinted with permission of ON THE RISK, Journal of the Academy of Life Underwriting.
Authors: Sylvain Goulet, FCIA, FSA, MAAA, Principal, Eckler Ltd. and Carmela Tedesco, Vice President, LOGiQ3 Corp.
The protective value that MIB’s Checking Service brings to the North American insurance industry is widely acknowledged and plays a vital role in ensuring its member companies can offer – and consumers can purchase – affordable life and health insurance protection. Despite this widespread belief, MIB considers it important to independently quantify the value of its services for its members. MIB does this by periodically conducting protective value studies using independent actuaries and consulting underwriters. This article summarizes MIB’s most recent study of a voluntary group product sold through the workplace and underwritten using a simplified process.
MIB has published studies demonstrating its protective value in both the individual fully underwritten life insurance market and the simplified issue life insurance market. In the latter half of 2013, MIB engaged underwriting consultants LOGiQ3 and the actuarial firm Eckler Ltd. to determine the protective value that MIB’s Checking Service brings to individually underwritten group voluntary insurance products sold through the worksite. The study examined a representative sample of employee and dependent applications for insurance, with average face amounts applied for of $163K, where evidence of insurability (EOI) is required, and standard underwriting evidence obtained included a short form application and MIB. The study credited MIB with protective value only in cases where it was the sole alert that triggered an investigation resulting in a decision other than issue, and it also employed a fully burdened cost model.
Similar to previous studies, this study validates the strong protective value of MIB and its role in ensuring overall product profitability. As expected, MIB’s protective value differs when compared to past studies due to the nature of the group voluntary insurance model. Group voluntary insurance is typically written on an ‘accept/decline’ basis, with no opportunity to collect extra premium commensurate with additional risk, whereas carriers selling individual insurance permit underwriting outcomes ranging from table ratings to flat extras. Group carriers account for this difference by setting the risk tolerance for ‘accept’ at levels ranging from 150% to 200% of standard mortality. Another significant difference is coverage duration; the ultimate age for individual coverage extends until age 80 and beyond, while group insurance coverage in this study ended at age 67, with conversions excluded, or an average age of 75 including conversion. Lastly, lapse rate assumptions are much higher for group insurance because they are driven not only by the consumers’ buying habits, but by employment duration and carrier changes made by the employer. All of these factors combine to paint a uniquely positive picture of MIB’s protective value in a very distinct market and distribution model.
During the design phase of this study, the study team decided to develop the protective value information from a candidate record set of all applications submitted to the subject companies over a period of twelve calendar months to provide a typical snapshot of application activity. In total, a sample of 20,000 applications were selected to represent individually underwritten group voluntary insurance from the larger pool of ‘all applications’ for the subject companies.
Protective value exists when the service, in this case the MIB Checking Service, provides an alert to the group insurance carrier that prompts further investigation and leads to an underwriting action other than an ‘approval’, and the estimated present value mortality savings of using the service exceeds the cost. Year-to-year savings were calculated in today’s dollars by discounting future excess mortality at 4% given today’s economic environment, and taking into account reasonable anticipated lapse rates of the policies over time. Lapse and mortality assumptions were provided by the subject companies and were used in calculating the present value of savings.
Through analysis of the data provided by the carriers on the 20,000 cases, the study was able to focus the review on cases where there was demonstrable MIB ‘exclusivity’. In the course of this study, ‘exclusivity’ was deemed to exist when the MIB Checking Service prompted an alert to the group carrier advising them of previously unknown information (medical or non-medical) pertaining to an applicant’s insurability. In general, the alert allows the group medical underwriter to re-question the applicant on the information received from MIB when the applicant did not declare the information on the EOI (evidence of insurability) at application time. When ‘exclusivity’ was established, the study then expressed the protective value of MIB for each particular case in terms of the additional mortality impact associated with the risk.
In practice, once the applicant responds to re-questioning, the group medical underwriter can determine if the information is satisfactory for risk selection and whether a decision can be made, or if additional information is required from a physician for risk selection. This process allows the group medical underwriter the opportunity to properly assess the applicant’s insurability. The selection of risk in group medical underwriting is on an ‘accept’ or ‘decline’ basis. Most group carriers will consider ‘accept’ risks from standard up to table 2 (150% of standard mortality), and sometimes as high as table 4 (200%). This study used table 3 (175%) as the basis for ‘accept’.
In the course of the study, when an applicant did not respond to re-questioning by the underwriter it was determined that these situations warranted inclusion in the study due to the potential mortality impact of the impairments that were previously reported on the individual. In these cases, had MIB not been utilized in the risk selection process, the applications would have been classified as ‘accept’. These misclassified cases were found to be impactful to the carriers’ block of business, and this is where the protective value becomes apparent and relevant to group voluntary life insurance.
When the study determined that MIB played a role in the underwriting process of a case, the next step was to determine the mortality impact. In the event that applicants did not reply to re-questioning, a conservative approach was taken to apply assumptions about the probable mortality impact based on the applicant’s current age and details of the MIB code. Underwriting and MIB guidelines advise that mortality cannot be determined by the MIB code(s) alone; however, for the purposes of this study, and to establish a probability of mortality, the information provided by the MIB codes was used.
Certain conditions that are chronic and progressive in nature, such as heart disease and diabetes, portend to an outcome that is easier to conclude based on the impairment code, whereas challenges arise in assessing the outcome of impairments that are acute in nature, and represent what might be termed a ‘decreasing’ risk. An example of an acute or ‘decreasing’ risk is the impairment risk posed by history of alcohol, drug or driving issues; it is entirely possible that the condition could have been resolved and, depending on the time frame since resolution, could have been assessed as an ‘accept’. However, the question remains, if it was resolved, why did the applicant not reply to the re-questioning to confirm that there was resolution? These impairments can range from standard to decline, depending on severity, time since diagnosis/treatment, and relapses, but for the purposes of this study it was decided to conservatively classify these risks at 200%, or just outside of ‘accept’.
All other increasing or decreasing mortality risks were applied the appropriate probable outcome, in the absence of current information, based on the following criteria: Current age of the applicant and the MIB impairment, including the modifying degree and time letters.
The infograph below illustrates the findings of the review. 115 cases were determined to be ‘exclusive’. However, after review, five cases were no longer eligible for ‘exclusivity’ because the information developed through the MIB codes was apparent to the group carrier through previous applications.
Actuarial Methodology Employed
In determining the protective value of MIB’s Checking Service, one must realize that there is no single definitive answer; results will vary by company and plan. Actuarial assumptions would take into account the circumstances and experience of each particular company and plan, and actuarial assumptions are key in the determination of the final results. Consequently, the methodology that the study employed was (a) to set a series of reasonable actuarial assumptions, (b) to determine the Net Protective Value Index (‘NPVI’) on that basis, and (c) to perform sensitivity testing by varying the assumptions from the study’s reasonable basis in order to assess their impact, and to present a range of possible outcomes.
The study defined the NPVI as: NPVI = (PVFEMS - MIBCost) / MIBCost, where:
- PVFEMS = the present value of all future expected mortality savings as a result of using MIB’s Checking Service for one year.
- MIBCost = the fully burdened cost of using MIB for one year.
In other words, for every $1 of MIB cost, the NPVI would be $X. A value of $X greater than $1 results in a better than break-even position. Because varying assumptions may have a significant impact on the NPVI, the margin must be over $1 in order to definitely state that MIB’s Checking Service is indeed a worthwhile investment.
Fitting the Data
The first step in the study’s actuarial work was to ensure that the sample selected was a good representation of the entire group. The length of this article does not permit a full documentation of these results; however, the following graphs illustrate two of the various criteria used in the study to ensure that both the sample data and the selected declined cases (or cases not placed) were an excellent representation of the entire group.
Other criteria included average age, average face amount, face amount distribution, percentage of males versus females, and percentage of employee versus spouse coverage for the sample group and the ‘not placed’ group.
Key assumptions used in the actuarial calculations are:
- A normal retirement age of 67 on the basis of US Social Security retirement age. However, it is reasonable to assume that a portion of those impaired lives will convert when they have the opportunity. The study has assumed a varying percentage of those applicants converting to a longer duration: 40%, 30%, 20%, and 10% to age 70, 80, 90, and 100, respectively. For sensitivity testing, the study has replaced these weights by an equivalent attained age of 75, which is not significantly higher than 67.
- As mentioned earlier, group carriers account for a risk tolerance to ‘accept’ at levels ranging from 150% to 200% of standard mortality. The study has assumed that any ‘not placed’ cases would be at the 200% level unless LOGiQ3 had additional information resulting in a higher mortality rating.
- Lapse rates: Used an average lapse rate of 10% annually.
- Pricing mortality: Used the carriers’ average pricing mortality. The study assumed a pure loss ratio of 60%, an expense and commission ratio of 25%, and an overall profit margin of 15%.
- Valuation mortality: Used 100% of the 2013 SOA Group Term Life (individually billed) Mortality Table as our base mortality to quantify the protective value.
- Used a discount rate of 4% in light of the current economic conditions.
- Considered the employee group only.
- Because the fully burdened MIB cost varies for group carriers depending on their size and volume, the study determined an assumption based on a medium-size company. Together with that, the study has also assumed a reasonable salary base for the underwriting staff to use the MIB Checking Service, including fringe benefits at 40% of salary.
On the above basis, the NPVI is 15.20, meaning that for every $1 of MIB cost, the Net Protective Value is $15.20. Even if the study assumes an ultimate age of 67, with no conversions of any kind beyond age 67, the NPVI is 9.83, still a significant protective value index in our opinion.
The study performed a number of sensitivity tests to demonstrate a possible range of results. The study results are summarized in the following table and two graphs. The table summarizes sensitivity tests corresponding to each point on the X-axis, so that the NPVI is in ascending order for each column. The first graph includes all sensitivity tests to demonstrate and compare the impact of the most important assumptions, such as (a) the lapse rate, (b) the mortality of cases not placed, (c) the discount rate, and (d) the ultimate attained age. The second graph focuses in on the narrower results.
*The values included in this table, and their impact on protective value, inform the x-axis of the following two graphs.
- The Net Protective Value Index has been calculated at 15.20 based on a reasonable set of assumptions. In other words, for every $1 of MIB cost, the Net Protective Value is estimated at $15.20. The underlying return on investment on MIB’s Checking Service is consequently quite large and this level clearly demonstrates the value of using MIB’s Checking Service.
- The NPVI is particularly sensitive to (a) the lapse rate, (b) the mortality of cases not placed, (c) the discount rate, and (d) the ultimate attained age.
- In reality, underlying assumptions would vary based on the circumstances and experience of each particular company and plan. However, it is clear that, under all sets of reasonable assumptions tested, the NPVI is significantly above 1.00. To use actuarial terminology, the study would say our ‘best-estimate’ of the NPVI is 15.20, to which one can add (a reduction) a margin for adverse deviation.
Published in the March, 2000 edition of the On The Risk journal.
Published in the September, 2010 edition of the On The Risk journal.
The study employed a ‘fully burdened’ cost model which included both direct and indirect costs incurred by the subject companies in their use of the MIB Checking Service.
About the Authors:
Sylvain Goulet, FCIA, FSA, MAAA, is a Principal at Eckler Ltd. and is responsible for actuarial consulting services at Eckler's Toronto office, focusing on life and health consulting in Canada, the US, the Caribbean and internationally. Sylvain has 35 years of actuarial experience in all actuarial and insurance fields in these markets and acts as Appointed Actuary for a number of companies. His expertise spans a number of disciplines from product design and pricing, valuation, M&As, actuarial and illustration systems, strategic opportunities and non-traditional actuarial areas.
Carmela Tedesco is Vice-President at LOGiQ3. She has managed various underwriting consulting and outsourcing projects at LOGiQ3 which include operations, auditing, training, risk management, project management, underwriting strategy, claims and workflow process review and implementation. Carmela has over 25 years of underwriting experience in life, DI, CI, structured settlements, health and group underwriting which spans both the reinsurance and direct insurance level for the North American, European, and Caribbean markets. More recently, Carmela has successfully brought underwriting training to a new level by offering an innovative web-based, video training option to life underwriters and insurance professionals around the world.