Research Paper

Critical Illness Capital Requirements

A recommendation by the CIA Committee on Risk Management and Capital Requirements

June 2011 Document 211060 Ce document est disponible en français © 2011 Canadian Institute of Actuaries

Research papers do not necessarily represent the views of the Canadian Institute of Actuaries. Members should be familiar with research papers. Research papers do not constitute Standards of Practice and therefore are not binding. Research papers may or may not be in compliance with Standards of Practice. Responsibility for the manner of application of Standards of Practice in specific circumstances remains that of the members in the life insurance practice area.

Memorandum To:

The Office of the Superintendent of Financial Institutions (OSFI)

CC:

Autorité des marchés financiers (AMF), Practice Council, CIA Members, Canadian Life and Health Insurance Association (CLHIA)

From:

Wally Bridel, Chair Committee on Risk Management and Capital Requirements

Date:

June 7, 2011

Subject:

Research Paper – A Recommendation on Critical Illness Capital Requirements

OSFI asked the CIA Committee on Risk Management and Capital Requirements (CRMCR) to review the current practices with respect to critical illness (CI) capital requirements being employed within the minimum continuing capital and surplus requirements (MCCSR) calculations by Canadian companies, and recommend a single approach, because: •

The MCCSR guidelines were not clear as to how CI should be handled; and

•

There was an apparent diversity in practice, leading to a wide range of results.

The CRMCR undertook this review, assessing the appropriateness of applying the existing life, disability income, and other capital approaches to CI. As a result of this review, the CRMCR recommended: 1. To use the same mathematical approach as is used for individual life volatility; and 2. That a catastrophe component is not required for CI. Based on these recommendations, the CIA conducted a quantitative impact survey (QIS) to determine the impact of the proposed change. There was a wide range of results, but it is noted that the change represents an increase in CI capital for most companies when compared to current practice. However, the impact from a total MCCSR/Test of Adequacy of Assets in Canada and Margin Requirements (TAAM) is generally negligible. There was some concern as to the impact this change may have on the CI market, as the change could potentially have a material impact on product pricing, internal rates of return, etc., for some companies, particularly smaller companies. Ultimately, the impact will depend on how individual companies choose to incorporate any change within the pricing and product development process. One argument for the change is that having one consistent approach helps to level the playing field. Pursuing this further was beyond the scope of the committee. In accordance with the Institute’s Policy on Due Process for the Adoption of Guidance Material Other than Standards of Practice, this research paper has been prepared by the CRMCR and has received approval for distribution from the Practice Council on April 14, 2011. If you have any questions or comments, please contact Wally Bridel, Chair, Committee on Risk Management and Capital Requirements, at [email protected]. WB

360 Albert Street, Suite 1740, Ottawa ON K1R 7X7  613.236.8196  613.233.4552 [email protected] / [email protected] actuaries.ca / actuaires.ca

Research Paper

June 2011

1. SUMMARY OF CURRENT SITUATION The determination of capital to support critical illness (CI) products is currently mixed. Some insurance companies follow the approach used for disability income (DI), some life insurance, and the majority other accident and sickness. This inconsistent treatment leads to materially different levels of capital being held. In 2008, the CIA’s Committee on Risk Management and Capital Requirements (CRMCR) undertook a quantitative impact study (QIS) with the industry to assess two alternative approaches: (1) volatility and catastrophe as per life insurance; and (2) DI approach. Results are attached (appendix B), and show how the capital derived from each of the alternative approaches compares to the current reported capital. 2. VOLATILITY Individual life volatility follows a binomial approach, as follows: S = 2.5 x A x B x E/F, where • • •

A is the standard deviation for the upcoming year’s projected death claims; B is the natural log of the Macauley duration of the projected death claims; and E/F is the ratio of net amount at risk to total face amount.

CI and individual life products have similar characteristics. Both products are based on an incidence rate times a policy exposure, generating current benefits. There is no runoff claim exposure for these products as there would be under DI, for example. Choices for the determination of capital include one of the two proposed survey approaches or a new approach outside the proposals in the survey. It seems reasonable to apply the same approach for determining capital for CI that is used for life insurance (i.e., the binomial approach). This approach can readily be applied for CI. It should be acknowledged, however, that this is likely to lead to relatively higher capital requirements for CI than for life, due to the smaller size of CI blocks for most companies compared to life insurance. The focus of the CRMCR’s discussion was on standalone CI, which is the most common form of CI product sold in Canada. Some companies have accelerated CI products (the CI benefit is an accelerated payout of the life insurance face amount). The CRMCR’s view is that the volatility approach would be applicable to both standalone CI and accelerated CI products. Recommendation: Adopt the life approach for CI volatility. 3. APPROXIMATIONS TO VOLATILITY It would be reasonable to provide for an approximation approach to be used in the instances that the detailed information is not available, or that materiality of the block warrants a simpler approach. Recommendation: The various approximation approaches that are available to individual life and group can also be used for CI. It is also possible for some group CI to require approximation approaches.

3

Research Paper

June 2011

4. CATASTROPHE COMPONENT Upon review, it is unclear whether a catastrophe component is required. Furthermore, if a catastrophe component is required, it is not apparent that the 10 percent of expected claims approach that is used for life is appropriate for CI. In the following section, we present comparative and quantitative information related to the catastrophe component. i) Single Discrete Event If a catastrophe is defined as a single discrete event that has an identified start date and fairly defined end date, then it is reasonable to conclude there is no catastrophe exposure for standalone CI. CI claims tend not to emerge immediately but instead take years to emerge in experience (e.g., cancers following a nuclear catastrophe). Immediate claims following a catastrophe will generally be focused on death, dismemberment, and burns; these claims are typically not covered under CI benefits. Catastrophe events generally have a geographical focus. The number of inforce CI policies is still relatively low such that there would not be geographic concentration of risk at this stage. Therefore, although some potential for large events may exist for CI, it is not similar to the magnitude of risk associated with life insurance. For accelerated CI, a catastrophe risk is still present for the life insurance component of the product. Thus, the committee believes that holding a catastrophe component for accelerated CI is consistent with current practice. ii) Single Discrete Event Leading to Long-term Change in Expected Claims Is a single discrete event that leads to a long-term change in expected claims considered a “catastrophe” event? Such a long-term change in expected claims must be capitalized in reserves at the valuation date. Questions arise as to whether this risk is covered in provisions for adverse deviations (PfADs) and/or capital ratio? The experience with prostate cancer, and a change in definition, was to see a short-term increase in incidence followed by a return to the original trend line. The view was that the increase accelerated the recognition of prostate cancer, but did not increase the overall total number of cases. Another example of a single discrete event that could lead to a long-term change in expected claims is a court ruling. Court rulings, however, would typically only affect one of the covered conditions and are usually taken in a very narrow context. The court decision may only affect one company or only a subset of the company’s policies. It is expected that the impact on overall total claims would be minor following a court ruling. Recommendation: For standalone CI, there is no requirement for capital related to catastrophe. For an accelerated CI product, it is appropriate to hold catastrophe capital. 5. DIVERSIFICATION WITH INDIVIDUAL LIFE AND GROUP MORTALITY There is a tie to accelerated product, and the 30-day survival period in CI. If volatility captures the randomness around CI incidence, there may not be significant diversification with mortality. Current MCCSR does not consider diversification across risks, which would support not diversifying. Furthermore, life and group blocks may dwarf CI add-on capital, reducing it to a very low incremental amount. Recommendation: Capital requirements for CI are not diversified with life and group mortality. 4

Research Paper

June 2011

6. PRICING/MARKET IMPACTS Late in 2008, the CIA undertook a QIS based on year-end 2007 results to gauge the impact on capital requirements of both the life volatility and DI approaches. A copy of the QIS is attached in appendix A, and the results are summarized in appendix B. Results are shown both gross of reinsurance and net of reinsurance arrangements. The first two columns show the results using 200 percent of the DI requirement versus the amount actually reported for CI in the 2007 MCCSR/TAAM return for each company. Columns (3) and (4) show results under the proposed life volatility approach versus reported in the MCCSR/TAAM return. The average result is 365 percent of the amount currently held by federally-regulated companies and 199 percent by Québec-based companies. The weighted averages are 107 percent and 160 percent for federally-regulated and Québec-based companies, respectively. The proposed change has a higher impact on smaller companies, which is expected under the volatility approach. When total MCCSR/TAAM ratios are examined (columns (5) and (6)), the survey results show that either approach has a minimal overall impact, with only slight reductions in ratios versus what was filed in the 2007 returns. Outside the average, there is a wide range of results (particularly under the DI approach, which is less influenced by the size of the CI block). This variability in results illustrates the diversity of current practices and approaches to determining CI capital requirements. The survey results indicate that the proposed change could have an impact on the CI market. Companies would likely review their product pricing since potential premium increases may be required to maintain internal rates of return and other pricing targets. This may be of particular concern for smaller companies. Ultimately, the impact of requiring companies to follow a common approach will depend on how individual companies choose to incorporate the change. The potential costs of a common approach for determination of capital would have to be weighed against the potential advantages of having one consistent approach which could help to level the playing field. Pursuing this further was beyond the scope of the committee.

5

Research Paper

June 2011 Appendix A

CRITICAL ILLNESS SURVEY COVER LETTER To: Canadian Life Insurance – Appointed Actuaries Re: Capital Requirements for Individual Critical Illness products In order to ensure consistent application of capital requirements for Critical Illness (CI) products, OSFI is considering redefining the regulatory capital requirements for Critical Illness products. The CIA Risk and Capital Committee has put forth a proposal such that the capital requirements for CI products should be calculated as either: 1. The capital requirement using the individual life formula for mortality risk, or, 2. The requirements for disability income insurance new claims risk multiplied by a factor of 2. For 1 above, the morbidity parameters (face amount, net amount at risk, duration, etc.) should replace the mortality factors. For 2 above, the factor should be applied to the premium related to the CI risk. The portion of the premium related to any return of premium of other ancillary benefits should be excluded. In order to evaluate the impact of this proposal on the industry we are asking all Canadian Life Insurers with products containing individual CI risk to complete the attached template. Results for the industry will be compiled by OSFI/AMF. Please send the completed template back by September 12, 2008: For federally-regulated and non-Québec provincial insurers, to: [email protected] For Québec-regulated insurers, to: [email protected]

6

Research Paper

June 2011

Company Name Morbidity Risk Capital Individual Critical Illness Products 2007 Year End (’000) Approach: Using 2007 Year-End Values

Gross (100% of MCCSR)

Net of Reinsurance

Volatility

Volatility

Overall MCCSR/TAAM Ratio (%)

Reported at 2007 Year End 200% of DI New Claims Risk Approach* Catastrophe

Catastrophe

Individual Mortality Formula Approach

* Please apply the adjustment factor of 75% for benefit periods not exceeding two years. Above values should reflect the capital for morbidity risk only. Should you have more than one block of CI policies, please report the combined results. Please provide additional information that may be relevant to this exercise.

7

Research Paper

June 2011 Appendix B

Critical Illness Survey Results – November 5, 2008 Federal Companies Total reported current CI capital required: $52.6 million

(1) Gross – required 200% of DI/Gross – required Company reported (%) A B C D E F G H I J K L M N O P

(1) 159.8 111.3 150.7 128.5 500.4 150.7 151.2 508.1 150.0 206.8 200.0 198.8 374.9 500.0 374.7 375.0 Average 265 Weighted average 214

(2) Net – required 200% of DI/Net – required reported (%)

(3) Gross – required Mortality (vol)/Gross – required reported (%)

(4) Net – required Mortality (vol)/Net – required reported (%)

(5) MCCSR ratio with 200% of DI/MCCSR ratio reported (%)

(6) Approx. MCCSR ratio with Mortality/ MCCSR ratio reported (%)

160.2 111.3 150.7 126.4 501.1 150.7 150.6 503.8 150.0 205.1 200.0 197.7 375.1 500.0 374.7 375.1

661.6 38.8 409.9 47.0 758.9 72.5 81.1 Na 51.1 149.0 Na Na 418.8 37.7 110.4 122.3

691.8 * 38.8 * 409.9 48.7 * 846.7 72.5 * 79.9 * 980.8 * 51.1 * 147.3 * 1,400.0 264.3 297.7 269.6 * 109.4 * 136.6

100.0 100.0 99.6 99.9 99.0 99.5 98.8 99.4 98.9 99.9 99.9 99.3 99.0 99.5 96.1 98.6

99.7 100.0 99.2 100.2 98.3 100.0 100.4 98.7 100.4 99.9 98.1 98.7 99.2 100.1 99.4 99.8

265

185

365

99

~100

226

121

107

99

~100

(2)

* CI not aggregated with mortality. # CI aggregated with other business.

8

Research Paper

June 2011

Québec Companies Total reported current CI capital required: $5.6 million for eight companies Average 271 Weighted average 275

241

267

199

100

99

256

222

160

100

100

9