What's Backing Your Life Insurance Guarantee?
What's Backing Your Life Insurance Guarantee?
by Carl Friedrich
Over the last 25 years, there has been a resurgence in the development of universal life products that prominently feature the strength of underlying guarantees with a special focus on products with secondary guarantees.
Life insurance historically has been characterized in terms of promises made by insurance companies to compensate policy beneficiaries under certain contingent events. Different forms of guarantees have been used in life insurance, and various means of analyzing and establishing financial support to assure these promises can be met. Although the prevalence of additional, non-guaranteed benefits has increased in the last 25 years, notably with the development of universal life products, there has been a resurgence of products lately that prominently feature the strength of underlying guarantees. Below, a variety of examples are presented with perspectives regarding the type of actuarial analysis necessary to assess the costs and risks related to those guarantees, and commentary regarding methods to secure financial support. The implications with respect to the requirements of various accounting systems and the views of different audiences, including regulators, reinsurers and the capital markets will also be covered.
Term Life Insurance
In contrast to the majority of ordinary life insurance coverage issued in the last 20 years, including par whole life, universal life and variable life plans, term insurance products written since the early 1990s have generally featured fully guaranteed net costs over their primary coverage period, with guaranteed level premiums over a 10, 20 or 30 year period. Typically, coverage could be continued beyond the level premium period, but this involved annually increasing premiums (YRT) that were loaded substantially for anti-selection. Much of this business was developed concurrently with the emergence of enhanced underwriting techniques and multiple risk classifications, which led to much lower premium rates for many applicants.
These plans generated GAAP reserves that built to substantial levels, reflective of a level premium structure that contrasts with underlying benefits that increase substantially with age. Statutory reserves held by many companies in the 1990s were based on the unitary reserving method, which relied on premiums that exceeded tabular expected claims in the ultimate YRT period. This methodology proved to be flawed in many cases, as significant lapses at the end of the level premium period could dramatically reduce the future expected margins, leaving deficiencies relative to the funding of guarantees that were provided during the level premium period.
Regulation XXX was implemented for various types of life business issued beginning January 1, 2000. This addressed flaws in unitary reserving methods as applied to some products. It imposed a humpback reserve requirement in basic reserve calculations over the guaranteed level premium period. It continued to require standardized mortality assumptions, based on the 1980 CSO Mortality Table, to be used with this methodology that had (in some cases) huge margins relative to expected mortality for certain classes of insureds. However, it allowed for individual company determination of "X factors" that were used to set mortality assumptions that were part of deficiency reserve calculations, providing some relief to companies.
Some further relief has been provided as the 2001 CSO table was adopted by the NAIC in December of 2002. However, even this updated table fails to recognize the benefits of modern underwriting techniques and distinctions in underwriting classifications beyond simple smoker versus non-smoker mortality differentials.
The resulting capital strain associated with term life insurance has left direct writers with the challenge of seeking capital support to back their underlying product guarantees. Reinsurance has provided a major outlet, with the majority of this business ceded through internal or external arrangements. This has led to increasing concerns for direct writers about credit risk exposure to external reinsurers, especially as consolidation in the reinsurance market is occurring. In addition, this has highlighted the importance of terms and conditions in reinsurance treaties.
Reinsurers' generally favorable views on mortality in the last 10 years have fueled some of the support being provided to direct writers. However, the major factor behind these reinsurance mechanisms is the use of offshore entities that are not subject to the same level of conservatism in accounting and reserves that are required in the United States. These advantages can apply to either external third party reinsurers or to reinsurers affiliated with the direct writer. However, for direct writers to reflect reinsurance reserve credits on their financial statements for reserves ceded to unauthorized reinsurers, including offshore reinsurers, one of three solutions must be utilized: letters of credit (LOC's) secured from a qualifying bank under prescribed conditions, assets held in trust or funds withheld reinsurance.
LOC's have been the most common mechanism in use. There are risks, however, as the majority of LOC's are written with a one-year term. Demand has dramatically increased, and capacity limits and costs into the future are in question. It has been projected that life industry demand for LOC's to support XXX-related and AXXX-related requirements will total $100 billion over the next five to seven years. Reinsurers typically have guaranteed their rates to direct writers in order to parallel base plan guarantees. This shifts the LOC capacity and pricing risks to reinsurers. However, the life reinsurance market has been hardening in the last year, and more direct writers are utilizing internal, affiliate-related reinsurance solutions, despite the LOC risk exposures.
Another capital solution has been emerging, notably with respect to XXX-related reserves. Rather than securing support from life reinsurers, some direct life writers are turning to the capital markets by developing securitization arrangements of the future cash-flows from specified blocks of business. These represent structured solutions that isolate a book of business, potentially secure a financial guarantee from a mono-line financial guarantee insurer to "wrap" the business, and secure financing through capital markets to support the reserve requirements of the business. The first major term life securitization occurred in early 2003, which triggered significant activity in the structuring of future potential deals on direct term life business.
The type of actuarial analyses necessary to understand the risks associated with this business has become increasingly sophisticated. Deterministic pricing, using static assumptions, has been complemented by stochastic modeling which examines the potential range of results generated under a set of assumptions regarding statistical distributions of key variables.
Corporate risk management, which requires a sound understanding of the range of potential results relative to certain exposures, is increasingly critical. The ability to affect external capital solutions typically demands considerable sensitivity analyses at a minimum, and usually stochastic analysis. Key variables normally examined include interest rates, mortality assumptions, lapses and other factors that depend on the coverage under consideration.
Universal Life Insurance Primary Guarantees
Basic guarantees provided in universal life (UL) plans include credited interest rate floors, cost of insurance charge maximums, static surrender charge schedules and expense charge maximums. Much in-force UL business was written in a higher interest rate environment than today's, with 4 percent or higher guarantees, and was not viewed as a corporate exposure when interest rates were in the high single digits or low double digits. One offset to the interest rate exposure was that multiple parameters could be adjusted by the life company, if necessary, to achieve target profitability. Asset hedging offers opportunities to reduce risks associated with a low interest rate environment, but few companies have pursued such programs because of the costs involved. In some situations, this has resulted in GAAP reserve adequacy problems for certain blocks of business as interest rates have dropped in recent years.
Increasingly, companies are conducting stochastic analysis to complement deterministic pricing of universal life and formulaic GAAP and statutory reserving standards. This analysis typically focuses on the interest rate risk inherent with the primary guarantees in the plan, but may include other key variables.
Another form of special guarantee emerging in the last decade involves "Guaranteed Maturity Extension Provisions." In order to avoid payment of an endowment benefit that would generate taxes for the policy owner at the stated maturity date of the contract (often the insured's age 95 or 100), these provisions continue death benefit coverage for an extended number of years, as long as coverage was still in place at the original stated maturity age. These extensions have often been provided free of explicit charge, with no additional charges assessed from the extension point forward.
The pricing of these provisions is rather simple if the insurer priced the product assuming a mortality rate of 100 percent at the maturity age. In that case, the maturity extension likely represents no additional cost to the company, since the assumption going in already overstates the true cost. Where other pricing assumptions had been previously utilized, adjustments will be needed to assess mortality rates into the extension period.
With the introduction and gradual implementation of the 2001 CSO mortality tables, more plans will naturally extend their maturity dates to age 120. It will be of interest to see whether this will eliminate the need for this feature, or make it easier for companies to rationalize offering maturity extensions beyond age 120.
Universal Life Secondary Death Benefit Guarantees
Also common is including additional short-term guarantees on universal life contracts that assure that if certain conditions are met (usually the payment of a specified premium level on a cumulative basis), coverage continues even if cash surrender values might not be positive. In recent years, these "secondary guarantees" have become more generous, to the point of becoming lifetime guarantees in many policy forms.
The lower interest rate environment has heightened attention on the guarantees inherent in universal life contracts. Some guarantees were directly impacted by the reserving requirements of Regulation XXX. For other designs, the impact was less obvious. This was one factor leading to a new generation of secondary guarantee structures, notably featuring shadow account designs. Shadow account calculations are independent from the regular UL accumulation value. Shadow accounts are, in fact, never used to determine cash surrender values–their only purpose is to determine whether secondary death benefit guarantees are in effect. Shadow accounts may use unique credited rates, cost of insurance factors and loads within this calculation. With most of these products, if the shadow account is positive, the UL secondary guarantee requirement is considered to have been met and the UL policy will not lapse regardless of an insufficient cash surrender value. Through the design of the parameters used in the shadow account calculation, the duration of the secondary guarantee can be controlled. This was one of the primary reasons for the development of shadow account designs, as companies sought to develop mechanisms that addressed consumer needs and allowed a range of guarantees to be delivered over a continuum of premium scenarios.
UL market share from 1999 to 2003 increased in large part as due to these attractive secondary guarantees. A new actuarial guideline (AG 38, or AXXX) was adopted effective January 1, 2003 (or for in situations earlier) that attempted to clarify the reserve requirements of Regulation XXX for many of these designs. AXXX details a complex nine-step calculation for shadow account designs, with the intent that reserves should approach Net Single Premium Levels as premium requirements for lifetime guarantees are met. In many cases this resulted in AXXX reserves that significantly exceeded cash values and CRVM reserves. In other cases, product designs were constructed that, using the formulaic approach specified in AXXX, led to AXXX reserves that were no higher than CRVM reserves, despite meaningful secondary guarantees. It should be noted that projected account values based on current interest rates for competitive policies typically show the secondary guarantee being triggered for attained ages in the early 90s, in essence providing free insurance from that point forward. In some cases, it became apparent to companies that the formulaic reserve calculations were insufficient to cover the true economic risks, and additional reserves were contemplated with "economic reserves" being established, either due to economic considerations, regulatory risk factors or both. This has led to questions regarding the derivation of tax reserves that may be treated as deductions in company tax calculations, as the current Internal Revenue Code requires the use of formulaic, seriatim calculations for deductible reserves.
A limited number of reinsurers supported direct writers on the guarantees of this business, gradually shrinking to one company during the last two years, and none currently. The typical mechanism had been to cede 80 to 90 percent of the net amount at risk to the reinsurer, with virtually all reinsurance charges waived in the event the secondary guarantee was triggered. This provided a rationale for the reinsurer to hold the incremental AXXX reserve (over CRVM).
With the fading of third party reinsurance as a solution to back these guarantees, affiliated reinsurers were established by a number of direct writers. While this provides capital relief, it exposes the direct writer to the LOC pricing risks discussed above and to the economic risks of triggering the secondary guarantee.
Stochastic modeling is even more critical for UL secondary guarantee risks than for term. Key variables include interest rates, funding patterns, mortality and lapses. There are a series of critical assumptions in performing such analysis, including mean reversion interest rate assumptions. Companies are spending much more time building models to understand exposures in this area, and as covered below, discussions are occurring as to whether a principles-based valuation approach should supplant the current formulaic standards.
With respect to securitization of UL secondary guarantees, the ability to demonstrate that these risks are minimal to an investor or a monoline financial guarantor is much more difficult than it is for term guarantees. I am aware of no AXXX securitization transactions consummated with respect to UL with secondary guarantees.
Concern has been expressed by some regulators regarding the potential for the spirit of AXXX to be circumvented under certain product designs. Discussions are continuing in the NAIC Life Health Actuarial Task Force (LHATF) regarding a potential rewrite of AXXX. Three different proposals were discussed in the September 2004 NAIC meeting, including a proposal that would redefine the calculation steps to be utilized, particularly for shadow account designs. A modified version of this proposal was adopted for exposure in the October 2004 LHATF meeting, with the potential for adoption in the December 2004 meeting.
As evidenced by a second proposal presented in September, there appears to be growing support from many in the industry for a principles-based, stochastically supported reserving methodology in lieu of formula-based approaches. An American Academy of Actuaries (AAA) Working Group has been established to address products with complex guarantees and/or tail risks. The scope of their efforts includes VUL and UL with secondary guarantees, and has been expanded to include term life products.
Short term, if changes are made to AXXX, the issue of retroactivity versus prospective application may become critical to many companies. Longer term, the potential change to a principles-based set of standards will be a challenging assignment requiring consideration of tax reserve issues, the level playing field issue across product lines, allowable assumptions and standards and an array of other issues.
Variable Universal Life Secondary Guarantees
The most recent life insurance product line in which guarantees have emerged is variable universal life insurance (VUL). These feature less competitive death benefit guarantees than under UL, but the gap appears to be narrowing. Reserving requirements for these guarantees were spelled out in Actuarial Guideline 37, which generally reflects an attained age level reserve methodology, along with a one-year term reserve. There is a unique provision in this guideline that allows companies to provide documentation in the state of domicile as to why the AG37 standard may be redundant.
There are a number of different VUL product design variations, including some featuring a UL-like guarantee embedded in the contract, with certain requirements regarding levels of funding into the fixed account. Other policy requirements, including minimum premium and fund diversification requirements, may allow for a more competitive guaranteed premium, but these are still likely to be higher than UL guaranteed premiums. I expect that additional creative design variations will emerge in the future in this product line.
The state of life reserving requirements for life products with long term guarantees appears be at a critical juncture, with future product evolution dependent on LHATF actions in the short term, and the AAA Working Group, LHATF, and other interested parties in the longer term. Support for the capital requirements of the business will be provided from an increasing variety of sources, including traditional reinsurers and capital markets. The analysis of life guarantees and the range of potential outcomes will become increasingly sophisticated out of necessity in order to address the requirements of emerging reserving rules and of the capital markets. We believe that the net result of these factors will be a better understanding of the risks being taken in the life industry. In some cases, guarantees may be curtailed relative to current levels as a result of this analysis. Nonetheless, I believe that strong guarantees have become an expectation in many life insurance markets, and companies will continue to explore and develop a variety of products with attractive guarantees that address the needs of their customers.
Carl Friedrich is a consulting actuary at Milliman, Inc.