Case Study: Efficiency and control through Milliman Mind
This briefing note provides a short case study on migrating an Excel/VBA model to Milliman Mind.
The recent volatility in the capital markets and the consequent adverse developments across the financial services industry have stimulated interest in more sophisticated risk-management techniques. Having spent our careers working in or for the life insurance industry, focusing on the pricing and valuation of life insurance and annuity products, we believe that much of the fundamental exposure facing life insurers arises from a failure to adequately understand the risks that are assumed. Such lack of understanding results in pricing products incorrectly.
Certainly, there are many types of risk, and each must be addressed to fulfill the requirements of a full enterprise-risk-management (ERM) solution. However, when focusing on the underlying drivers of primary risk, we believe that understanding which risks are being taken, charging a suitable price to offset them, and then managing the business to keep risks within the expected range—these factors together present the ultimate challenge.
Pricing risk relates directly to the long-term nature of a life insurance or annuity-type policy. By contrast, a candy bar manufacturer presumably knows or can ascertain all the costs associated with the development, manufacture, and distribution of the candy bar when setting its price. If any of the knowable costs change, that information can be used to adjust the price of the product. Unlike a candy bar manufacturer, "the cost of goods sold" inherent to a life insurance policy will probably not be known with certainty for many years and should be projected over a long period of time when setting the policy price. The failure to fully appreciate the risks being taken, not to mention their value, can lead to the underpricing of the policy itself and the company's ultimately incurring a loss on the sale. Given the sophistication of the insurance marketplace, the more underpriced a product is, the more policies a company may sell. While some elements of a life insurance policy may not be guaranteed and can be adjusted based upon events subsequent to the sale, the management of these nonguaranteed elements creates risk. A company's inability or unwillingness to manage such nonguaranteed elements can also lead to losses.
Market-viability risk occurs when a company cannot find a market for and sell its policies to a given constituency on a profitable basis. An insurance company will fail if it cannot find a continuing market for the products it sells. It also faces strategic risk from choosing the wrong markets in which to participate. These risks are closely related to and arise from pricing risk. There may be a market for a particular product at a given price that is not profitable, based upon the efficiencies and risk management competencies that the issuing company brings to the table. Likewise, a market for that particular product may evaporate if a company, when charging a price it believes to be profitable, cannot communicate the product's value to potential clients. Consequently, there is a natural friction in the management of each of these risks. Obviously, the ability to sell products that are profitable is the key to the long-term sustainability of a company.
Asset-related risks are associated with the products sold by a life insurance company. Given the long-term nature of a life insurance contract and the level premium charged for what is normally an increasing risk (i.e., mortality increasing with attained age), a life insurance company typically accumulates a substantial number of assets (premiums) to invest. The higher the rate it assumes it can earn on the assets, the lower the premium it may charge. The ability to meet the assumptions utilized with respect to investment return will determine whether the policy is as profitable as it was projected to be in the pricing process.
Successful risk management requires that the assumed investment rate (for guaranteed premium products) and the investment spread (for nonguaranteed products) are set consistent with the company’s realistic ability to achieve both objectives. Many of the historical failures of life insurance companies can be traced to the assumption of an unrealistically high investment rate/investment spread that resulted in investment in assets with excessive duration mismatch or dubious quality, or that became extremely illiquid. A "run on the bank" can jeopardize even the best-managed company. This risk can be reduced by avoiding excessive concentration in any one asset class and taking care to anticipate and plan for liquidity requirements under a range of different scenarios.
"Reaching for yield" to meet assumptions that were set during the pricing process has been a primary cause of life insurance company impairments over the last 25 years. Examples of this include investing in real estate to support interest rates credited on guaranteed-investment contracts, in junk bonds to support interest rates credited on universal-life and deferred-annuity contracts, or, more recently, in long-term assets supported by short-term borrowing to extract the difference in yield.
Those responsible for product development, pricing, and overseeing the overall risk-management function in a company generally recognize the impact of interest-rate movement and its potential negative effect on the projected profitability of certain types of products within certain markets. However, many of the assumptions with respect to policyholder behavior are still based on informed judgment rather than reliable experience. Consequently, the sensitivity of results under differing environments should be assessed and reflected in the product-development and pricing processes, because these form a key part of the risk-management process of the company.
Risks associated with other pricing assumptions must also be addressed in the risk-management process.
Lapse rates - Many of the most popular products being sold today (e.g., long-term care, level-premium term, no-lapse-guarantee universal life) are lapse supported (i.e., profits increase if the ultimate-duration lapse rate is increased). Lapse support has had a negative connotation among some, with an implication that the pricing of such products was somehow flawed. However, lapse support is a consequence of the product's design. Specifically, any coverage that charges level premiums for an increasing exposure (i.e., increasing probability of claim), without providing nonforfeiture values commensurate to the "equity" the policyholder has generated in the policy, will be lapse supported. Profits increase as the ultimate-lapse rate increases, because fewer policyholders are in force in policy durations for which the revenue collected is less than the benefits and expenses paid.
Conversely, if lapse rates in ultimate durations are less than were assumed in the pricing process, the profitability of the product will suffer. In some instances, the repercussions can be quite substantial, with reductions in the ultimate-lapse rates wiping out the projected profit of the product and creating a substantial loss. This has been particularly true with no-cash value life insurance (i.e., term to 100) sold in Canada several years ago and in the early-generation product offerings of longterm-care policies in the United States. Consumers, or their agents or brokers, tend to recognize a good deal when they see it, and do not lapse these policies, with lapse rates for some of them falling below 1% annually.
Assuming a low rate of ultimate lapse in the pricing process produces a higher premium, given a stated profit objective with all other assumptions remaining equal. Consequently, the natural friction between the salability and the profitability of the product emerges during the pricing process. The risk-management process should recognize this and avoid the utilization of unrealistically aggressive (i.e., high) ultimate-lapse rates in the pricing of these products.
Mortality assumptions - Insured mortality has improved significantly over the last 25 years. The extrapolation of this improvement into the future during the pricing process creates risk for the insurance company. It follows that if any projected mortality improvement does not emerge, the negative impact can be significant if ultimate-lapse rates also fall below pricing assumptions. The splitting of cohorts into more underwriting classes, the uncertain impact of medical technology, and the lack of credible mortality experience for older ages have resulted in the development of assumptions based on informed judgment rather than historical experience. This makes the monitoring of experience as it emerges critical to the risk-management process. Risk management may require the hedging or balancing of this risk among different lines of business.
Severe events - The effect of infrequent but severe events, such as epidemics or terrorist attacks, should be considered in both the risk-management process and the product-development/pricing process. The cost of stop-loss reinsurance can be incorporated into the pricing of life insurance products to reflect this risk.
New business - The absolute level of new business produced can have a significant impact on surplus levels. Tactics such as the financing of new agents and the payment of annualized first year commissions can produce a substantial effect on the level and quality of new business produced.
Counterparty risk - The potential negative consequences of counterparty risk have become clear over the past several months, as financial institutions with the highest ratings have failed or were acquired or bailed out in some form. Historically, the biggest counterparty risk within life insurance companies was associated with ceded reinsurance. The risk-management process should consider the reinsurer's ability to pay claims, even during times of economic or catastrophic distress. Risk management should also assure that pricing the guarantees embedded in variable annuity products with living-benefit guarantees recognizes the cost charged by well-capitalized, reliable counter parties.
Hedging cost volatility - Stock-market volatility directly affects the cost of hedging the risks embedded in variable-annuity products with living-benefit guarantees. As market volatility increases, the cost of hedges increases. Product pricing should reflect the long-term nature of these guarantees and the volatility of the cost of hedging them.
Risk contagion - The correlation of various risks has historically been underappreciated. The implosion of the subprime mortgage market led to the bursting of the housing bubble, which resulted in the tightening of credit standards, reduced consumer spending, increased stock-market volatility, decreased interest rates and increased interest spreads, asset devaluation and illiquidity, and recession. Causation can be debated, but the correlation of these events cannot. The magnitude of these developments occurring together is much more virulent than the effect of each occurring separately. The subtle interactions between risk factors often go unnoticed until emergence of a complex pattern that can be difficult to understand and anticipate.
Acquisitions - In an acquisition exercise, the premiums are set and the amount to be paid for the business is determined. The primary objective during an acquisition process is not for the company to understand the risks of the entity being purchased, but rather to understand how the risk profile of the new combined post-acquisition entity is different when compared with the preacquisition risk profile. The new entity will entail different risks and, presumably, different skills, so successful integration requires these to be optimally allocated and priced into the acquisition.
This article has stressed that many of the risks faced by life insurance companies can and should be reflected directly in the product-pricing and product-development processes. Nonetheless, not all risks can be successfully identified early in the process. The promises made by a life insurance company to its policyholders are by their very nature long-term. The effect of such things as agent misselling, interruption of business processing, and litigation can be managed and reflected indirectly in the product-pricing process. However, other risks—such as emergence of a new competitor, enactment of unfavorable legislation, development of new technology (e.g., the Internet), failure or bad behavior of a competitor (Madoff effect), war, economic depression, or an unsolicited offer to purchase the company—are more difficult to manage, avoid, and reflect in the pricing process. These risks can imperil the future of any one industry or company. It is therefore important for every company to make sure to understand the risks being taken and to try to reflect suitable allowances in pricing to ensure the longterm ability of the company to manage those risks for society.
Once understood, it becomes clear that it is difficult to differentiate solid everyday management of a company from an effective risk-management process. Risk management must be integrated into the everyday operation of the company—and indeed is the everyday operation of the company. Strategic success is possible when companies understand which risks are being taken and can appropriately charge for those particular risks. They can then focus on the important day-to-day task of managing those risks.
Bradley M . Smith is Milliman's chairman. His primary area of expertise is individual life, annuity, and health insurance, including the development of products as they relate to the overall strategic direction of a company. He has assisted numerous firms with their surplus management strategies, quantifying the impact of tactics such as financial reinsurance, merger, acquisition, demutualization, and divestiture.
Neil Cantle is a principal and consulting actuary with the London office of Milliman. He leads the development of Milliman's CRisALIS™ methodology for analyzing and quantifying enterprise risk. In addition to risk management, he has extensive knowledge of the U.K. retirement market, and experience with mergers and acquisitions, strategy, product development, and corporate restructuring.
Managing enterprise risk: For life insurance companies, pricing demands clear assessment of possible threats
The recent volatility in the capital markets and the consequent adverse developments across the financial services industry have stimulated interest in more sophisticated risk management techniques.