Exploring the complexities of emerging product risks with actuarial
consultant Martin Snow
Through the years, life insurance and annuity products have become more complex. For example, customers can benefit from the upside of the stock market without participating in the downside by purchasing certain types of life insurance and annuity products, and customers can obtain very inexpensive guaranteed premium permanent death-benefit-only life insurance. These features place increased risk on the insurer—both for market risk and insurance risk. In addition, customers can now buy policies that combine life and health benefits, and companies are starting to rely more heavily on artificial intelligence (AI) and machine learning in the underwriting process. All of these changes bring increased risk that a good pricing actuary needs to know how to consider.
Risks associated with valuing the guarantees provided, including the development of new experience studies and models, marketing issues related to agent and policyholder understanding of new policy provisions and their interactions, and the risks associated with administering complex policy provisions and providing accurate customer service to policyholders, are all concerns that companies must address with respect to these products.
To examine the topic of product risk more fully, The Actuary asked Martin Snow, FSA, MAAA, several questions about product risks.
The Actuary: What emerging risks do you feel are most significant for product actuaries?
Snow: The list of emerging product risks is fairly long. One important issue is product and distribution relevance, which refers to the degree to which potential customers understand what our products can do for them and can relate to our methods of distribution. We must find ways to balance this relevance to potential customers with the risks we are accepting.
The potential impact of big data to disrupt our industry is significant. Will we choose to find ways to improve our company processes and products using new data and methods, or will we adopt a “wait and see” approach? Are the roles of the actuary, data scientist and statistician clearly delineated in our companies? Do our companies have a comprehensive strategy about how to use and efficiently integrate big data?
Speed to market has long been a concern of companies. Have we identified the primary impediments to getting products out there more quickly? What are we doing to improve our company’s speed to market?
Theoretically, nonguaranteed elements (NGEs) in a contract provide a company with greater flexibility and less risk. However, do customers understand the implications of nonguaranteed elements? Do our companies review NGEs as frequently as our contracts or procedures say we will? What risks are our companies exposed to by our current NGE practices?
For many years now, there has been talk of the inevitability of the return of high interest rates. To date this has not happened in a significant way. Companies may wish to explore whether our product portfolios are well-positioned for a sustained low interest rate environment. Additionally, companies that priced products over the last decade or longer assuming that interest rates will rise should analyze what the impact on the embedded value or other profitability measure is likely to be. Similar thoughts would apply to companies that assumed significant mortality improvement at ages where it may not be emerging as anticipated.
For companies that are making a big push into the longevity business, have we adequately assessed the potential impact that longevity emerging differently than anticipated may have on our portfolios?
Other product risks on the list include whether certain risks are insurable, the potential impact of principle-based reserves (PBR) on pricing and reinsurance, the emerging changes in the life risk-based capital factors and calculations, the ability of our models to accurately measure the emerging risks, the appropriate balance between expense management and having the resources necessary to do the pricing work, if the models and assumptions have been validated, and the credibility of individual actuaries and the profession. Do we have the influence we need in our companies to properly measure and price the risks our companies are accepting? Will regulators gain greater confidence in our professionalism to move to simpler, yet viable and sound, regulation?
In the Chief Actuary PBR Survey I recently completed, one chief actuary stated: “I believe the actuarial profession has a tremendous opportunity to demonstrate its competence and integrity with PBR. I hope we collectively live up to it and further distinguish ourselves as trustworthy and highly capable professionals.”
The Actuary: What factors raise red flags for you about the possibility that a product has been underpriced?
Snow: Certainly, a dramatic and unexplained improvement in competitive positioning should be examined closely. If the company’s competitive position is one or two, it should be examined closely as well.
Independent third parties can provide useful feedback on pricing. If reinsurers, finance providers or competitors have exited a market, then their evaluation of product risks may be significantly different than the risks shown in company pricing.
Sensitivity testing is a key tool for evaluating the risk profile of a product. If sensitivity testing results cannot be readily explained, or small changes in assumptions result in big swings in profitability, then careful examination of the assumptions and models is warranted. Similarly, there should be concern if we are unable to produce a change flow showing the impact of each major assumption change on the new price.
Producers are experts at finding the situations where a product is particularly competitive (or underpriced). We should be skeptical when the assumed sales distribution cannot be readily supported, especially if cells with higher profitability have a higher assumed concentration of business. Another statement that should be viewed with caution is when we are told that an unprofitable feature is highly unlikely to result in any sales.
We also need to be on the lookout if the company is lending too much credence to anticipated improvements in experience (without strong evidence).
The Actuary: What advice would you give to a company that discovers that pricing assumptions are not being realized and a product has been underpriced?
Snow: It depends on the relative volume of sales. If it’s a major product, it certainly needs to be repriced expeditiously for new issues going forward. For policies already sold, it depends on how much was sold and what the impact on profitability is. The company also needs to quickly identify where there was a failure in controls (if this is the result of an error) and tighten them before the next pricing, or why the assumption was approved yet changed so quickly. The company might wish to review its assumption-setting procedures and put tighter assumption governance rules in place. For example, the company could require that:
- Assumptions be well-justified, documented based on emerging experience and sensitivity tested.
- The impact of assumption changes be quantified.
- The assumptions be reviewed more frequently.
- The assumption oversight be at a higher level in the organization or otherwise strengthened.
The Actuary: How could the risk of underpricing be reflected in a company’s risk management program?
Snow: Any major product change (e.g., new product, introduction of new policy feature) and assumption change should be reviewed from a risk management perspective as part of the company’s formal risk management processes. Significant changes in the external environment (economic and regulatory) should also be reviewed regularly to see what effect they may have on product risk.
It is important for the company to have a clear delineation between the actuary and risk manager roles, and for the actuary and risk manager to work closely together where their respective roles overlap.
In addition, companies need to be comfortable that their actuarial software and administrative platforms are equipped to effectively and—in a scalable and controlled manner—appropriately handle the complex features of emerging products.
The Actuary: What have been the main drivers of additional complexity?
Snow: Some of the chief drivers of additional product complexity have been the desire of companies to gain a competitive edge, the availability of financial instruments to support product designs and the regulatory environment.
The Actuary: How has product complexity affected pricing methodologies?
Snow: At some companies, riskier products are required to have higher returns. There is always the question as to whether this should be established at the corporate level or at the line of business (LOB) level, and whether this is at the marginal level or at the fully allocated level.
It is also important to ensure that the pricing models can adequately handle products as they become more and more complex. One example where this might be particularly relevant is in products with both death benefits and living benefits, such as chronic illness benefits. Is your pricing software able to effectively model a product that has both of these benefits?
The Actuary: What has been the impact of more complex products on sales illustrations and other initial client communications?
Snow: Sales illustrations are extremely difficult to read. Many policyholders may not fully understand them (perhaps an understatement). Industry, regulators and professional groups need to work together to ensure that the right information is presented to the customer in a way that the customer understands and absorbs. With complex products, this can be a very difficult task.
The Actuary: Do you see any reversal in trend to products becoming simpler in certain markets?
Snow: Companies would certainly like for this to happen. Whether it will happen and how quickly is another question. Perhaps getting potential customers to return to our industry through the use of AI will enable the development and sale of simpler yet comprehensive products.
Subject Matter Expert
Martin Snow, FSA, MAAA, consults for insurers, software developers and startup companies, and he is a member of the Advisory Board of Atidot, an agile machine learning InsurTech company.
He is a seasoned insurance industry executive with key roles at MetLife, TIAA and Prudential. He has extensive pricing, valuation, reinsurance and regulatory experience for life, annuity and long-term care products. In addition, Snow is an expert in principle-based reserves (PBR). He led the development of the industry’s first universal life policy with secondary guarantees (ULSG) priced with PBR and the conversion of the pricing models to a new platform.
Snow is a member of the ASOP 11 Task Force of the Actuarial Standards Board and is a frequent speaker at Society of Actuaries (SOA) meetings.
He is also a member of the Big Data Task Force of the American Academy of Actuaries (the Academy). In this capacity, he is leading the preparation of the portion of the Academy’s Big Data Monograph related to Current and Emerging Practices for Life Insurance.
The Actuary: If a company has an economic incentive not to have policyholders exercise product options efficiently, how can the company appropriately balance the interests of the company
and the policyholder?
Snow: Some companies are using AI to help develop buyback programs. These programs should factor in both what it will take to get the customer to agree to the buy-back as well as ensuring that it is truly based on a fair value. Companies will need to exercise care to ensure that consumers are compensated adequately and are not taking undue risk by agreeing to the buy-back.
Some companies are educating consumers about what is in their policies. From customer service and fairness perspectives, this is clearly the right thing to do. These companies need to be careful to ensure their pricing anticipates this approach early on so there are not big profitability and risk surprises.