How much risk should an enterprise carry? Most insurers would say: "As much as possible to maximize returns without jeopardizing the financial solvency of the enterprise." But this approach fails to recognize the individual risk tolerance of investors, bondholders, regulators, and insurance company decision-makers, each of whom have their own perspective of risk. Nor does it take into account the sophistication called for by increasingly volatile markets and ever-demanding regulatory standards. Defining risk and determining an appropriate risk profile has become a complicated process, requiring far more robust and dynamic tools. This is why economic capital analysis holds so much promise for insurers.
Determining economic capital is perhaps the most fundamental risk-management activity that an enterprise undertakes, for it seeks to quantify the amount of capital that an enterprise needs as a financial cushion against potential losses. What is an adequate level of capital given the risks that the enterprise has or may assume? Is its capital allocated effectively among its products? Or are low-performing products eating up too much capital, while top performers go hunting for resources? How these questions are answered affects not only the financial solvency of the enterprise but also its ability to compete. Maintaining insufficient capital jeopardizes the viability of the enterprise. Retaining too much capital or allocating it ineffectively hampers the enterprise's ability to compete.
From maximizing return for equity investors to minimizing risk for policyholders and regulators, insurers are often pulled in different directions in an effort to define an acceptable level of risk for the enterprise. Economic capital gives managers a realistic approach with which to evaluate and integrate the risks the organization faces.
A leap forward in risk assessment
One of the primary advantages that economic capital models have over conventional tools is their capacity to quantify the risk-reward tradeoff of an insurer's strategic choices. Based on stochastic or probabilistic analysis, economic capital models provide a distribution of loss outcomes for different risk scenarios. In analyzing these loss scenarios, managers can compare the risks associated with a product, for example, and quantify its capital needs.
This ability is a huge advance over conventional tools that typically provide a limited range of loss outcomes with no information about the probabilities that those losses will occur. With conventional tools, each loss outcome within the range has an equal chance of occurring.
Traditional deterministic (or formulaic) methods may have been adequate at one time, but as profit margins narrow and markets become increasingly skittish, insurers have sought out more sophisticated risk-assessment tools that could measure the impact of risk scenarios on their strategic choices. How would their products react if interest rates spiked or if the stock market tanked? Which product would suffer? How much? Which products would hold up?
Implied in these questions is the need to know whether the products in a portfolio have an offsetting effect under a given loss scenario. If an increase in interest rates bolsters the performance of an insurer's disability products, for example, but depresses annuity results, what are the offsetting benefits? Traditional measures lacked the capacity to quantify the diversification benefits of an insurer's product portfolio; this is a crucial blind spot, as correlating diversification benefits can lower (or in some cases increase) an insurer's capital needs.
Conventional tools, which rely heavily on industry ratios and averages, don't allow managers to look beyond general comparisons of capital adequacy.
Entering an information-rich age
Economic capital models overcome many of these shortcomings because of the rich information they provide about the probability of loss outcomes.
Economic capital models generate potentially thousands of loss scenarios, which are compiled to form loss distributions. However, only a narrow band of extreme scenarios—those worst-case outcomes in the tail of the distribution—are analyzed to determine their impact. This process allows an insurer to examine a product's loss probabilities and better understand the loss characteristic of a product. In this way, managers can hone their risk tolerance and risk appetite and then determine whether the capital allocated to a product line is worth the risk.
Moreover, loss-distribution scenarios are also aggregated across product lines, an advancement that allows managers to determine whether offsetting or diversification benefits will mitigate certain risks or if losses across lines will deepen under a certain loss scenario. Managers' view of risk is expanded to the potential linked or diversification effects among a company's products. Instead of trying to approximate capital allocations using industry averages, an insurer—now armed with a virtual universe of losses—can manage its capital based on its unique risk and product profile.
This ability shifts the management paradigm to a risk-adjusted platform. Decisions pertaining to reinsurance, investment hedging strategies, product portfolio, and entry into new products or markets can now be grounded in a quantified analysis of the potential trade-off between risk and reward of a decision rather than on vague notions of risk.
As good as it gets
A model is only as good as its underlying assumptions. For some exposures, estimates of loss have been fairly reliable. Over the decades, mortality data have shown that this exposure behaves in relatively predictable ways, allowing for a high level of confidence in the loss estimates. The same holds true for a number of other exposures, such as morbidity or voluntary surrenders.
But the reliability of projections related to strategic and operational risks has long worried insurers. Many catastrophic risks seem to take shape out of thin air, making for pie-in-the-sky projections. How do you estimate losses stemming from the malfunction of a hedging software program that goes undetected for months? How deep would losses be if a back-up computer system should fail to kick in when needed? Or if a pandemic emerged amid an already jittery stock market? Losses would be severe, but conventional models left managers wondering how severe. And if a model couldn't quantify risk, how could an insurer manage it?
Using the bottom-up approach made possible in economic capital models, the analysis of these high-impact, low-frequency operational and strategic risks has become grounded in reality. This new approach relies on the premises that, despite their volatile nature, many operational and strategic risks emerge over time. By tracking the chain of events that precipitate an eventual blowup, managers can gain an understanding of which levers trigger certain events and what costs are associated with each phase in the development of the risk. This process promotes the assignment of more realistic probabilities to once-elusive operational and strategic risks.
An equally important feature of economic capital models is their ability to integrate catastrophe loss estimates into the overall analysis. Unlike earlier economic models that merely tacked on a crude estimate of an enterprise's strategic and operational risks, economic capital models incorporate more realistic risk estimates into the model's overall framework, and thus remove many concerns about a model's risk-assessment capabilities.
Model uncertainty will always surround risk assessment. However, recent advances in how businesses think about their risk make it possible for managers to quantify risk with more confidence than before. With increasingly credible tools comes the ability to define and explain an enterprise's risk tolerances down the level of command and among its external stakeholders. And as awareness of the risk tolerances spreads throughout the organization, insurers can move ever closer to true enterprise risk management.
MARC SLUTZKY is a principal and consulting actuary with the New York office of Milliman. He consults to life insurance companies and investors on enterprise risk management, mergers and acquisitions, capital management strategies, and reinsurance. He has extensive experience in the industry as an actuary and financial officer and has worked with clients to develop economic capital models.
JAMES G. STOLTZFUS is a principal and consulting actuary with the Philadelphia office of Milliman. His professional experience includes actuarial appraisals; product development for life, health, and annuity plans; asset and liability analysis; cash-flow testing; statutory and GAAP valuations; embedded value analyses; and claim liability analyses. He has developed economic capital models for clients and assisted clients with their enterprise risk management.