Reassessing the value of insurance: the current economic climate has commercial insureds looking at risk financing in a new light. (Property/Casualty: Commercial).

With the confluence of a hard market, pressure on corporate profits and a pervasive feeling of greater uncertainty, many eyes are focusing on the value of insurance. Suddenly, insurance appears to be more and less worthwhile at the same time. These are ideal conditions for rushing to judgment, which means it is precisely the right time to take a step back and reassess the value of insurance before making any decisions.

First, let's begin with a review of current thinking and then outline a new framework for understanding the value of insurance that clarifies when insurance definitely makes sense, when it does not make sense and when it might make sense--in effect, a triage.

Current Thinking

It is human nature to ignore the real value of something until the unexpected happens. While a lackadaisical attitude may be OK in a static environment, a more thorough understanding is a source of competitive advantage in a dynamic environment. Rather than attempt to understand the economic value of insurance, most of the current thinking about insurance focuses on its relative value using some or all of the following approaches.

Lottery. One mind set is that the value of insurance is obvious. Everyone knows someone who has won or lost the insurance lottery--collecting hundreds, maybe thousands, more than he or she had paid in premiums or the opposite: suffering a big loss that could have been insured for peanuts. Referencing these examples, there is no question that insurance is valuable and that "you have to play to win."

Renewal. Another mind set is characterized by a renewal mentality, for which the value of insurance is assessed by reference to itself. Premium, coverage, limits and security have gone up or down so the value of insurance is thought to be higher or lower, and one buys more or less insurance than before.

Retention. This approach uses rules of thumb to determine the optimal retention level by considering how much risk managers think the corporation can retain, based on the impact that losses would have on various performance measures. This way of thinking, combined with knowledge of premium credits for higher retention levels, usually supports the status quo or results in marginal adjustments to retention levels.

Benchmarking. The herd mentality is alive and well, and conformity matters. Occasionally, there are those who value being different. Whatever the decision, it is comforting to know what others, similarly situated, are doing.

Quantification. Some large companies are able to determine the probability of annual aggregate losses within retention levels for which they have a high frequency of claims and the claim-generating processes remain stable. Others assume they can estimate the probability of large losses by adjusting population data to reflect their individual characteristics. In this case, the value of insurance is assumed to be the difference between the premium and the expected loss or is related to the probability that losses will exceed the premium.

Efficient frontier. Another approach involves plotting expected cost and variability of the cost of alternative retention and insurance structures, as measured by the difference between expected cost and, say, the 95% confidence level cost. The efficient frontier is defined by a set structure for which it is not possible to lower the expected cost at the same level of variability, or lower variability at the same expected cost, or lower both expected cost and variability. Using this approach, insurance structures on the efficient frontier are preferable to those that are not.

Tax and accounting. Insurance premiums are tax deductible and count as an expense for financial-reporting purposes, whereas self-insured losses, generally speaking, are not deductible until paid and can be expensed only when reasonably estimable. For these reasons, structures for retaining risk that offer tax and accounting benefits--such as a captive insurance company, retrospectively rated insurance or finite insurance--may be of particular value.

Service. The value placed on insurance services goes both ways. Some policyholders value the greater experience and efficiency of an insurance company when it comes to services such as loss control and claims management. Others don't need help and may see insurers as interfering with or potentially acting against their interests.

Requirement. In some cases, insurance is required by regulation or mandated by contract. For example, states require that employers maintain workers' compensation insurance or qualify as a self-insurer. Similarly, contractors often are required to maintain insurance. Here, the value of insurance is defined externally.

Signal. Insurance helps build internal and external stakeholders' confidence by assuring them that funds are available if something goes wrong. This may have great value. On the other hand, insurance may signal the presence of a soft and rich target attractive to plaintiffs, encouraging run-of-the-mill lawsuits and creating a settlement floor on larger suits.

With the exception of analogizing insurance to a lottery, which has a certain madcap appeal, all of these ways of thinking have merit. None of these approaches, however, answers the more fundamental questions of whether and by how much insurance adds to or subtracts from the bottom line. These questions demand answers, especially in today's market.

While the cost of insurance is reasonably clear--some would say painfully clear--the cost of managing insurable risk and running a business without insurance is not as apparent. It is not obvious at what point risk, beyond its manifestation in losses, has a cost in terms of diverting scarce financial and operational resources from projects with upside potential. A practical way to triage the value of insurance is needed so energy devoted to making decisions about insurance is focused on what matters and not dissipated on things that only appear important in the fog of renewal negotiations. Seeing through this fog is critically important for buyers, sellers and intermediaries alike.

To understand the value of insurance, one must compare the cost of insurance premiums with how much it would cost to hold capital to fund the same types of losses. This cost of literal self-insurance equals the cost of expected losses plus the "cost of risk capital"--that is, the cost of the capital multiplied by the amount by which actual losses may exceed expected losses. Self-insurance is rarely taken this literally in practice, however, and this approach is used solely to understand where insurance coverage is worthwhile.

The chart "Insurance vs. Literal Self-Insurance," below, depicts the relationship between insurance and literal self-insurance over the relevant range of per-occurrence limits. Assuming it performs as expected, the cost of insurance is known and is represented by the solid blue line on the chart. The cost of literal self-insurance only can be estimated, since expected losses and the cost of risk capital are both uncertain, Therefore, it is useful to put upper and lower boundaries around the cost of literal self-insurance.

In this example, the value of insurance is shown in green and equals the amount by which the cost of self-insurance is greater than the cost of insurance. Here, insurance definitely provides value over a wide range of per-occurrence limits, because it costs less than even the lower bound of the self-insurance cost. Above and below the green area are yellow areas, where insurance may or may not be of value, and some judgment is required. The bottom and top of the chart are depicted in red and indicate where insurance definitely does not provide value. An analysis of this triage is warranted.

Triage Analysis

There are two red areas in the graphic where the cost of insurance is definitely greater than the cost of self-insurance. One occurs at the lower per-occurrence limits, where claims frequency is usually high and uncertainty--both as to expected loss and variability of actual loss from expected losses--is relatively low. Here, even if the buyer and seller agree on the same level of expected loss, the transaction costs of insurance render its purchase an unwise financial decision in comparison with self-insurance. Furthermore, insurers are typically reluctant to provide coverage in this area, due to concerns about potential moral hazard.

The red area at the top of the chart begins where claim frequency is very low or nonexistent, even for the population at large. At this point, buying insurance isn't a good financial decision for the buyer, because the organization's ability to create value would be totally vitiated in the event of a loss of this magnitude or because a loss of this size is viewed to be impossible. In either case, literal self-insurance would have no cost, therefore insurance, even if priced to reflect capacity commitment only, has no value.

A triage pattern with a large green area is typical of insurable risks that meet the classic conditions for insurability, for which risks are fortuitous, independent, homogenous and quantifiable at the collective level. In this area, insurance is definitely a good value. The insurance "sweet spot" starts where the insurance cost line intersects the lower boundary of the self-insurance cost envelope and ends where the marginal cost of insurance (that is, premium per million of limit) begins to exceed the marginal cost of self-insurance. For classically insurable risks, the "sweet spot" remains pretty much intact even if the buyer has a more favorable view of expected loss than that implied by the seller's pricing. Nevertheless, the insurance "sweet spot" is reduced and may vanish altogether if the price of insurance goes up dramatically, particularly if it goes up in conjunction with coverage restrictions or rationing of limits.

Caution, Discretion Is Advised

The graphic depicts two yellow areas or "caution zones" where insurance might or might not provide value. Making good decisions in these areas calls for expertise, discretion and judgment.

More advanced loss modeling is a partial answer. While most companies do not have enough data to estimate the probability of large losses based on their own experience, it is possible to gain some insight as to the probability of large losses by tailoring population loss data to a particular company or by building cause-and-effect models of particular exposures. In this way, the difference between the buyer's view of expected loss and the expected loss implied by the premium may make it clear whether insurance has value or not.

Companies can reduce the size of the lower caution zone by developing a database of historical losses both for insurable and noninsurable events, such as computer system outages, contract disputes and unanticipated maintenance costs. A review of this type of event database will reveal the size threshold below which an insurable loss is just one of many losses suffered each year and above which an insurable loss either stands alone or is one of only a few other losses. Below this threshold, the insurable risk in question is largely diversified within a portfolio of nonhomogenous risks and little or no capital should be attributed to it. Above this threshold, capital should be attributed on a nearly dollar-for-dollar basis. Again, this may make it clear whether insurance has value.

Developing a better understanding of the size and other characteristics of potential insurable events by exploring various scenarios can help companies reduce the boundaries of the upper caution zone. Using this technique, senior management may determine that the occurrence of a particular insurable event is either impossible or highly remote. Similarly, they might conclude that a particular event would irreparably damage the organization's ability to create value or that it would make the company's future so uncertain that a decision to commit additional funds to the business should be made only after an event occurs. Understanding the potential ramifications for various types of extreme losses provides greater clarity about whether further attribution of capital makes sense and whether insurance provides value.

A final way to reduce the size of the caution zones is to improve the company's understanding of its cost of capital. Rather than estimating the cost of capital as a single number, effort can be made to explore in more detail the cost of various sources of risk capital across the full range of potential loss experience. These sources would include cash flow from operations, funds held in liquid assets, lines of credit, debt, equity and various types of derivatives.

Reassessing the value of insurance is useful for buyers and sellers of insurance, because it requires both parties to take a step back and consider the role of insurance as a means of financing risk. It is in both parties' interest to understand when and where insurance adds value and to take active measures to expand its relevance over the full range of potential losses.

In the final analysis, exercising good judgment is paramount.

[GRAPH OMITTED]

L. Ware Preston and Bruce B. Thomas are senior vice presidents in Marsh's Enterprise Risk Consulting Practice, Stamford, Conn.