The Evolution of Finite Reinsurance and Financial Accounting Statement (FAS) 113

Finite Insurance Primer

For those of you who don't know what finite reinsurance is, here's a very brief primer. The hallmark of any finite risk deal is how its loss reserves are funded. Instead of paying a relatively small premium in exchange for a huge limit of liability (real insurance), in finite deals, the premium and the limit of liability were the same. Here's how it worked. Think of a risk, any risk—environmental impairment, for example. You determine that if the former landfill on which you built those 10,000 apartments starts leaking toxic substances, you really should have some sort of reserve fund from which to buy off potential claimants. Because you have more money than Bill Gates, you stick $20 million into a finite reinsurer.

You've been told that this "premium" might just be tax deductible, which sweetens the deal immensely. The offshore reinsurer sets up your own "experience account" into which the funds are deposited, provides you with a guaranteed rate of return on your funds, and tells you that at the end of the 3- or 5-year policy term, you can get it all back—whatever wasn't paid to the odd disgruntled tenant, that is. As they say back in the old country … such a deal! But it gets better.

You didn't even have to deposit the entire $20 million, just its present value based on a discount rate and the number of years in the policy period. This assured that, given the guaranteed interest rate and no losses, the fund would equal $20 million at the end of the policy term. The expenses associated with such arrangements varied, but in your view, it was well worth it. The upshot is this: real insurance operates on the principle of the "law of large numbers"—the premiums of the many pay the losses of the few. Finite reinsurance, however, relied on the time value of money, and not what we generally call the "insurance mechanism."

For about 5 years, many of us in the alternative risk financing game became intoxicated by the sheer possibilities of finite reinsurance. Centre Re's annual reports during the late 1980s and early 1990s were unlike that of any other insurer or reinsurer. They were brilliantly written and included ever more exotic case studies of actual clients, both primary insurers and noninsurance entities, solving seemingly intractable problems though finite reinsurance.

FAS 113 Arrives on the Scene

Eventually the Financial Accounting Standards Board (FASB) realized that we needed new guidance, beyond that provided in FAS 5, to determine whether a transaction could assume insurance accounting. Aimed squarely at finite reinsurance, FAS 113 was promulgated in 1992. While the statement also includes a cash-flow test, its most onerous requirements involve its definition of insurance as it pertains to the likelihood of a loss. It says that there must be a reasonable chance of a significant loss for insurance accounting to be used (which is the Holy Grail, by the way). Huh? What does this mean, exactly? Typical of those august organizations that write the rules, the FAS purposely did not elaborate on whether a given transaction did or did not qualify.

Meanwhile, back on the farm, few of us paid much attention to FAS 113, and mistakenly reasoned that it would take years for the FAS to catch up and actually enforce this thing. Moreover, few of us thought that what we were doing couldn't be justified as real insurance under the broad FAS 113 guidelines.

Fortunately, the American Institute of Certified Public Accountants (AICPA) came to our rescue in the form of what became known as the "10/10" rule. Contrary to popular opinion, FAS 113 does not say anything about the so-called 10/10 rule—it was a construct devised by accountants who determined that if a finite transaction exhibited a 10 percent chance of a 10 percent loss, it was defensible in a challenge. Mind you, this means that 10 percent of the risk must then be actually transferred to another risk-bearing entity. Over time, various accounting organizations developed their own opinions. One rule of thumb said that for the transaction to qualify for insurance accounting, the transferred portion of the risk should be equal to the (primary) loss fund.

The main problem we faced was not quantifying 10 percent of the possible loss—that was easy; in the above example, this figure was $2 million (10 percent of $20 million). The sticky wicket was coming up with how to determine the likelihood of such a loss. Actuaries are used to predicting future losses based on historical loss data. But, in the vast majority of finite deals (if not all of them), there was no reliable historical loss data; not even proxy or industry data.

So, instead of looking backward, we started looking forward, and asked this question: "even though we can't quantify it, isn't it safe to say that, given the exposure (the key), this bomb could drop at any time, causing considerable loss?" Why, of course it could! Yes, that's the ticket, and if we come close to purchasing a reasonable amount of excess risk transfer, we're good to go.

So let's revisit our example. Is there a chance that a former landfill would eventually begin to bubble up a bit of primordial ooze? You bet there is.

Back to the Future

Fast forward through the 1990s and wham—Enron happens. Enron's meltdown, while not directly related to the misuse of finite reinsurance, ignited the New York Attorney General's conflagration that spread through the financial services industry, extracting enormous fines from the like of Merrill Lynch among others, and eventually focused on the business of insurance. Along with other infractions (bid-rigging, etc.), finite reinsurance came under the microscope. In April 2006, ACE Insurance Company admitted wrongdoing in a 2000 finite reinsurance transaction involving another insurer. ACE admitted that the finite arrangement was designed to bolster its financial position, and ultimately contained no actual risk transfer. It was fined $80 million. Zurich, St. Paul, and Chubb were also nabbed doing these deals. They were fined $153 million, $77 million, and $17 million, respectively.

Needless to say, this sort of thing rather shut down the market for finite reinsurance. Already prior to the attorney general's investigation, conservative financial consultants were warning against relying on the aforementioned 10/10 rule, recommending something like 30/30 or higher.

Looking Ahead

So, what's the future for FAS 113? Back in May 2006, the FASB issued an Invitation to Comment on proposed changes. Specifically, it proposed that every insurance transaction should be examined, and if a noninsurance component is present, it should be recognized separately and accounted for using deposit accounting instead of insurance accounting—bifurcation. As of December 2007, the FASB announced the following.

The Board plans to issue an Exposure Draft [of its revised Statement 113] in the second quarter of 2008 that will (1) clarify the level of insurance risk transfer required for a contract to be accounted for as reinsurance, (2) clarify that noninsurance entity policyholders must evaluate whether contracts they hold transfer significant insurance risk, and (3) improve insurance and reinsurance disclosure requirements.

Stay tuned, my friends. When this comes down, you'll be one of the first to hear about it.