Not very long ago, the use of finite insurance looked much better than it looks today.
For years before the current scandals at American International Group Inc., the arrangements were touted as a sleek way for a corporation to control its risk funding and shed liabilities from its balance sheet. During times like the post-9/11 period, when property and liability coverage was overly costly or hard to come by, finite risk seemed a legitimate, more predictable alternative to traditional insurance.
That’s not to say that critics haven’t long questioned one key aspect of the deals: They don’t tend to transfer much actual risk from insured to insurer. As the name implies, finite coverage sharply limits the amount of loss the insurer can suffer. Instead, what the carrier does cover is a timing risk — the chance that over a specified period a bunch of claims will have to be paid at an inopportune moment.
In its most common form, corporations plunk down a premium nearly big enough to cover all the expected losses into an account held with the insurer. If the cost of losses turns out to be less than the premium, the carrier gives back the difference to the insured; if the losses turn out to be greater, the insured pays an additional premium to the insurer. To accounting skeptics, finite insurance can look like a way for a corporation to retain risks without really retaining them and transfer risks without really transferring them.
To be sure, there’s little doubt that finite coverage can enable executives to spruce up their companies’ financials if they can record the arrangements as insurance rather than as deposits. Indeed, until recently such balance-sheet cleansing was a prime selling point of finite-insurance marketers. “The product was designed to improve financial statements,” recalls Andrew Barile, an insurance and reinsurance consultant in Rancho Santa Fe, California, who has worked on finite deals.
During the go-go late 1990s, the idea of using insurance to help companies hit net-income targets could be alluring to management, and regulators and auditors could be counted on to let the deals slip through without much scrutiny. But in the post-Sarbanes-Oxley era, the mere suggestion of “earnings management” gets enforcement officials riled.
How Many Brightpoints?
The 2003 Brightpoint case seems to have fueled a fair amount of the regulatory activity involving the use of finite coverage. Without admitting or denying guilt, AIG and Brightpoint agreed to pay $10 million and $450,000 in civil penalties, respectively, to settle fraud charges brought by the Securities and Exchange Commission. The SEC had charged that AIG created an “insurance” product that enabled Brightpoint to shave $12 million in losses from its books and overstate its 1998 earnings by 61 percent.
Even though Brightpoint had merely deposited cash with AIG that the insurer refunded to the company, the insurer made it seem as if the wireless-device distributor paid premiums for the coverage, the commission charged.
How many more Brightpoints will there be? Citing unnamed sources, a recent account in The Wall Street Journal forecast that federal and state authorities will turn their attention to investigating the uses of finite coverage by non-insurance companies once probes of AIG and other insurers are finished. But state insurance regulators are hard-pressed to produce examples of abuse other than Brightpoint. “We have no evidence that it’s a pervasive problem,” says Joseph Fritsch, director of insurance accounting policy at the New York State Insurance Department.
At the moment, the government investigations that have grabbed most of the headlines are focused on finite-risk reinsurance arrangements among insurance-industry players. New York Attorney General Eliot Spitzer’s probe of a $500 million finite reinsurance deal between AIG and Berkshire Hathaway’s General Re division, for example, played a role in the departure of Maurice Greenberg, the former chairman and chief executive officer of AIG.
If that were all, the matter would be a tempest in an insurance-industry teapot. But current probes by Spitzer and the SEC involving “non-traditional, or loss mitigation” insurance products employed by AIG, General Re, ACE Ltd, and St. Paul Travelers, along with abundant press coverage, are bound to taint finite insurance even when it’s purchased by non-insurance companies.
That would be a shame, maintains Christopher Culp, an adjunct professor of finance at the University of Chicago Graduate School of Business who specializes in alternative risk transfer, if the stigma deters finance executives and risk managers from using finite coverage. “It would be a huge mistake,” he wrote in an E-mail to CFO.com, if companies were “scared away from what is fundamentally a legitimate and potentially quite beneficial type of transaction, just because of some regulatory attention and purported abuses.”
Sealing the Cookie Jar
Culp suggests that adroit use of the coverage can protect companies from their managements’ own worst impulses. That’s because, in effect, the corporation hands over the cash to an insurer with instructions not to return it unless it’s being used to cover losses or the risk is over.
Far from being motivated by a desire to smooth income, most executives use it as “a way to take a reserve against a loss credibly,” he contends. “If you just set cash aside for a loss, investors are suspicious you might spend it on something else.”
The arrangement “leads to a higher quality of earnings than if the firm doesn’t reserve for a major loss or just tries to set money aside internally,” according to Culp, who notes that having finite coverage can shield companies from charges of setting up “cookie jar” reserves to inflate future results.
Another advantage is that the coverage can help companies that don’t want to buy traditional insurance to fix the cost of hard-to-quantify exposures over a span of years. Barile, the consultant, says that in the late 1990s, he worked on an arrangement to insure gambling casinos against a loss of revenue resulting from a downturn in the Las Vegas economy, for example. Such a risk, he notes, could include the possibility of a terrorist attack — an extremely difficult exposure for traditional insurers to price.
Further, finite insurance can come in especially handy in covering severe risks that are outside the core functions of a company, observes Culp. For instance, chemical marketers and distributors could incur latent pollution liabilities that might not crop up for many years. “To assure investors that they have credibly managed any such losses, finite is a good answer,” he adds.
With merger activity heating up of late, corporate negotiators could also rediscover finite arrangements as a way to grease a deal. In the past, a self-insured company facing many workers’ compensation claims could overcome resistance to a merger by “selling” its loss portfolio to a finite-risk insurer, says Barile. By transferring the claims to the insurer, a suitor or a target could present itself as liability-free.
Transfer or Restate?
Despite its benefits, however, finite coverage carries with it the distinct risk of a restatement. Under generally accepted accounting principles, payment from an insurance company isn’t an “insurance” recovery for accounting purposes unless the policy transfers some risk to the insurer. If a policy doesn’t transfer risk, it’s a financing arrangement in which all premiums are treated as deposits.
CFOs and risk managers should be particularly careful to account properly for finite policies that insure past events rather than future probabilities, experts say. Auditors and the Internal Revenue Service have frowned on such retroactive policies because it’s hard to determine whether risk has actually been transferred, says Joel Chansky, a principal in the Boston office of Milliman Inc. At the same time, for example, “no one knows the ultimate value of workers’ comp liabilities or claims reserves, so there is risk there,” the consultant adds.
To make sure that enough of a peril is transferred, risk managers can blend different kinds of coverage into the policy, according to Chansky. They might, for instance, combine a prospective earthquake exposure for 2005 with known workers’ compensation claims from 2004. In that case, the quake hazard might provide a risk-transfer justification lacking in the workers’ comp losses.
Other potential accounting foul-ups can occur in connection with the use of finite coverage by a company’s captive insurance subsidiary. Under a typical arrangement, the captive might insure the company’s cheaper workers’ compensation claims itself and buy finite-risk reinsurance to cover the more expensive losses, says Keith Buckley, a group managing director with Fitch Ratings.
Since captives are subsidiaries, their results are typically reported together with those of the parent company. If the captive’s reinsurance arrangement is found to not involve enough risk transfer, the parent company would have to make an adjustment “to take all the losses into the income statement at one time again,” notes Buckley.
Compounding the risk of restatement for all finite deals is the currently vague definition of “risk transfer.” While a “10/10” rule — a 10 percent probability of a 10 percent loss of premium — is most often used, the standard is ultimately “what auditors and regulators say it is,” according to a Fitch Ratings report issued late last year. Some clarity might be on the way, however. Earlier this month, the Financial Accounting Standards Board reportedly announced that it would look into what risk transfer actually means when it comes to finite coverage.
It’s the lack of risk transfer, then — and not the much-dreaded label of “earnings management” — that turns out to be the greatest threat to finite insurance. Merely using the product to make financial results look better might not be a deal killer, according to Michael Moriarity, director of the capital markets bureau of the New York Insurance Department.
If a finite arrangement transfers enough risk, “it can smooth earnings,” he says. “There’s nothing wrong with that.”