cfo.com

Print this article | Return to Article | Return to CFO.com

Infinite Risk?

The abuse of finite insurance uncovered at AIG and other insurers may be just the tip of the iceberg.
Ronald Fink, CFO Magazine
June 1, 2005

What do many corporate buyers of insurance have in common with American International Group? Perhaps more than they would like to admit. Like AIG, many companies in the past few years have bought finite insurance, which transfers a prescribed amount of risk for a particular liability. What regulators now want to know is, how many companies, like AIG, have used finite insurance to artificially inflate their financial results?

It's a tough question to answer, since there's nowhere near enough disclosure in corporate financial statements to determine whether buyers are properly accounting for such contracts. Consider the case of Delta Air Lines. Several years ago, when the financially troubled carrier wanted to remove from its balance sheet at least part of its growing liability for free mileage grants to frequent fliers, Delta used its captive insurance subsidiary, Aero Assurance, to buy reinsurance for the purpose. (Reinsurance was necessary because a captive's results are otherwise consolidated on a parent's balance sheet.)

The deal enabled Delta to take "a very large liability" off its balance sheet, the company's risk director at the time, Chris Duncan, told CFO in an interview for the March 2001 online article "Delta's Strategy for Reducing Turbulence." But Duncan would not go into further detail, and nothing about the transaction was disclosed in Delta's 10-K. Indeed, Delta has said very little about the deal. Duncan has since left Delta to join insurance brokerage giant Marsh Inc., a division of Marsh & McLennan, and did not respond to a request for comment. Delta also failed to respond to such a request. But experts say a frequent-flier program is the type of liability that lends itself to finite coverage.

Loans in Drag
Delta, of course, is hardly alone in using insurance in this way. Indeed, there's nothing legally wrong with doing so. The trouble is, regulators suspect that much of the activity in this arena is really financing masquerading as insurance, because it involves little or no transfer of risk. Without more significant risk transfer, the accounting rules say there should be no change in the liabilities recorded on a buyer's balance sheet.

Given the general lack of disclosure, it is impossible to say how far the abuse goes. But AIG has been on the other end of similar deals with PNC Bank and with a Plainfield, Indiana-based cell-phone distributor named Brightpoint. In settling charges of aiding and abetting accounting fraud in the PNC case in late 2004, AIG acceded to a Securities and Exchange Commission demand that it install an independent monitor to oversee its operations. Reportedly, the monitor is not only supervising new AIG deals, but also investigating others that the company engaged in around the same time.

Meanwhile, the SEC and other authorities, including New York State Attorney General Eliot Spitzer, the National Association of Insurance Commissioners, the Financial Accounting Standards Board, and the Federal Bureau of Investigation are hunting for problems elsewhere. General Electric, which like AIG both sells and buys finite insurance, is the latest company to be subpoenaed by the SEC in connection with the product. As Scott Taub, the SEC's deputy chief accountant, told a conference on financial reporting at Baruch College in early May: "Finite insurance is not solely an insurance company issue. It affects buyers as well as sellers. Any number of companies have this issue." The potential upshot for CFOs of offending companies: a new round of SEC enforcement actions, followed inevitably by shareholder lawsuits.

In the most worrisome cases, premiums for finite insurance fully cover the discounted value of all potential losses, and any unclaimed money at the end of the contract is returned to the buyer. All too often, say critics, the only cost to the buyer is the interest earned by the seller on the money, an arrangement unmistakably a loan rather than insurance.

And if the liabilities covered by the policy are nonetheless removed from a buyer's balance sheet, a company would make its financial condition look stronger than it really was. To be more precise, loans are counted as liabilities, whereas insurance is treated as an asset, so a company's net worth would be overstated as a consequence. This, in fact, is exactly what AIG has admitted doing when it bought such a policy from Berkshire Hathaway's General Reinsurance subsidiary several years ago. In that case, the policy inflated AIG's reserves and other assets by $495 million.

Artful Dodgers
So far, regulators haven't reported much abuse beyond what has been revealed at AIG and a few other insurers. For that matter, even evidence of legitimate use of finite insurance is slim. Of the 524 active captive subsidiaries domiciled in Vermont as of last March, Leonard Crouse, deputy commissioner of captive insurance for the Vermont Department of Banking, Insurance, Securities, and Health Care Administration, estimates that 5 or 6 used finite insurance, but declined to identify them.


If the same proportion were found among captives elsewhere, the number of those using finite insurance would rise to about 50. That, of course, is a tiny minority. "A lot more [captives] look at" finite insurance than buy it, asserts Kathryn Westover, a captive insurance consultant and business-development representative of Barbados, where 257 active captives were domiciled as of last March. (Bermuda was still far and away the most popular domicile, home to 1,150 active captives, or 24 percent of all active captives, followed by the Cayman Islands, Vermont, and Guernsey in the Channel Islands.)

Nonetheless, a report in the June/July 2004 issue of the industry publication Captive & ART Review suggests that the number of such deals might be much higher, since most finite-insurance business has been transacted in locations outside the United States. In any case, the report suggested that the leading provider of finite coverage for captive insurers, Munich­American Re, had written almost $25 million in finite premiums in the previous 12 months. And a March 2004 survey of captive insurance subsidiaries by The ACE Group of Cos., another large provider of finite insurance, found that almost 12 percent were planning to buy finite coverage in the coming year (see "Just How Finite?" at the end of this article). "There's pretty extensive use of it" among captives, asserts a former industry executive who asked not to be identified.

Despite the problems with finite insurance uncovered so far, supporters staunchly defend its use. "Finite is a legitimate method of financing risk," insists Westover. Her view was echoed in mid-April by Ellen Vinck, vice president of risk management and benefits at U.S. Marine Repair, who was recently elected president of the insurance industry trade group Risk and Insurance Management Society. "I'd hate to see all finite insurance and other forms of ART [alternative risk transfer] cast in the same light," Vinck told CFO.

10 Percent or 1 Percent?
But the line distinguishing the legitimate use of finite insurance from the illegitimate is extremely fine. To remove a liability from its balance sheet, according to FAS 113, the relevant standard under U.S. GAAP, the buyer of such a policy must transfer a "significant" amount of risk to the seller policy, and the seller must face a "significant" likelihood of a loss.

Absent more specific guidance, the insurance and accounting industries have established a rule of thumb that says that the seller of a policy must face at least a 10 percent chance of a 10 percent loss on the liability. But as one FASB insider who asked not to be identified put it, "10 percent times 10 percent equals 1 percent. Is that a significant amount of risk?" No, say critics of finite insurance, such as Frank Cacchione, a partner in the New York office of London-based PA Consulting Group. "That's pretty thin," he observes. Hence FASB's recent decision to put the issue on its agenda for possible new rulemaking.

But Cacchione points out that even some companies that profess to adhere to the "10-10" rule-of-thumb may agree in side letters that they will make sellers whole for any losses. (Note that AIG did that with its General Re finite coverage, and that Enron had a side agreement that masked the true nature of its spurious off-balance-sheet deals from auditor Arthur Andersen.) Such letters, of course, escape the attention of auditors.

When asked how many captives that buy finite insurance agree through side letters to make sellers whole for losses, the former industry executive who requested anonymity replies: "Do you have a phone book?" Says Cacchione, "More disclosure is required" of the terms of finite-risk contracts.

Changing the Rules
Of course, that's an issue that must be addressed by the SEC, not FASB. "That's one for enforcement," Taub reminded CFO after his remarks at Baruch College. At this point, what any of the rule makers will do is anyone's guess. At the minimum, however, FASB is likely to require buyers to meet a minimum threshold for the estimated severity and frequency of the potential losses, ventured FASB chairman Robert Herz after addressing the Baruch conference. But Herz also told CFO that the board might follow the lead of the International Accounting Standards Board (IASB) on this question, in which case insurance would be treated like any other financial instrument. That would mean that insurance buyers couldn't move liabilities off their balance sheets even if 100 percent of the risk for them had been transferred.

"Guess who most loudly opposed that [IASB] rule" when it was being debated? Herz asked, and then answered his own question: "AIG."

Without a clearer picture of regulators' intentions, it's impossible to predict what companies would do in response. But if FASB goes as far as Herz suggests, it is safe to say that companies would have to unwind or restructure many existing policies, finite or otherwise, and then restate their results. Vermont regulator Crouse, for one, doubts such an outcome. He asserts that all of the captives domiciled in the state that use finite insurance account for it as financing. At the minimum, tougher rules are likely to discourage new transactions.


"Auditors are paying more and more attention to these transactions," notes Arthur Koritzinsky, a managing director of Marsh, which manages more than 1,000 captive subsidiaries for their corporate parents. And because of the extra scrutiny, adds Westover, finite-risk buyers "are far more cautious than they used to be." Given the rising regulatory concern over yet another questionable method for engineering financial results, that is hardly surprising.

Ronald Fink is a deputy editor of CFO. David M. Katz, deputy editor of CFO.com, contributed additional reporting.

Just How Finite?

There is anecodtal evidence that interest in finite insurance was rising among companies in the pharmaceutical, consumer-products, and energy industries prior to the fallout over AIG's deals.

"While finite risk is not a new arrival in the world of captive insurance, it's becoming more widespread as it becomes better understood," reported industry publication Captive & ART Review in an article posted on its Website in June 2003. The article also stated that provider The ACE Group of Cos. had recently structured a finite-insurance program for an unidentified pharmaceutical company that had seen its captive subsidiary's own liabilities rise from $50 million to $400 million because of such exposures as product liability and product recall.

The same article reported that Mark Lima, then executive vice president of ACE Financial Solutions, had had discussions earlier in 2003 with a major consumer-products company about using finite insurance as a result of the rising cost of casualty coverage provided by its captive subsidiary. "The company was concerned about whether the level of capital in the captive was sufficient to support the additional risk that would be assumed," explained Lima. —R.F.




CFO Publishing Corporation 2009. All rights reserved.