September 10, 2005

Hidden Loans in the insurance Industry

On October 29, 2004, Wall St. Journal reporters Theo Francis and Christopher Oster authored an article titled, Hidden Loans May Be Common Practice In Insurance Industry. (The article is accessible via the WSJ archives for a fee but can be found other places on the net. See this Google search.

The article begins:

Insurance companies for years have been buying insurance policies for themselves that are akin to the product at the center of a criminal investigation into whether American International Group Inc. helped a cellphone distributor manipulate its earnings. Critics say the policies are sometimes insurance in name only. That is because the premiums or other payments are so big that the seller assumes little or no risk, making them like loans that help buyers smooth their earnings and shore up their stock price. The reason: Insurance proceeds count as income and offset losses, while a loan must be counted as a liability -- a debt that must be paid off over time.

I emailed the authors (see the text of my email below -- "continue reading...") and Theo Francis replied back, "Ed, thanks for your very helpful note -- I've gotten a lot of feedback from people on this article, both in the industry and outside, and yours was easily one of the more informed. Would you have time in the near future to talk about some of these issues on the phone?"

I'm posting all of this to my new weblog because I'm planning to do a follow-up blog post on the issue, now that nearly a year has gone by. Stay tuned.

My email reply to the WSJ reporters:

Very good and timely article. Keep digging, you've just begun unraveling the ball of string...

You got it. Finite is a loan (usually with indeterminate due date). Finite "smoothing" has spread beyond reinsurance into mainstream corporates world-wide, to the detriment of F/S users. Most main (re)insurers actively underwrite these products in several forms. Paradoxically, few insurers "make money" on finite because they don't manage it as credit risk. FASB itself, is one of the causes of finite's murky disclosure.

For follow-up articles, you may want to consider that..........

"AIG [refers to a] one-of-a-kind policy". AIG is INCORRECT. AIG, and many other major (re)insurers have several groups that offer finite insurance, many in competition with each other, at the same company. Just check leading insurer websites and you will see references to finite solutions. And many conventional product lines (e.g. D&O, surety, workers comp, business interruption, etc.) are structured in finite form to achieve the same result. This enables carriers to truthfully claim: "we don't write finite anymore [we closed the finite dept]", while writing earnings-smoothing covers in other guises. Continuing investigations [SEC, Spitzer, CT, Cal, NJ] will likely show that finite earnings-smoothing structures have spread far beyond the P&C industry, where they originated and are pervasive, into mainstream corporates, in the US and overseas. Virtually all (re)insurers are selling these.

"the economics favor the insurer". NO--they are neutral. That's like saying "economics of debt favor borrower over lender" or vice versa. The market clearing price doesn't favor either. Finite is oddly an under-priced loan and capital markets penalize under-pricing quickly, with disasterous results. Consider Centre Solutions--long the leading light in finite insurance. Look at their WSJ ads circa 2002: (effectively) "we can do things the capital markets can't". Centre lost Zurich over $1BN from credit-related covers (one wonders how many of these were reinsured by Converium). In practice, very, very few (re)insurers manage finite as credit risk; those that do get the capital market spread, nothing more.

"disclosure is murky...industry specialists have a hard time deciphering [significant risk transfer from in-substance loans]". YES and NO. Financial (re)insurance disclosure is governed by FASB 113, which practitioners find easy to structure around. Deciphering is a problem because the insurance industry is complex to outsiders [even experienced industrial analysts who are unfamiliar with it].

* So, a P&C securities analyst reviewing a P&C insurer can possibly identify finite exposures in the insurer. However, not being familiar with commercial practices in other--energy, transportation, retail, banking, etc,--sectors, he would find it difficult to identify accounting effects on earnings or f/s elsewhere in industry (in the absence of a meltdown). Similarly, an industrial securities analyst familiar with his industry specialty will have a hard time finding earnings smoothing via a finite arrangement.

* This is because the finite "policies" masquerade in different forms. One major Financial Services company recently entered into an earnings-smoothing arrangement [i.e. finite deal] with a major reinsurer, under the guise of a "marketing joint venture", where credit losses in one year were, through the arrangement, spread over several years.

In my view, FASB itself is a cause of "murky disclosure". FASB 113 and FASB 133 (for derivatives) contain quantitative tests which structurers/auditors model, pass and take comfort in. Perhaps recent SEC decisions moving auditors to consider qualitative materiality of F/S presentation will clarify these and other disclosure issues?

This arcane area needs some sunshine to get it up to capital market--professional and disclosure--standards. Hope this helps your pursuits. Thanks and regards.

Posted by at 05:32 PM | Comments (4202)