Arnold Kling  

AIG, CDS, Suits, and Geeks

Bailout Rage... Krugman's Tin Ear...

The Washington Post is in the middle of a three-part series on AIG and credit default swaps. A lot of the focus is on corporate infighting.

A Brooklyn College graduate, the 42-year-old [Joseph] Cassano was not one of the "quants" who had mastered the quantitative analysis and risk assessment on which the firm had been built. He had no expertise in the art of hedging. But he had excelled in the world of accounting and credit -- the "back office," as it is known on Wall Street.

Cassano became head of the unit at AIG that did credit default swaps. Actually, I think that a strong back-office guy would be pretty good to have in that position. You don't just have economic risk. You also have operational risk. It sounds to me as though Cassano would have been good at controlling operational risk.

The man with whom I identify in the article is Tom Savage, a geek who headed the financial products unit in the late 1990's. He thinks that credit default swaps were highly profitable and low risk when they started, but seeing what ultimately happened, he regrets getting into them.

I feel that way about my role at Freddie Mac (although I was never a senior executive, as Savage was at AIG) in promoting credit scoring for mortgage underwriting. It seemed like a great idea at the time, yielding fewer errors and lower cost than traditional underwriting. But it helped pave the way for private securitization, because Wall Street came to treat FICO scores almost as a substitute for the guarantees of Freddie Mac and Fannie Mae. And, somehow, everybody forgot about the importance of borrowers putting up equity to buy homes, and the rest is history.

The Post story leaves more questions unanswered than answered. It refers to hedging strategies that the folks at AIG believed in prior to 2005 (when this part of the series ends. The next part picks up tomorrow.) What were those hedging strategies? Was it dynamic hedging, meaning that if your credit default swap started to go bad you would short similar securities?

Another question concerns the destruction of AIG's chairman Maurice Greenberg by Elliot Spitzer. The article mentions a transaction involving "finite insurance," which is something I have never heard of. The article suggests that it was an accounting trick, designed to dress up the balance sheets of both sides of the transaction. I want to know whether this was a "foot fault" (Greenberg's term) or a serious fraud. Most importantly, I want to know who created the incentive to do this sort of transaction--was it insurance regulators in the first place?

Mostly, I am waiting (presumably this will be in the series finale tomorrow) to learn how AIG got into mortgage-backed securities and whether it was the suits or the geeks who got it wrong. That is, did the geeks under-estimate the risks, or did the suits run past warning signals flashed by the geeks?

Comments and Sharing

COMMENTS (6 to date)
Grant Gould writes:

That is, did the geeks under-estimate the risks, or did the suits run past warning signals flashed by the geeks?

I'd bet on some of both: Nothing in this world is ever truly monocausal. The geeks disagreed (when on earth has a large group of geeks ever been unanimous?), the suits picked the geeks whose answers they liked to listen to, and the rest fell into line. Enough blame to go around, particularly with nobody keen to claim leadership.

Robert Scarth writes:

I'm not an expert but as far as I understand it "Finite Insurance" is (was) basically a financing deal dressed up as an insurance contract. Usually the insurance risk transferred is quite small, the risk transferred usually involved the timing of payments, interest rate risk, or some other financial risk. It is sometimes called "Financial Reinsurance". From

"A pure financial reinsurance contract for a non-life insurer tends to cover a multi-year period, during which the premium is held and invested by the reinsurer. It is returned to the ceding company - minus a pre-determined profit-margin for the reinsurer - either when the period has elapsed, or when the ceding company suffers a loss. Financial Reinsurance therefore differs from conventional reinsurance because most of the premium is returned whether there is a loss or not: little or no risk-transfer has taken place."

The "Finite" part of the name denotes that the loss to the insurer was quite limited.

As far as I understand these products purely existed due to regulatory arbitrage. There may also have been some tax advantages.

The wikipedia article gives a reasonable overview

taper writes:

Finite reinsurance is when there is little or no probability that the insurer will lose money.

You're on the right track with your question on whether this is a regulatory thing. From an accounting perspective, insurance is simply the trick of paying someone to take a potential economic risk for you and recording the transfer of those risks in the income statement as opposed to the balance sheet.

When claims comes up, the insurer records the claim as an expense and the insured as income. Bingo, smoother results.

There are lots of ways this happens: guarantees that all claims will be repaid or extremely large premium payments (say 90% of insured limit) for events that you know will happen or you know HAVE happened.

It seems that AIG and other institutions became lazy and relied on the investigations and reasoning of others. Fannie and Freddie were recklessly buying subprime mortgage backed securities, funded by their government guarantee. The major credit ratings agencies were eager to go along with government pressure to approve subprime MBS bonds. They were comfortable having a an effective monopoly as government-approved companies.

So, I think the suits got it wrong, thinking: Why stay out of a market that everyone else has confidence in.?

AIG got into trouble evaluating the risk of default for banks holding mortgage backed securities. Banks bought AAA rated mortgage securities, supposedly rock-solid. The AAA ratings were given by government-approved ratings agencies. Insurers like AIG and other institutions relied on those ratings, and insured those banks by selling credit default swaps, which seemed to be a conservative risk. The actual risk of those mortgage securities was far higher.

The agencies said that they relied on facts presented by issuers, and that they are not responsible for conducting their own factual review. So when S&P and Moody's rate a pool of mortgage loans, they don't examine any of the individual mortgages within the pool. Amazingly, S&P told the Wall Sreet Jounal that
"We are not auditors; we are not accounting firms."

Andrew Garland
Credit Rating Agencies Misled Investors

dWj writes:

My understanding was that finite reinsurance wasn't so much a way around regulations as it was a way around accounting rules; it was relatively straightforward to structure a contract that had relatively little substance other than to make accounting measures diverge from economic reality, so that AIG could report higher GAAP earnings than it should have. (This surely has regulatory implications as well, but I think it was more about management being able to lie to investors, and possibly themselves.)

I'd also go with Grant Gould's intuition; someone along the line said, "well, this is likely a bad idea, but it could be okay", and it got passed up the chain to people decreasingly able to understand hedged predictions. (Pretending to be sure about things is endemic to politics and to the upper levels of business, who are driven as nuts by accurate conveyances of uncertainty as I am by refusals to accurately convey uncertainty.)

Richard writes:

Most of the comments on finite coverage (reinsurance in the case of AIG) are accurate. The irony here is that AIG was the assuming company, i.e., it acted as the reinsurer of General Reinsurance. Gen Re ceded (reinsured) approximately $500 million of its loss reserves with AIG. AIG put this amount on its balance sheet thereby increasing its loss reserves. Though seemingly counterintuitive, it was prompted by an analyst report indicating that because AIG's loss reserves were less than the analyst anticipated, AIG's balance sheet was lacking conservatism. The finite transaction therefore increased reserves which, of course, is atypical. Interestingly, Gen Re accounted for this transaction properly. It neither reduced its reserves nor took any P&L benefit, accounting for the cash transfer as a deposit. Despite this proper accounting by Gen Re of the "non-risk" transaction, federal prosecutors have successfully litigated against four of their executives for essentially enabling AIG in its transaction.

Comments for this entry have been closed
Return to top