Selected writings by David Fiderer
Joe Cassano is a very good liar, which is why it would be so hard to prosecute him for perjury. When testifying before The Financial Crisis Inquiry Commission, the former head of AIG Financial Products kept blending in half-truths with his audaciously dishonest claims, so that the overall effect was nonsensical. For instance, to justify his outrageous claim that, “the books were generally considered fully hedged,” he explained that “we were using it basically in actuarial basis …[so] it’s not hedged in the conventional sense.” (Translation: The book was never hedged in any sense. Nor was there any actuarial analysis, only a reliance on triple-A credit ratings.) These rhetorical tricks were designed to throw sand in everyone’s face. But his tactics seem to have worked. The staunchly unregenerate Cassano framed a media narrative that deflected away from his dishonesty and gross incompetence.
Here’s a reality check on some of his more ridiculous claims, in order of appearance:
1. Cassanos’s Claim: AIGFP never compromised its high underwriting standards.
The Truth: AIGFP had no underwriting standards pertaining to the most important risk, which affected AIG’s liquidity.
Commission Chairman Phil Angelides asked Cassano if he understood the subprime risks he insured. Cassano stonewalled with a lot of doubletalk:
Angelides: I want to talk to you about this, that these were represented as multisector CDOs. But if you look at — we did a sample of some of these in 2004, 2005, 2006, they were almost overwhelmingly residential-backed and very substantially subprime. For example, in the survey we did of some of these CDOs that you issued protection on, 84 percent were backed by RMBS residential mortgages in ’05, 89 percent in ’06. And just as an example, while you indicated you decided to stop writing on subprime instruments in January of ’06, for example, you backed an instrument called RFC III where that CDO was 93 percent subprime and seven percent HELOC home equity loans.My question for you, Mr. Cassano, is was there — you said you did thorough due diligence. Were you aware of the quality of the mortgages? Do you do direct analysis of the loan data? Were you confident that you had a full understanding of the nature of what you were backing?
Cassano: Chairman Angelides, the numbers that you are referencing in these portfolios, I don’t know specifically. I’m happy to look at them again and go through that with you.
Reality Check: Cassano insults everyone’s intelligence by refusing to admit that he insured tens of billions of dollars of toxic investments that were primarily comprised of subordinated tranches of subprime mortgage securities. The CDOs that caused the collapse of AIG were no more “multi-sector” than the government of Iceland is multiracial. His unwillingness to acknowledge the obvious truth is a rhetorical device intended to cast doubt and cause confusion among listeners and the media.
Cassano: But I think to answer your question more directly, we never diluted our underwriting standards at any point in time. We had rigorous standards, standards set by the AIG credit risk management that we then employed in underwriting these transactions.
Reality Check: Cassano said he would answer the question “directly” and then didn’t. The question asked whether he personally understood the risks associated with the subprime mortgages embedded within the CDOs. It wasn’t about what “we” did, and it wasn’t about some dilution or non-dilution of some undefined underwriting standards that may have had nothing to do with subprime risk.
What remains indisputable is that there were no standards for protecting AIG’s liquidity. AIGFP was in the business of trading derivatives. The liquidity risk, pertaining to collateral postings, was never even considered when these deals were approved. It sold $78 billion worth of long-term credit default swaps that were unhedged. As part of those swap agreements, AIGFP agreed to post margin, or cash collateral, without ever attempting to define the basis by which those collateral postings would be calculated. The stupidity of Cassano and other top managers at AIG cannot be overstated. They operated like an 11-year-old driving a motorcycle. The current chief risk officer at AIG explained:
Angelides: Mr. [Robert E.] Lewis, you are the chief risk officer. Anything you want to add to this?
Lewis: I would state that the risk issues that were the focus of the attention at AIG were around the actual credit risks in the underlying portfolios. And our — the rigorous work that we did together with FP was to determine what the likelihood was of suffering credit losses through defaults and losses in the underlying mortgages.
The liquidity aspects were something, quite frankly, just didn’t focus on to the extent that we now know we should have. The — these instruments up until the time of the crisis had traded in very narrow bands, highly liquid AAA securities, until the crisis occurred when they traded off quickly and then there was no market. So —
Angelides: But were you — but you — were you aware that there was a liquidity provision, you weren’t, were you?
Lewis: No, I was not until —
Angelides: All right.
Lewis: — till the date I just testified. [i.e. July 2007, several years many of the deals were booked]
Reality Check: Lewis, like Cassano, had no idea what he was doing. Every trader, every junior risk analyst, every deputy assistant treasurer with the most minimal level of competence knows the dangers of selling an unhedged derivative. You can lose money when the swap terminates, and you can lose liquidity before the swap terminates. The longer the tenor of the swap, the bigger the risk. This isn’t some honest mistake, some detail that could ever be overlooked. A financial company depends on liquidity the same way that a mammal relies on oxygen to stay alive. Lewis acted like the traffic cop who looks at cars in the left lane and ignores the vehicles in the right lane. The only plausible explanation why Lewis still has his job is that the AIG does not want to expose more dirty laundry.
Their stupidity was compounded further their willingness to post collateral based on the “market value” of the CDOs. Lewis’s claim that “these instruments up until the time of the crisis had traded in very narrow bands, highly liquid AAA securities, until the crisis occurred,” is further demonstration of his cluelessness and/or dishonesty. The triple-A tranches of these CDOs didn’t trade. Why would they? AIG had assumed virtually all the credit risk in the most senior tranches. These CDOs never had an ascertainable market value based on comparable sales or industry benchmarks, according to PriceWaterhouseCoopers, AIG’s auditor, and Deloitte & Touche, Maiden Lane III’s auditor. Both accounting firms designated the CDOs as Level 3 assets, explained here.
Another level of stupidity was their disregard how residential mortgage-backed securities work. These mortgage bonds are valued according to the credit losses that are expected in the future, not according to the actual losses that have already been recognized. It takes a long time, typically about a year, to recognize an actual loss on a loan after the borrower first becomes delinquent. Usually, a notice of foreclosure is first presented after the 90-day delinquenciy period has passed, later, the lender commences a procedure whereby it “buys” the residence, and then the property sits on the market until it is sold to partially pay down the loan. Because of the time lag, you need to be concerned about paying cash margin long before the final tally of actual losses on a mortgage pool. Goldman had always understood this; Cassano’s people didn’t. Cassano and his management team, who were in the business of trading derivatives, didn’t know squat about liquidity risk.
Because AIG never understood the risks it was taking on, it agreed to contract language that gave Goldman and other CDO banks the opportunity to jerk the company’s chain indefinitely. Since the valuation of the CDOs could be debated endlessly, there was an ongoing risk that, at any given moment, Goldman could declare its unmet demands for cash collateral to be an event of default. One default can quickly trigger a series of cross-defaults forcing a bankruptcy. This was why the rating agencies told AIG and the New York Fed that the contingent liabilities tied to these CDOs needed to be removed no later than November 10, 2008, or AIG would suffer further downgrades.
2. Cassano’s Claim: The CDOs held by Maiden Lane III performed in line with his expectations.
The Truth: The CDOs performed in line with the $35 billion write-down taken by AIG when Maiden Lane III was created.
Cassano: [M]any of these multi-sector CDOs that we did now reside in Maiden Lane III…And to date that vehicle is performing. I think, you know, I’m sure the commission knows the statistics, the federal government lent that vehicle $24 billion. To date that vehicle has repaid $8 billion through the performance of these transactions. And as far as I can see from where I sit when I look at the portfolio residing in Maiden Lane III, I don’t know — I don’t think any of the transactions have pierced the attachment levels that we had set in our underwriting standards… And as we move through this and we come through the financial crisis, the only thing I can do is look at the existing portfolio and say that it is performing through this crisis and it is meeting the standards that we set.
And it’s not the credit risk here that eventually became the issue at hand. These — my point has been that the underwriting standards and the credit risk within these transactions have, to date, been supported and still perform.
Reality Check: To recap simply, Cassano insured the CDOs acquired by Maiden Lane III for $62 billion. AIG had paid out $35 billion in cash collateral to the CDO banks before Treasury and the New York Fed began negotiating with the banks. When Maiden Lane III paid the banks an additional $27 billion to acquire the CDOs and tear up the credit default swaps, AIG recognized a loss of $35 billion. Deloitte & Touche valued the CDOs at $27 billion on December 31, 2008, and that value more or less held steady as of December 31, 2009. Cassano wants to make it sound as if the CDOs’ performance, after recognition of the $35 billion loss created by him, somehow validates his own reckless performance.
When Cassano said, “the only thing I can do is look at the existing portfolio and say that it is performing..” it became obvious that he was lying. Cassano can’t look at the performance of the CDO portfolio because he no longer works at AIG and the performance reports on all CDO portfolios are kept secret. The reports are only disclosed to actual investors of CDOs, who are bound by nondisclosure agreements. (The reports are still kept secret in order to protect the banks and rating agencies from lawsuits.) Cassano was blowing smoke in everyone’s face; he has no idea whether these deals have “pierced the attachment levels,” i.e. crossed the loss threshold when a credit default swap provider must pay out. Again, any half-wit in finance understands the idea of discounting future expected losses to present value.
3. Cassano’s Claim: His books were fully hedged.
The Truth: They were never hedged.
Commissioner Brooksley Born: With respect to your portfolio as a whole, did you hedge any parts of that portfolio?
Born: Which parts?
Cassano: Much of that…But we ran — you know, nothing is 100 percent hedged, but the books were generally considered fully hedged.
Born: Well, let’s look at your credit derivatives portfolio. I think there was something like more than $560 billion in notional amount of credit derivatives in your portfolio in 2007. Were you actually hedging in the conventional sense or were you relying on tranching and the level at which you were insuring? I want you to answer as to whether you were hedging the way you were hedging your interest rates by taking offsetting positions.
Cassano: Perhaps the best way to delineate this is that the super senior credit derivative book, which is the book you’re — the super senior credit derivative book globally, which is the book you’re referencing, had $560 billion. We were using it basically in actuarial basis in order to secure that business. So it wasn’t — it’s not hedged in the conventional sense that you’re talking about buying and selling interest rate risk.
Reality Check: Cassano went Orwellian, labeling his unhedged portfolio as “hedged” the in the same manner that East Germany called itself the German Democratic Republic. Every hedge has two offsetting positions. A single position is always unhedged, period. The distinction between a hedged position and an unhedged position that you deem to be low-risk is as big as the Grand Canyon. A hedge protects you against a market shock, when the markets freeze or act in an unexpected manner. A “low-risk” unhedged position has no such protection. Of course, the “actuarial basis” that Cassano relied upon was also bogus.
4. Cassano’s Claim: The risk exposure on the credit default swaps was managed on an actuarial basis.
The Truth: They took $78 billion worth of unhedged exposure based on the CDOs’ triple-A ratings.
The “actuarial basis” by which these CDOs were evaluated was AIG’s secret financial model developed by Professor Gary Gorton of Yale. It was one of those “Monkey-See-Monkey-Do” models that regurgitated credit ratings but tested nothing. The truth was revealed by Andrew Forster, the former CFO of AIGFP, in testimony given on July 1, 2010. The questioner was Commissioner Peter Wallison:
Wallison: So the Gorton model now evaluated the risk of loss on super senior portions of these CDOs. Did the model evaluate the assets or the composition of the assets in the CDOs?
Wallison: So it just — let me go on a little bit further then and ask — so in your testimony you said that in the summer of 2005 you began thinking more about the multisector CDOs, and you began to question whether the modeling that was needed — the additional analysis of deals — was sufficient, or were they sufficiently taking account of interest-only loans? I think that’s how you phrased it in your testimony.
Were you then beginning to ask whether the model was actually looking at the underlying loans and how it was functioning at that point?
Forster: I think — just to take a step back, if I may — the — through any business that we did, it always made sense to take a step back at different times and question the assumptions that we were using in any of it. And I think that’s — that’s what we did in July 2005. Some of the questions that I posed at that time, we probably knew the answers to. Others of it was just reinforcing the assumptions that we were making.
At the time, what we wanted to do was — the model is obviously only as good as the inputs that you put into it — we wanted to make sure that the underlying loans, underlying reference obligations, we were still comfortable with those and we still felt they — you know, the ratings and things like that reflected the risk that was inherent in it.
Wallison: Let me see if I understand correctly. The model did look at the underlying loans, the kinds of loans that were being made. And when you were talking about interest-only loans, for example, those were taken account of in some way in the model, so that if the model was made up of 95 percent interest-only loans, the model would have reflected the risk associated with that? Is that correct?
Forster: It’s not quite correct, I think —
Wallison: Good. Please correct me.
Forster: Sorry. The underlying ratings of the obligations — if you had the subprime obligation — if it was all interest-only or heavily concentrated in certain areas, then the rating of that obligation would reflect that. So if it was all interest-only, the rating agencies would see that as more risky. It would likely then get a lower rating. The model would just take the rating of the instrument.
Wallison: Oh, so the model relied on the rating agencies?
Forster: Yes, the model — I mean, to a large extent. We made additional changes to it and we stressed the rating agency’s assumptions and we checked that we we were comfortable with the rating agency ratings. But the model basically uses the ratings of the underlying data.
To clarify further, if you let the bogus triple-A ratings define the range of possible outcomes, a “stressed” scenario is meaningless.
5. Cassano’s Claim: His people did not rely on the rating agencies to evaluate the risk.
The Truth: They sure did.
Cassano: We did a fundamental analysis of the transactions. My team reviewed the underlying portfolios and the underlying assets within the portfolios directly. So we were not reliant on the rating agencies to tell us what was good or bad in these portfolios.
Reality Check: See 4 above. This is how a liar defends his lies with more lies.
6. Cassano’s Claim: He arranged the CDOs to benefit from structural seniority.
The truth: The opposite is true. Cassano insured the senior tranches of CDOs compiled of deeply subordinated claims of risky subprime mortgage portfolios.
Cassano: I think what you need to look at within these transactions is the underwriting standards that we committed to, to do these transactions. I’ve heard this phrase that it’s a one-sided bet. But when you think about the protections that we built into the contracts through the subordination levels, through the structural supports that we built into the contracts and then through this very, very strict underwriting standards we performed, it — this was extremely remote risk business.
Reality Check: Any residential mortgage asset that was not initially rated triple-A is deeply subordinated, at the bottom 20% of the capital structure. The deeply subordinated tranches of subprime mortgages were packaged into CDOs, with senior tranches that were rated triple-A and insured by AIGFP. Structural seniority in a CDO only indicates that you’re better off than the suckers beneath you. It’s like having the top bunk in steerage on the Titanic.
7. Cassano’s Claim: The New York Fed handed over $40 billion to the CDO counterparties.
The Truth: He inflated the number by 50% to give the false impression that most of the cash had been paid out at the initiative of the New York Fed.
Cassano: For the credit default portfolio the federal government paid $40 billion. But one of the things I wonder about when I look at that is I’ve never understood why the $40 billion was accelerated to the counterparts. Now, I haven’t been involved in that and I’m only looking at it from afar. But when I think about the contractual defenses and the contractual rights we had in the contracts, it has caused me to scratch my head and ask why was it that $40 billion was accelerated to the counterparts.
Reality Check: A lying liar lied about the dollar amount paid by the New York Fed, and about the criteria that drove negotiations. The CDO counterparties received $35 billion, paid out by AIG before it allowed the New York Fed to take the lead in negotiations, on Thursday November 6, 2008, well past the point when there was time to negotiate anything. There were no clear-cut contractual defenses because there were no clear standards for calculating collateral. The rating agencies told AIG, Treasury and the Fed that AIG’s failure to remove the contingent cash calls from the CDOs by Monday November 10, 2008 would cause them to further downgrade the company, and thereby precipitate a bigger liquidity crisis from collateral calls imposed by AIG’s ratings triggers. Cassano acts like Heinrich Himmler critiquing the Marshall Plan.