I have stated in an earlier article
my belief that SEC has a very strong case. I also believe that the Goldman Sachs (GS
) defense I’ve read thus far is inadequate.
So what is wrong with the Goldman defense? In my opinion, Goldman is putting its own spin on the subject matter, rather than addressing the SEC complaint directly. In fact, they seem to be regurgitating the same boilerplate arguments I outlined in the earlier article without really addressing the specifics.
Let’s break down the defense Goldman camp has used to get a better view of what we’re talking about:
The “But This Wasn’t Material Information” Defense
This has been an important focus of discussion in legal blogs and opinions. I seem to be one of the rare souls who believe with certainty that this information is absolutely “material.” Misleading investors in a structured security underwriting on top of excluding pertinent information in an offering memorandum in my opinion should pass the test of “materiality.”
A collateralized debt obligation is a structure that repackages loans into a new instrument. These portfolios could be made up of pools of corporate debt, auto loans, credit card debt, or subprime mortgages.
The synthetic Abacus CDO was composed of credit default swaps written on subprime debt. These credit default swaps (CDS) were insurance bought by Paulson on the housing market from Goldman. For this short position, Paulson paid premiums just like regular insurance.
Paulson bought insurance on what he thought was most likely to fail. In other words, these CDS instruments suffered from what is known as adverse selection problem in the insurance industry. People tend to buy insurance in bad products or what is more likely to fail, but they don’t tend to buy protection on good products or successful investments. It is allowed and perfectly legal.
So it was this insurance guaranteed by Goldman, which was repackaged into the Abacus CDO. This is where the Abacus CDO deal-making process entered into the picture. Goldman found long investors for the deal to transfer its short CDS liability to Paulson, without telling them that Paulson had the majority of say in what products would go into the CDO structure.
Instead, the CDO was represented and marketed to investors (IKB, and ABN Amro who assumed Goldman’s credit risk with ACA) as having been designed and selected by ACA Management, an independent firm with a long interest in the underlying securities.
SEC alleges that Goldman not only hid Paulson’s role from all long parties via making it appear like a third long party (ACA) picked the portfolio, which is a material misrepresentation on its own, but it also made one investor (ACA) believe that Paulson was interested in the long side of the deal.
Some legal experts say the case may very well come down to the definition of “material information.”
First of all, a specific information does not have to act as a threshold to make or break a deal for it to be considered material.
Such information could also make a difference in the pricing of a deal, or the market price of a proposed deal or transaction, if we were talking about a publicly traded firm acquiring another firm.
This information could have caused investors to require higher yields and better negotiating terms from the other side, i.e., Paulson, to be compensated for the extra risk they were carrying, if they had known that someone with a short interest was influential in choosing the securities. But no. Instead, they believed an independent collateral manager with a long interest had performed that job.
This information would also influence the grading of the bundled product besides yields and structure. A Moody’s (MCO
) official has testified
in a US Senate panel that “he would have liked to have known that billionaire hedge fund manager John Paulson was shorting the Abacus product when Moody's was rating it”.
What has been done in this deal is not equivalent to failing to inform that Paulson the hedge fund is on the other side of an investment. If this had been a simple trade in the capital markets, confidentiality of Paulson’s position would have been rightfully protected.
However, this is not a simple trade or market transaction, it’s an underwriting with an offering memorandum that also gets graded depending on collateral and structure. The material information at hand is that this portfolio has been misrepresented as having been designed and selected by an independent third party with a long interest, ACA Management. The email trail presented by SEC shows that Goldman staff has gone through some pains to conceal this information, and this action is not limited to Fabrice Tourre.
Gregory Palm, the general council of Goldman Sachs, insisted in the quarterly conference call that ACA Management was the sole party responsible for selecting the securities, despite the SEC claim that ACA chose securities from the pool of mortgage securities recommended by Paulson, whereby ACA accepted 55 of the 123 initial securities proposed by Paulson (Complaint No: 30).
The rest of the securities (35 in total) were decided back and forth between ACA and Paulson. On top of that, Paulson exercised the freedom to delete “safer” securities suggested by ACA from the portfolio. (Complaint No: 34) Thus Paulson had a very significant say, if not full control, in picking a total of 90 names for the Abacus portfolio.
According to the transcripts of the conference call available at Seeking Alpha, when asked about exactly what Paulson’s involvement in the portfolio selection process was, Palm says this
In this market there has to be a long and a short. That is perfectly clear. Other point I would really emphasize is in order to have a transaction in this market you have to have some reference portfolio of securities which is satisfactory to both the longs who are looking at the portfolio; they are not really looking at really anything else and the short who are looking at the same portfolio and deciding that. As you know, whether the shorts are us or anyone else.
If I understand correctly, he is saying (or rather going through great pains to avoid saying) that it is natural for the long and the short side to come together and pick a portfolio that is mutually acceptable to both the long and the short sides of the deal.
That’s great. Sounds quite acceptable to me so far. But why was the deal represented to the outside investors as if ACA Management, a company who had long interest in the deal, had sole responsibility in picking all of the names in the portfolio? SEC alleges that it was because German bank IKB specifically required an independent collateral manager as an investor. All official marketing materials represented ACA as the only party (with long interest) who selected securities for the CDO structure.
Goldman Sachs bought insurance from ABN Amro to cover for ACA’s credit risk, thereby exposing this firm to a long position in the CDO in case ACA failed to make through on its payments. Would ABN Amro have been equally willing to insure this transaction at the spread it agreed to if it had known that Paulson with a short interest had a significant say in picking the securities in this deal? It was ABN Amro (via RBS) who eventually recouped the losses of ACA Management in this deal. (Complaint No: 66)
Remember, Paulson picked 55 out of the 90 names, and it vetoed the “sound” RMBS securities suggested by ACA from the rest of the 35 that eventually made it into the portfolio.
Later on in the same conference call Palm offers the following
In very simple terms the portfolio here was not selected by John Paulson. The portfolio here was selected by ACA. ACA had the responsibility to do that. They were paid to do it.
Wait, didn’t he just say minutes before something about “the longs who are looking at the portfolio and the short who are looking at the same portfolio and deciding that”?
Must be difficult to stick to your initial story about ACA’s portfolio selection responsibilities AND try to sound consistent when asked about Paulson’s involvement at the same time.
If anything, it appears as if the general council is failing a simple cross-examination regarding this matter.
The charges have to do with failure to disclose relevant information, and “recklessly, negligently and knowingly misrepresenting” the offering, which fall within the scope of the Section 17(a) of the Securities Act of 1933 and Section 10(b) of Securities Exchange Act of 1934 and Rule 10b-5. Private placements are not exempt from these laws.
Also, the fact that these investors were “sophisticated” does not relieve Goldman of its duty to comply with the laws stated above. Finally, the fact that IKB, ABN, and Moody’s failed to do proper due diligence on the selected securities, while noteworthy, does not give Goldman the license to misrepresent the deal making process.
The “But We Lost Money On The Deal” Defense
Goldman Sachs has affirmed that because it has lost money on this deal, the “intent” to defraud is proven to be false. My first reaction to this defense was the irrelevance of this fact (or claim).
However, Goldman is treading a dangerous territory by taking this route. If the firm maintains innocence partly because of its long position in this specific deal, then it will have to explain profits made via any offsetting short positions in other proprietary deals or trades besides this one, as well as insurance “hedges” bought in the form of CDS contracts. On top of that, despite claiming to have “skin in the game” via this position, Goldman has tried to unload
its remaining stake in the Abacus investment “almost from the start”.
In fact, there are very good accounts (here
) by William Cohen of the Daily Beast and Steven Davidoff and Peter Henning of The New York Times about how Goldman’s relationship with its short-betting client Paulson may have helped shape the firm’s own opinions regarding the mortgage market. The former also explains in detail the linkage of that outlook to Bear Stearns’ eventual collapse.
Goldman Sachs’ infamous decision to reduce its exposure to the mortgage market after a ten-day losing streak at the end of 2006 was regarded as evidence by some (including yours truly
) as an astute use of risk management by the firm.
Taking a better look at the inside of the firm gives a better view of the fact that Goldman was a significant originator, intermediator, arbiter, marketer, buyer, seller, and cheerleader of this opaque market of mortgage securities, where barriers to entry made it impossible for most people to participate in. Goldman not only had a far better knowledge of the junk it was selling, but it also knew which players were betting for and against which instruments, and thus the timing and size of important bets.
A report released by Permanent Subcommittee on Investigations notes
that a transaction by the name of Timberwolf Ltd. (along with Abacus deal) was completed after
one department shifted its proprietary position in the mortgage market from $6 billion long to $10 billion short by the end of the fateful quarter that ended February 2007.
The subcommittee has released new information that enables us to take a closer look into the workings of the investment house. In an internal email, a supervisor refers to the Timberwolf CDO as a “sh**y deal”. The same supervisor sends an email congratulating sales staff for getting rid of a specific position in this CDO created by Goldman itself.
This is purely conjecture, but Abacus CDO could also be seen in the bigger picture as an effort for Goldman to unload a CDS liability it acquired in earlier dealings with Paulson.
Interestingly, Andrew Butter has provided a short and important timeline
at Seeking Alpha that puts the Abacus deal into a new perspective, into a deal that was designed to unload an earlier CDS exposure Goldman acquired in trading with Paulson in 2005 and 2006.
I don’t know whether Goldman had any naked CDS exposure to Paulson before the Abacus CDO transaction was completed. This is speculative at best, but if Goldman had full exposure, then this transaction would save them $900 million as opposed to costing them around $90 million.
But for that to happen, Paulson would have to allow Goldman to buy back the naked CDS, thus allowing Goldman to get out of its original position. After the CDO deal gets done, Paulson would buy a similar CDS from Goldman, and this CDS would now be appropriately hedged in Goldman’s books via the special purpose vehicle created for the Abacus CDO. Simple, no?
Interestingly, Blankfein himself sends an email
dated Feb 11, 2007, wondering whether “could/should we (they) have cleaned up these books before and are we (they) doing enough right now to sell off cats and dogs in other books throughout the division?”
The head of the correlation desk sends an email on the same day to Fabrice Tourre stating, "the cdo biz is dead we don't have a lot of time left" (SEC Complaint No: 18). If anything, these emails are indicative that Mr. Tourre was not acting alone in his over-extended enthusiasm to get the CDO deal done at all costs. They are also indicative of the firm trying to reap commissions and/or trying to get rid of “cats and dogs”, rather than trying to serve clients’ best interests during this timeframe.
Goldman has certainly profited at the expense of its clients via unloading mortgage securities they helped create. Even if these actions themselves are not against the law, attempting to unload such positions in the inventory (including the remaining Abacus position that eventually cost Goldman $90 million), and email trail released by the SEC and the Senate subcommittee, assist in establishing the intent to misrepresent the Abacus deal to investors.
Disclosure: No positions