The Goldman Sachs scandal: the void of responsibilities
Posted on 21. avr, 2010 by Jean Jacques Ohana in Weekly Focus | 3 comments
In 2006, Paulson asked Goldman Sachs to structure a Collateralized Debt Obligation (named ABACUS), made of a pool of Residential Mortgage Backed Securities in order to place a bet on the burst on the real estate bubble. A Credit Default Swap (CDS) on the CDO (or “synthetic CDO”) was then structured, so that Paulson could benefit from the default of the RMBS (being on the long side of the CDS). Goldman Sachs in his defence document describes the synthetic CDO as having « characteristics much like that of a futures contract, requiring two counterparties to take different views on the forward direction of a market or particular financial product, one short and one long”. Paulson has an interest on the short side of the CDO (he will be long the CDS).
Goldman Sachs selected a supposedly independent selection agent (ACA Management), which by miracle, is the same entity as ACA Capital, the major long investor. On top of being a monoline insurer, ACA used to be a CDO Manager before its whole CDO management business was smashed by the credit crisis and lately dismantled and sold off. Even more striking, ACA received fees to select the pool of RMBS included in the reference portfolio of the CDS. There were meetings between Goldman Sachs, ACA and Paulson where the latter may have influenced the portfolio selection without explicitly telling he was on the short side according to the SEC.
Goldman Sachs did not intend to keep a long position on the CDO (short position on the CDS) in its book. Contrary to the classical hedging business of a market maker, this product is too complex to be hedged with standard product. Therefore, Goldman Sachs tried to find a long interest in the CDO (or an interest to go short the CDS in front of Paulson). ACA Capital took 950 Millions of the CDO and the structure was successfully marketed to IKB, a German bank, which would later be saved by the German Government. Goldman Sachs retained a small part of the long side, and stated in his defence that it lost 90 million USD in this specific transaction, failing to reveal how much it won from the different favours possibly paid back by Paulson in return of the fabulous service provided by Goldman Sachs (Paulson earned $1 bn out of this deal). Besides, Goldman Sachs omits to say that the bank shorted many other CDO structures under the supervision of David Viniar, Goldman’s chief financial officer who suggested that it adopt a more bearish posture on the subprime market as of December 2006. It is therefore highly plausible that this “90 million USD loss” was offset in some way by “hedging” operations in other departments of the bank.
Last but not least, ABN Amro guaranteed ACA’s default under a CDS negotiated with Goldman Sachs. Indeed, Goldman Sachs was the global counterpart of all players and hedged against ACA’s credit signature. When writedowns eventually occurred in the reference portfolio, ACA did not meet its obligation and ABN made payments to Goldman Sachs.
Goldman Sachs acted simultaneously as underwriter of the product and broker since it wrote marketing materials to find long investors opposite Paulson. In the marketing document of ABACUS 2007-AC1, they explicitly recognize the possibility of “conflict of interests” (Goldman Sach ABACUS documentation to investors), which looks like an acknowledgement of their ambiguous role in the transaction.
The deal structure is represented below:
In our view, the number of parties involved into this highly complex transaction illustrates the institutional void surrounding Investment Banking business. This void was favourable to the emergence of extremely complex transactions where responsibilities are disseminated in every part of the chain.
A certain number of questions are raised by this scandal.
The first set of questions concern the synthetic CDO themselves. How come virtual CDS betting on someone else’s portfolio are authorized? Economist Paul Krugman wrote in April 2010 that the creation of synthetic CDO should not be allowed: « What we can say is that the final draft of financial reform…should block the creation of ’synthetic CDO’s,’ cocktails of credit default swaps that let investors take big bets on assets without actually owning them.”
The second set of questions concern the banking practices. Why are investment banks involved into a mix of advisory, structuring, flows trading and proprietary trading roles? Ironically, Goldman Sachs is accused by the SEC of breaching laws established in 1933 and 1934. The regulatory framework needs an urgent update to catch up with the complexification of financial transactions…
3 Responses to “The Goldman Sachs scandal: the void of responsibilities”
About us
Visit our corporate website :
Subscribe to our blog posts :
Why does it matter if Paulson selected the most toxic tranches to bet against? – of course he did, as he wanted to make money of it. ACA could have easily decided not to enter into a deal with such toxic tranches – who forced them?
They did not care because they were going to sell it to a third party and just collect the fees, just like Goldman did (and that is fine with me). IKB could and should have reviewed what they were buying and nobody forced them to buy it.
None of the players here were innocent people taken advantage by a scam artist – they were sophisticated institutional investors who should have known better (Goldman and Paulson clearly did). Innocent people who invested with IKB may have lost money, but that is because IKB failed in its fiduciary duty to them (not to invest in stuff they do not understand).
Paulson was looking to make money, just like a professional gambler may bet more on a soccer match if a certain player is injured that if he was not. Had ACA insisted on putting the best tranches on the CDO Paulson would have likely walked away from the deal, it was not just a good bet for him.
Why does it matter if Paulson selected the most toxic tranches to bet against? – of course he did, as he wanted to make money of it. ACA could have easily decided not to enter into a deal with such toxic tranches – who forced them?
There’s no crying in baseball. But that does not mean that there’s no cheating or that cheating should not be punished for the benefit of the game. The SEC represents the markets, not the victims of any particular scam. The foolishness of the victim says nothing about the thief.
But I think Mario ignores the subjective nature of what ACA and IKB do for a living. ACA was acting as an insurer of the portfolio. Yes, it can inspect the property, but it can also legitimately inquire into matters of adverse selection. Does the insured have an insurable interest? Is guy helping to select the portfolio (WHY was JP invited to that table?) long or short the portfolio? Are we working together to select the best reference portfolio, or are we negotiating one? These are matters that insurance underwriters routinely consider to bolster their confidence in their own assessment. Non-insurance types idealize the process as based simply on the merits of the risk, but insurers know better than to rely entirely on their judgment of the implications of who is buying the insurance and why.
As regards IKB, we are talking about a foreign investor who came to a gilt-edged investment bank, demanded that an independent agent select the portfolio of US RMBS, and further demanded that their tranches be rated AAA. If that’s not enough, who needs investment banks, selection agents, and ratings agencies? And what foreign entity will want to hold dollars if the US financial services industry cannot be trusted to deliver good paper when asked – and paid! – to do so?
Dear Mario and Lawrence,
The points you raise are quite interesting.
I think that conflicts of interest and loss of confidence could be avoided if we prevented market makers from marketing their unhedgeable risks to other investors. Here, this would have prevented Goldman from dealing with Paulson in the first place as the risk they took (long position in the CDO) was not traded in a liquid market. Of course, this radical proposal would be an obstacle to the market making activity in investment banks and insurance companies as it would prevent them from transferring such risks as dividend or correlation risks to investors. But the benefits could be important for society as a whole: more trust, less conflicts of interest, less moral hazard (as
the new risks created will have to be borne by the entity that originates and (hopefully) understands them).
This proposal has to be included in a broader set of reforms whose goal would be to better circumscribe and separate banking/insurance activities (market making, advisory and prop trading).
Yours sincerely
Steve Ohana
Professor of Finance at ESCP Europe