Wednesday, July 4, 2012

Abacus 2007-ACI for Dummies (Part I)

(Editor's Note: The original Parts I & II of this series can be found on the author's instablog here and here.)

Was ACA needed?

I strongly believe it was not needed.

To create a synthetic CDO we need an underwriter (in this case Goldman Sachs (GS)) and investors. Sometimes investors require that the collateral is selected by an independent party (the ACA in our case). Generally, if the underlying assets are selected by an independent agency, there is an alignment of interests between the underwriter, the collateral and the investors. In the ACA case, the SEC argues that because Paulson selected (or helped select) the collateral there was no such alignment, and in particular, the interests of the underwriter clashed with those of the investors.

Technically, Goldman Sachs could have acted as the Portfolio Selection Agent. However, if it was done without a Portfolio Selection Agent, IKB would not have invested in this synthetic CDO. The SEC claims (.pdf) that IKB told Goldman Sachs and Tourre that

it was no longer comfortable investing in the liabilities of CDOs that did not utilize a collateral manager. Tourre and Goldman Sachs knew that ACA was a collateral manager likely to be acceptable to IKB.

The Collateral

The SEC argues that Paulson heavily influenced ACA’s selection. Regardless of the illegality of Goldman’s actions, we know that on the 26th of February 2007, Paulson and ACA came to an agreement on a reference portfolio of 90 RMBS (paragraph 35 – SEC) with an average life of 3.9 to 4.9 years (according to the flipbook).

There is not much information regarding the 90 RMBSs. The only coherent list is the one provided in the flipbook and is very hard to understand. The RMBS have rather arcane names like ABFC 2008-OPT1 M8 for example. I have tried searching for further information but there is nothing available.

Who are the counterparties?

The parties involved in this deal are Goldman Sachs, ACA and IKB. Goldman Sachs structured and marketed the synthetic CDO. ACA acted as the Portfolio Selection Agent (that is they chose the CDS which were included in the reference portfolio).

IKB is a German commercial bank which bought $50 million worth of Class A-I notes at face value and $100m worth of Class A-2 notes also at face value. The Class A-I Notes paid a variable interest rate equal to LIBOR plus 85 basis points and were rated AAA. The Class A-2 Notes paid a variable interest rate equal to LIBOR plus 110 basis points and were rated Aaa by Moody's (MCO) and AAA by S&P.

It is clear from the transaction that Paulson and ABN AMRO did not take part in the ABACUS 2007-AC1 synthetic CDO. ABN AMRO lost a significant amount of money because it entered into CDS contracts with Goldman Sachs and ACA referencing the super senior tranche of ABACUS 2007-AC1. But clearly, this is a separate trade that has nothing to do with ABACUS.

It is important to be clear on the role of ABN.

ABN intermediated a $909m CDS referencing the portfolio between GS and ACA. ABN assumed the credit risk that ACA might not be able to pay if its obligations under the CDS came due.

Finally, in late 2007, RBS unwound ABN’s super senior position and paid GS $840,909,090.

It can be argued whether or not Paulson took part in this transaction. In my opinion as he did not buy any security, he should not be included. On the other hand, the SEC accuses him of influencing the selection of the collateral and misleading ACA into believing that he was going to invest in the CDO. Furthermore, he paid all the structuring fees to Goldman Sachs ($15m). This is all very strange.

Why would you pay for something that you will not invest in? I’m not the only one who has this question. See Felix Salmon’s blog entry: “Goldman’s ABACUS fee”. Was $15m a normal fee? It seems high to him. Since the total cash invested was about $200m, this implies that Goldman Sachs' fee is about 7.5% of the cash size of the deal.

The final numbers

The SEC argues that investors in this transaction lost approximately over $1bn. Paulson's opposing CDS positions yielded a profit of also $1bn. We also know that GS lost $90m (according to them) and earned $15m in fees (according to the SEC).

According to GS's answer, we see that ACA was the largest investor in the transaction ($951m), followed by IKB which invested $150m.

The main question here is whether or not ACA charged a fee. It seems hard to believe that they took such a large role without receiving any fees in return.

Let us see if the numbers add up:

Total Cash Invested: $192m (nearly $200m).

Super Senior Tranche: $909m (ACA)

Investors: $909 (super senior ACA) + $42 (ACA) + $150 (IKB) = $1,101 = $1.1bn.

Synthetic CDO Notional Total: $1.1bn

Fees:

- Paulson paid GS $15m in fees.

- ACA ?

Losses: Everything. $1.1bn (which ended up being Paulson’s profit).

The winners: Goldman Sachs ($15m) and Paulson ($1.1bn)

The money ($90m) that Goldman Sachs says it lost in this transaction is irrelevant. The loss comes from a separate trade, not from the ABACUS CDO itself.

If ABN ended up paying $840,909,090, it implies that ACA paid approximately $68m before it defaulted.

Finally, the structure of the synthetic CDO might look like: (adapted from Sandrew on Finance)

Tranche

Investor

Amount

Rating

Attachment Point

Detachment Point

Super Senior Tranche

ACA

$909m

45%

100%

Class A1

IKB ($50m)

$50m

AAA

38.75%

45%

Class A2

IKB ($100m) and ACA (42m)

$142

AAA

21%

38.75%

TOTAL

$1.1bn

Disclosure: No positions

No comments:

Post a Comment