Goldman CDO suit throws focus on collateral manager conflicts

Goldman Sachs fraud allegations show portfolio managers credit selection interests are often not aligned with benefiting CDO note-holders, say lawyers.

A recent high-profile legal action launched by the US Securities and Exchange Commission against Goldman Sachs has thrown fresh scrutiny on apparent conflicts of interest between synthetic collateralised debt obligation (CDO) collateral managers and note-holders on whose behalf the portfolios are supposedly being managed.

A civil lawsuit filed by the US SEC at the US District Court for the Southern District of New York on April 16 alleges Goldman Sachs violated securities law by failing to disclose to investors the involvement of New York hedge fund Paulson & Co in selecting the credits underlying a synthetic CDO called Abacus 2007-AC1.

The Goldman case hinges on the contention that the collateral manager responsible for running the Abacus deal – New York-based ACA Management – would not have agreed to lead the transaction had it known the role Paulson played in adversely selecting residential subprime mortgage-backed securities (RMBS) it wished to short.

Investors that agreed to buy notes issued by Abacus did so in the understanding ACA had conducted due diligence on the underlying mortgage securities to gauge their probability of default.

"Many investors looked at the quality of the collateral manager as the main determinant in whether they chose to invest in structured credit instruments. They were aware they weren't able to comprehend all the underlying instruments, so they selected a manager they had confidence in to differentiate good assets from bad – although there were disclaimers in the documentation a mile long that warn caveat emptor," says a London-based hedge fund manager.

In the case, Paulson selected a portfolio of 123 RMBSs it wanted to short and apparently approached Goldman Sachs to help it buy credit default swap (CDS) those names via a CDO. ACA conducted its own analysis of the securities, and included 55 of the original 123 names in the final portfolio.

Lawyers say the case has put a focus more generally on the role of CDO managers, and whether they discharge their duty in the interest of CDO investors or CDO arrangers.

"The key question is on which party's behalf and for which party's benefit is a CDO managed? Before 2007, CDO documentation did not mention for whom the structure was managed. In later vintage CDOs, they expressly state the portfolio is managed for the benefit of the CDO issuer, the special-purpose vehicle (SPV) itself, not the CDO investor. Today, those statements are standard in CDO documentation," explains Dario Loiacono, senior partner at law firm Loiacono e Associati in Milan.

Offering documents for managed CDOs – in which a collateral manager is permitted to make substitutions to the names underlying the structure once the deal closes – obtained by Risk confirm this analysis.

"The portfolio manager has been appointed by the issuer under the portfolio management agreement to manage, on behalf of the issuer and for the sole benefit of the issuer, the reference portfolio with the objective of approving the initial selection of, and thereafter maintaining a mixture of, reference entities in the reference portfolio, in order to minimise the probability that the relevant reference portfolio will contain reference entities that suffer credit events," reads one of the documents.

While credit substitutions that benefit the issuer would seem to also be in the interests of note-holders, it depends on the part of the capital structure both parties are invested in.

Investors in the equity tranche generally prefer high correlation in the underlying portfolio, as that would imply either no defaults or many defaults – no defaults would mean the equity tranche would continue to perform. In contrast, an equity investor would be disadvantaged by low correlation, as that would imply a greater chance of idiosyncratic events that would wipe out the investment.

The opposite is true for investors in the super-senior tranche. High correlation means there is greater probability of many defaults, which could eat into the senior tranches. Low correlation implies a smattering of individual credit events, which would be unlikely to wipe out all the layers of subordination.

"I've seen cases where a CDO was managed differently on behalf of different investors in different tranches. So, for the benefit of one super-senior investor, substitutions would be made in the knowledge those changes would be detrimental to the interests of the equity investor. I've seen other cases where there is nothing in the documentation about which party a transaction was managed for, but the salesperson assured the investor the CDO would be managed in their interests. Unsurprisingly, that one became a misrepresentation case," says Loiacono.

Arrangers would attempt to align the interests of portfolio managers and other parties in a variety of ways. The manager would collect both a senior component fee based on the income the CDO receives and an incentive component based on whether there was cash left in the structure to distribute at the maturity of the trade.

Collateral managers would often also take long positions in the equity tranche of a CDO, partly to illustrate its own faith in the structure to note-holders.

In some cases, collateral managers simultaneously bought CDS protection on the underlying portfolio, making them largely agnostic on the performance of the CDO.

"It was not unusual for managers to buy the equity tranche of a CDO and then short the entire reference portfolio. Depending on the pricing, if there were no losses on the equity notes and the carry to go short the reference portfolio was less than what it was receiving on the equity position, then the trade came up trumps. At the same time, if the manager lost all its equity position but spreads blew out enough on the short protection, it would still make money on its CDSs. It would make money on either extreme and only lose if there were a few defaults – effectively it was a straddle trade," says a New York-based credit structurer.

Despite questions over incentives, the involvement of independent third-party agents was preferable to non-managed trades. In these transactions, the arranging dealer would manage the portfolio selection for its own interests.

An offering document from one such deal that Risk has seen explicitly states: "The portfolio adjustment agent is only obliged to consider the interests of the credit default swap counterparty and is not obliged to consider the interests of the issuer, the trustee or the note-holders."

Despite such conflicts being spelt out in black and white to alert note-holders the deal would not necessarily be managed in their interests, investors still entered the transaction, most having forgone legal counsel beforehand.

"A portfolio adjustment agent is clearly not an independent third-party collateral manager that has a responsibility to maximise returns for noteholders – it is probably an entity like a prop desk of a bank. I've seen this kind of language several times before. Most incredibly, there are deals that state in large script on the first page of the prospectus that decisions made by the portfolio manager or agent may be adverse to the interests of noteholders, and yet investors still buy these products," observes the credit structurer.