Abacus 2007-AC1 — The Goldman Sachs CDO Fraud

In late 2006, John Paulson was convinced: the US subprime mortgage market would collapse. The problem was how. Short-selling individual RMBS securities was market-inefficient and costly.

Paulson's team approached Goldman Sachs with a proposal: build us a portfolio to short. Specifically — create a synthetic CDO referencing 90 subprime RMBS securities most likely to fail, and we'll take a short position via CDS on that portfolio.

Goldman turned this idea into a business. It hired ACA Management as the CDO manager to handle portfolio selection. But it failed to disclose a critical fact to ACA: the entity that had the most influence on the actual portfolio composition was Paulson, who planned to take a short position on this deal.

In April 2007, Abacus 2007-AC1 was issued.

Abacus 2007-AC1 was a 'synthetic' CDO. Rather than holding actual RMBS securities, it created synthetic exposure to 90 subprime RMBS through credit default swaps (CDS).

Structure: Goldman established an SPV → the SPV acted as 'protection seller' under CDS contracts → Paulson was the 'protection buyer' → if RMBS defaulted, Paulson received payment from the SPV.

Investors (IKB Deutsche Industriebank, ACA Financial Guaranty, etc.) invested in this SPV → effectively taking a long position on those RMBS securities. If RMBS held up, investors received spread income; if RMBS defaulted, investors lost principal.

Paulson was deeply involved in portfolio selection, handpicking the most likely-to-fail RMBS. Simultaneously, he asked Goldman not to disclose this to other investors. Goldman complied. ACA believed it was acting as an 'independent manager taking a long position.'

Nine months after the April 2007 issuance, by January 2008, more than 99% of the 90 subprime RMBS portfolio underlying Abacus 2007-AC1 had defaulted or been severely downgraded.

IKB lost approximately $150M; ACA Financial Guaranty lost over $900M. Combined investor losses exceeded $1B. Meanwhile, Paulson & Co. earned approximately $1B from this deal alone. Paulson's total 2007 profits reached approximately $3.7B.

Goldman also held some long positions and lost around $90M on this deal. However, Goldman maintained net short positions in subprime overall and was profitable during this period.

The core problem: investors had no idea that the entity selecting the portfolio planned to short it. This was the crux of the SEC's case.

In April 2010, the SEC charged Goldman Sachs with securities fraud. The core allegation: Goldman had misled investors by omitting from Abacus marketing materials the fact that Paulson had been involved in portfolio selection and intended to take a short position.

Goldman's position: "There was no legal obligation to disclose that Paulson would take a short position. ACA independently reviewed and approved the portfolio."

In July 2010, Goldman settled for $550M — at the time, the largest securities fraud settlement in SEC history. Goldman admitted that marketing materials "contained incomplete information" but denied intentional fraud.

Paulson was not charged. As a counterparty, he owed no legal duty to the investors.

This case underscored how critical 'conflict of interest disclosure' is in structured finance, and became a centerpiece of post-crisis financial regulatory reform discussions.

Abacus 2007-AC1 simultaneously demonstrates three fundamental vulnerabilities of structured finance.

First, information asymmetry: complexity creates information gaps. If investors had known Paulson designed the portfolio, would they have invested? The more complex the structure, the greater the information asymmetry risk.

Second, ratings limitations: even AAA-rated tranches suffered total loss in nine months. Rating models failed to capture the high correlation between subprime mortgages. "Ratings evaluate structure, not intent."

Third, incentive misalignment failure: Goldman earned fees from selling the deal. Paulson profited from designing it. Only investors bore the risk without full information. Conflicts of interest were embedded throughout the structure.

What Michael Burry, Paulson, and Steve Eisman in The Big Short recognized was not simply 'housing prices will fall.' They identified that 'the structure itself was corrupt.'