There are various ways to describe the synthetic collateralised debt obligation that Goldman Sachs constructed for John Paulson, the hedge fund manager who bet on the collapse of the mortgage bubble.

Goldman itself terms it “nothing unusual or remarkable”. The US Securities and Exchange Commission describes the Abacus deal that closed in April 2007 as securities fraud. I call it short-term greedy.

Goldman rose to its dominant position on Wall Street through the dictum of Gus Levy, its former senior partner, that it should be “long-term greedy”. He meant that it should forego quick gains for enduring profits.

The bank's expression of that principle on its website is: “Whether a mid-size employer in Kansas, a larger school district in California, a pension fund for skilled workers, or a start-up technology firm, our clients' interests come first.” So what about a Düsseldorf bank?

In late 2006, after an internal debate on its mortgage desk, Goldman made a far-sighted decision to protect itself from what it had come to fear would be a severe downturn in the housing market. It pulled back from taking risk and hedged its mortgage book, saving itself from the deluge.

At exactly the same time, it was approached by Paulson & Co to structure a CDO that the hedge fund could make money by shorting. Paulson offered it a $15m fee to assemble the deal and find an investor that would take the risk from which Goldman was simultaneously fleeing.

Fabrice Tourre, Goldman's

front-man on the deal, went to IKB, a bank that had invested in several of its Abacus CDOs and was a valuable Goldman client. He offered it a package which it took without looking closely enough at the dodgy mortgage swaps inside.

When the smoke cleared, Paulson made a profit of about $1bn while IKB lost $150m and had to be bailed out that August. ACA Capital, which selected the securities, lost $900m and later failed (with ABN-Amro assuming liability). Goldman, which tried to shed its position but could not hedge it exactly, lost $100m.

The SEC will find it difficult to make a charge of fraud stick against Goldman and Mr Toure. IKB was a “qualified” professional investor that was not supposed to need its hand held by an investment bank (although there always seems to be a German bank in such cases, eager to blow a hole in its balance sheet).

But judged by Levy's dictum, and the principles of John Whitehead, another Goldman senior partner (“We are dedicated to complying fully with the letter and spirit of the laws, rules and ethical principles that govern us. Our continued success depends on unswerving adherence to this standard”), Abacus was a tawdry episode.

None of the deal documents made clear that Paulson had called the tune and Goldman had initially acted on its behalf. Goldman has been reduced to brandishing the letter of the law and insisting that it adhered to “market practice”, which is not the same as ethical principle.

Goldman gave 8m documents to the SEC, but did not include the overarching philosophy that brought it to Abacus, the strategy laid out by Lloyd Blankfein, its chairman and chief executive, in 2005.

As Charles Ellis's The Partnership records, Mr Blankfein feared that if Goldman stuck to its traditional separation of agency and principal businesses – keeping its client advisory and asset management work isolated from risk-taking with its own capital – it would be overtaken by commercial banks.

“Its complex variety of many businesses was sure to have lots of conflicts,” Ellis writes. “Goldman Sachs, Blankfein said, should embrace the challenge of those conflicts.” It definitely embraced them with Abacus, keeping IKB in the dark both about Paulson and its own view of mortgage CDOs.