Can you stand reading one more thing about the Goldman Sachs (GS) scandal? Try to bear with me. I'll tell you something about it you haven't heard before.

As you know, as you've heard a million times over the last week, the Securities and Exchange Commission has sued Goldman Sachs for fraud, in connection with a complex subprime mortgage security with the decidedly unsexy name of Abacus 2007-AC1. If you want to read the SEC's formal complaint click here. It'll only take you 10 minutes to read, and you don't have to be a lawyer to understand it. I recommend taking a look, rather than letting the media tell you what to think.

That said, speaking as a card-carrying member of the media, let me proceed to tell you what to think. Or more specifically, what not to think.

Here's what to think. If the SEC is right, then Goldman failed to disclose all the relevant facts about Abacus 2007-AC1 to investors. Under the securities laws, if those facts are found to be material to making an informed decision about investing, then failing to disclose them is fraud.

Here's what not to think. A lot of commentators are saying that Goldman designed Abacus 2007-AC1 as an investment time-bomb, deliberately created to be as worthless as possible. Why would Goldman do that? Because there were not only people who wanted to buy subprime mortgages like Abacus 2007-AC1 -- thinking they'd be great performers -- there were also people who wanted to sell them short -- thinking they'd be dogs. Supposedly Goldman was catering to the shorts, not the buyers.

First, let's be clear on one critical reality. No matter what the media mouthpieces are saying about it, the fact is that the SEC is not accusing Goldman of this. It's a simple fraud case based on failure to disclose. No one is accusing Goldman of created financial weapons of mass destruction.

Second, let's be clear on what Abacus 2007-AC1 is, and what it isn't. It's a very unusual and complex investment instrument -- and if you just think about it like a plain vanilla stock or bond, you won't understand what Goldman did, or why.

When Goldman or any other investment bank underwrites a new issue of stock or a bond, it has a duty to offer its customers a sound investment. Sure, there are always risks. That's where disclosure comes in. But Goldman's customer is the buyer, so Goldman basically owes that customer something that's worth buying, at least given the risks.

But that's not Goldman's duty when it acts as a broker for trading in stocks and bonds that already exist -- rather than new issues that Goldman is underwriting. With existing securities, Goldman's duty might be to a buyer, or to a seller, or to both at the same time. The buyer might think the stock or the bond is wonderful, and the seller make think it's the worst -- Goldman just stands in between them and makes the trade happen, without passing judgment.

Goldman's role in Abacus 2007-AC1 is more like that, even though it was a new issue that Goldman was underwriting. The reason why is that Abacus 2007-AC1 wasn't a stock, and it wasn't a bond. It was what is called a synthetic collateralized debt obligation, or "synthetic CDO." It's a derivative security, simply a bet between its buyer and its seller. It wouldn't even exist if there were not someone who wanted to be long and another someone who wanted to be short. So Goldman owes a duty to both buyer and seller symmetrically.

Should Goldman have made sure that Abacus 2007-AC1 was the best damn synthetic CDO in the world, practically sure to earn a great return for its buyer? No -- because that would have been disloyal to the seller. Neither should Goldman have tried to make it a time-bomb. That would have been disloyal to the buyer.

Here's exactly what happened, at least according to the SEC's complaint. A hedge fund called Paulson & Co. believed (correctly) back in 2007 that the subprime mortgage market was going to blow up, so it was looking for ways to be short subprime mortgage backed securities (MBS). Paulson went to Goldman and asked them to create a derivative security that would act like an index fund that tracks a hundred or so subprime MBS. Paulson would take the short side of the deal, and Goldman would find someone else to take the long side.

Now it gets complicated (it wasn't complicated enough already?). If Abacus 2007-AC1 was to be like an index fund, someone would have to pick the MBS that go into the index. So Goldman hired an investment manager who was an expert in this field, called ACA. While ACA was assembling the index, Paulson worked with them on it. According to the SEC, Paulson tried to get the worst subprime MBS into the index, while ACA was trying to get the best.

Here's the catch. According to the SEC, ACA didn't know that Paulson wanted to be short the index they were creating. In fact, ACA thought Paulson was going to go long (according to press reports this week, the SEC has testimony from a Paulson employee that emphatically contradicts this).

Then here's another catch. When Goldman marketed Abacus 2007-AC1 to buyers, the SEC says it never disclosed that Paulson had been involved at all. The idea is that a buyer may well have felt differently about buying if he thought that a fast-money New York hedge fund had a hand in it, as opposed to only a respectable MBS manager. Especially if that hedge fund was going short!

You can see that the disclosure issue makes Goldman look pretty bad. In my opinion, Goldman should have told the buyers. I don't know if this case will ever go to trial, but if it does, my hunch is that a jury will agree with me. If it doesn't go to trial, that's only because Goldman felt the same way and decided to settle instead.

But in my view that's a pretty small crime in the grand scheme of things, and not a particularly unusual one in the securities business. It's a far cry from accusing Goldman of completely betraying its customers by disloyally creating and marketing a bad product.

Far from it. Goldman oversaw the creation of the perfect synthetic CDO. A derivative like that is at its best when the buyer and the short have maximum disagreement about it -- when one just loves it, and the other just hates it. From that standpoint, Abacus 2007-AC1 is a masterpiece, and a completely fair deal.

This distinction is critical, because Goldman wasn't the only firm to create synthetic CDOs like this. They all did it. Hopefully everyone else had better disclosure than Goldman. But if merely creating these securities is a crime, then the whole U.S. banking system is going to jail. As I said, the SEC isn't looking at it that way. At least not now. But maybe they will start to do so, as a way of turning up the heat on Goldman. Or maybe some other regulator will get into the act and take on all of Wall Street over this, the way former New York Attorney General Eliot Spitzer did after the dotcom bust, claiming conflicts of interest by stock analysts. In the 1980s, when he was a U.S. Attorney, Rudolph Giuliani did the same thing with insider trading, and ended up driving Drexel Burnham Lambert into bankruptcy over it.

Watch the headlines on this. If it turns into a big witch hunt, Spitzer/Giuliani style, last year's banking crisis will look like a buggy ride by comparison. It was bad enough then for Wall Street to take its own losses on subprime MBS. Let's hope Wall Street doesn't get sued into taking all its customers' losses now, too.