Barely a day goes by when you can't marvel at Goldman's PR, they never miss an opportunity to polish their narrative. Here's what their CFO said last week:

"Primarily in response to our client needs, our equity derivatives business was short volatility entering the second quarter and posted poor results," Chief Financial Officer David Viniar said on a conference call with reporters on July 20. "We took the other side because you know we deal with our clients all the time," he said. "We had that position going into the quarter and volatility just spiked."

As "technically true, but substantively crap" statements go, this is a really good one.

Just to provide a little of the context here, Goldman's results last week were impacted by pretty ordinary trading profits - and here in this call, David Viniar is expanding further on where certain larger trading losses crystallized.

And obviously, Goldman is still dealing with the Abacus mess - the SEC fine was one thing, mollifying client concerns that it screws them over is the meta-battle right now for the firm. So what better way than to tie a relatively disappointing trading quarter with the line, "what can I say, we're the kind of folks that will take one for our clients."

Let me explain what actually happened.

Goldman sells equity derivatives by the truckload to clients, and life insurers are particularly reliable purchasers of these things called variance swaps. As the linked Bloomberg article explains, this is a product that helps insurers hedge against a stock market collapse, although the true risk is actually explained by a complicated sounding thing called "implied volatility". In plain English, implied volatility is essentially investor expectations on the volatility of the stock market, and the purchasers of variance swaps are taking a position that will benefit them if actual volatility rises above expectations (they lose if volatility ends up lower than expected)

Goldman Sachs, as Viniar says, takes the other side of the trades. This is self-evidently true, Goldman trades as principal, so if a client goes long volatility, Goldman is short.

What's deceptive is presenting the notion that Goldman was overall short volatility "primarily in response to client needs". You see, when Goldman takes the short side of the trade, it doesn't have to remain short. It can hedge the risk, probably with another bank, and it will hedge itself if the firm's view (agreed by trading heads and risk managers) is to go long volatility (or at least, not go short).

What actually happened last quarter is Goldman sold the derivatives and collected its usual premia on the variance swaps, and had separately established a view that it would remain overall short volatility, so it sat on an unhedged position of an agreed magnitude. Maybe they thought hedging was too expensive, far more likely, they just wanted to be short. And why not? Historically, at least looking back the last 5 years and stripping out the mayhem of H2 2008, VIX at the start of the previous quarter was still rather higher than usual. What's not to like about that trade?

A lot, of course, when it ends up losing you possibly in the region of nine figures. The spike last quarter was pretty extreme, and anyone short volatility like Goldman would have been in fat-tail territory.

Still, polishing this turd the way Goldman has with the "we did it for our clients" spin is splendidly brazen. The truth is, they'd have done these trades with clients regardless of their view on the market - they might have done the trades at marginally higher levels to squeeze about a bit more of a premium, but they'd still have done them.

Lastly, and this again is where Goldman is as usual very finely attuned to the political atmosphere, Goldman is suggesting that these trading losses are not proprietary trading losses. Because they came about from client-driven business. This is Grade A bull, but right when the stringency of the Volcker rule is under the microscope, the timing is impeccable.

The Volcker rule proposes that proprietary trading be banned (or at least isolated) by firms, like Goldman, that are backstopped by the Fed. And here's Goldman spinning a large proprietary-trading loss as a client-trading loss. The truth is, no-one in the industry believes that unless you tackle head-on firms which trade as principal (i.e. all the big firms), can you realistically clamp down on proprietary trading; this is a prime example why.

Anyways, next conference call when probably Goldman will be reporting another successful quarter, maybe some bright spark can ask David Viniar if their profits came from positions that existed "primarily in response to client needs". I'd expect they'd have a ready and polished answer, but I'd still be interested to hear it.