Why Goldman's $10b Repayment Offer Is a Bad Deal

Goldman Sachs has always had a lot of swagger, and for good reason. It's long been the richest and most storied investment bank on Wall Street, producing Croesus-like profits and a string of Treasury secretaries, to boot. So it was perhaps no surprise this week when Goldman announced its plans to be the first bank to pay back the $10 billion in TARP money it took from taxpayers last year. Here are some good reasons Tim Geithner shouldn't let Goldman throw off the government shackles quite so quickly—for all our sakes.

Goldman's announcement came off the back of a stunning first-quarter earnings report, in which it announced net revenues of $9.43 billion, and earnings of $1.81 billion. Not bad at a time when most everyone else is still running for the hills. So how did they make all that money? According to CFO David Vinair, "The environment in the first quarter was such that … there were so many opportunities in truly liquid assets that there was no need to use liquidity to buy illiquid assets, and there weren't a lot of good illiquid assets for sale." Translation for those who don't speak financese: "We're the hottest traders on the planet."

That may be true. But there is also likely much more at work here than that. We've already noted in the Wealth of Nations blog how a fortuitous shift in accounting dates allowed Goldman to leave December—a month with a lot of write-offs—mostly off the books. And that may not be the only form of accounting sleight of hand behind the good news on Goldman's balance sheet.

On Friday, a high-level source in credit research (who didn't want to be named speaking about Goldman) walked me through the bank's recent financial statements. This person noted that Goldman has some $585 billion worth of what are known as "level two" assets—securities which may not have a clear market price but which can be accounted for in the company's books by comparing them to a similar asset (Goldman would neither confirm nor deny that $585 billion number, which isn't public). That leaves a lot of wiggle room when it comes to recording the value of such assets. If a small fraction of those assets are actually worth less than Goldman assumes, it could account for a few billion worth of the bank's first quarter revenue. A spokesman for Goldman would not definitively confirm if anything had been marked up (mark ups don't necessarily have to be announced immediately), but did say, "Goldman Sachs did not announce any material write ups for the first quarter."

I ran this scenario by another source, Sean Egan of the Egan-Jones Group, one of the country's largest independent credit-ratings firms (hold on to that "independent" notion, we're coming back to that). Not only is it plausible that Goldman's level-two assets have been overvalued, Jones said, but it is impossible to know based on public paperwork. So, when will we know if Goldman's stellar quarter actually represents some end to the crisis—at least for the world's most-storied bank—or just more of Wall Street's too-familiar accounting highjinks? "You won't know until it doesn't matter," was Egan's verdict. Government money has helped Goldman out of the crunch; meanwhile, problematic practices like the lack of accounting transparency or clear oversight of banking practices continues. As Martin Wolf so smartly summed it up in the Financial Times this week, if Treasury's current plans to get the financial system up and running work, it will be because they are a "non-transparent way of transferring taxpayer wealth to banks."

We all know that the bankers at Goldman are smart enough to make money in a very down market. And most of us can see that chairman Lloyd Blankfein's posturing about being "humbled" is less about learning any particular lessons than about wanting the government and its pesky rules about compensation off his back as soon as possible (leaving Goldman free to scoop up even more talent from competitors, further solidifying its market dominance). But what we need to know in order to get credit markets as a whole up and running is whether we can trust what's going on inside of banks.

That gets us back to the point about independence. The essential conflicts of interest in the ratings industry, for example—in which major players like S&P, Moody's and Fitch receive compensation from the very banks and corporations whose financial health they are rating—has already been well publicized. Yet, to date, nothing about that system has changed. This is a fundamental flaw, and one that desperately needs to be addressed if we are to reestablish trust in the U.S. banking system (which controls, by far, the largest chunk of the world's financial flows).

Another—and, perhaps, more fundamental—problem is the revolving door between the public and private sector, and, in particular, between banks like Goldman and the government (for those who may have forgotten, this bank that has produced so many Treasury leaders was also allowed to pay off its credit default swaps with its holdings of AIG—now owned by the state). Goldman may be eager to pay back its TARP money. But we shouldn't necessarily let it off the hook so quickly. Regardless of whether Goldman can cough up the $10 billion, poor Tim Geithner, who is still a one-man shop (perhaps, not surprisingly, he's finding it hard to hire deputies) needs time to evaluate what's really changed—if anything—at Goldman and the other banks now on government life support. The only thing that's really clear so far is that what looks good in a quarterly earnings report may not represent an end to our financial troubles.

All of which leaves us with one clear conclusion: as nice as it might be to have some of the TARP money back in government coffers early, letting Goldman slip away from Washington's intrusive gaze at this point won't help anyone outside of Goldman.