Goldman and the Financial Crisis

Back in 2010, in a hearing in front of the Senate Finance Committee, Goldman Sachs was accused of wrongdoing that helped precipitate the economic collapse of 2008. Goldman CEO Lloyd Blankfein nonchalantly dismissed these charges, at times looking confused over why anyone would be mad at him. People were – probably still are – mad because they think he was unethical in betting against the very securities he was selling.

The securities were the notorious collateralized debt obligations (CDOs) whose toxicity helped spawn the systemic collapse of some of Wall Street’s biggest institutions. When housing prices declined these CDOs – packages of mortgages of varying levels of risk – plummeted in value. The resulting undercapitalization of banks caused the credit markets to seize up, and the reverberations were felt across Main St. But Goldman, having placed bets that this would happen, profited.

Goldman is a market maker. They quote a price at which they’re willing to buy and sell a particular security and follow through when orders come in. Goldman in this regard isn’t a direct speculator in the market. It profits by charging a higher price for a buy order than a sell order, the difference in which is known as the bid-ask spread. It is therefore subject to loss when the order volumes are skewed against them. In order to defray this risk and put itself in position to profit on the fate of the security, the market maker alters their respective long- and short- positions on the securities they sell.

Some sources place blame for the financial crisis on excessive lending to the underqualified. On this view the borrower’s default was the direct cause of the crisis. But the consequences of these lending practices were multiplied by the structure of our financial sector. Goldman Sachs, along with the other big investment banks on Wall St., swapped bets on these mortgage packages. They referenced the mortgages as an underlying asset without holding the actual debt. They played mortgages like gamblers play the ponies, with just as much regard for the asset’s as on a race’s prize money. In this way they created a second-tier of risk, over and above the market for the asset itself, based only on other banks’ desire to gamble.

The industry says that in making these bets they are providing liquidity to the market – allowing the underlying asset to be more easily bought and sold at a more efficient price. By making these second-order bests on mortgage-backed securities, the story goes, the asset is more accurately priced to reflect the market’s belief of their value. As a result, capital is more efficiently allocated in the market. There are two potential problems in the proposition that Goldman was furthering this liquidity goal by making the market.

The first is the problem of fraud. If Goldman purposefully packaged their CDOs in such a way as to make them fail, and that likelihood of failure was misrepresented to their clients, that would be fraudulent behavior. Whether Goldman’s actual behavior – failing to disclosure their involvement in the origination of, and position with respect to the CDOs – constitutes fraud was left to the courts, though never decided. A $500 million settlement prevented a definitive ruling on Goldman’s fraud, I turn to the other concern.

The second problem is that of inefficient, and dangerous, information asymmetry. An information asymmetry exists when the seller has more information about their product than the buyer. Take for example Goldman’s so-called Hudson C.D.O., a bet  Goldman made with rival bank Morgan Stanley on the fate of $2 billion worth of subprime mortgages. Goldman took a $2 billion short position in the deal (meaning a $1 drop in value yielded a $1 Goldman profit), while Morgan Stanley took a $1 billion long position. Goldman provided in this Hudson CDO the level of risk exposure Morgan Stanley wanted, based on the credit ratings of the 3 institutional rating agencies on Wall St – they acted as a neutral market maker to provide Morgan Stanley liquidity, while defraying what it thought was substantial risk.

The Senate Committee wanted to accuse Goldman of conflict of interest in this market making – betting against the product you’re selling creates adverse incentives and is unethical, they said. But Goldman responded that market making is about providing the client with what they want, owing no duty to that client’s financial interests. So maybe Goldman didn’t do anything legally, even ethically, wrong outside of the Fraud accusations. But even those are tenuous on the assumption of working credit agencies; the exposure calculation should be neutral, the hand that packages irrelevant.

So the question remains, why did two sophisticated investment banks working on the same rating information make such big bets against each other? Next time: the role of credit agencies and proprietary information in this mess.