Thursday, January 21, 2010

The Selection Bias in Goldman's Argument Against the Obama Regulation

Subscribe in a reader

The proposed Obama regulations that I'm referring to are the statements that came out today that he would:
"seek to prevent banks that have special access to low-cost Fed funding from operating or investing in hedge funds, private equity funds, or trading purely for their own benefit in a way that’s unrelated to serving customers." [Bloomberg]

While this is sure to be contentious, I think it's important to take a step back and look at what especially discounted Fed funding is intended for. I'm not talking about the funding that the Fed provides during a "business as usual" scenario, but in scenarios such as this recent credit crisis, when the discount window's possible term length was raised to 90 days and interest rates slashed, the goal was presumably to get some emergency liquidity out there to try and unfreeze credit markets and stimulate lending again.

I won't try to quantify the potential for ruin that hedge funds, private equity, and proprietary trading might potentially represent to a bank, although this is certainly possible. If the goal is to stimulate lending, however, having all these components as parts of banks that have the discount window available to them would make it difficult to monitor where these emergency funds might be going once they leave the Fed. While these other activities no doubt have the potential to be economically important, I don't think there would be significant objection that the Fed's primary objective in a crisis, is to funnel funds to firms that are going to use that money to lend and keep credit flowing, might be hindered if these funds are instead going to any number of other activities available to some of the firms that received emergency Fed funding.

This leads to the Goldman comments coming from Chief Financial Officer David Viniar. He argued:
"If people are focused on things that caused, or were real contributors to the crisis, it wasn’t trading...Most trading results were actually pretty good, not just at Goldman Sachs but at most firms and that’s not really where the problems were." [Bloomberg]

Looking at the results of trading from the crisis as a sample of typical trading is a vivid example of selection bias. It seems that Viniar would like us to think that because most trading was profitable during the crisis, it will be profitable in the future and doesn't doesn't pose systemic risk to banks. While on average trading profits might be positive, that says nothing for the skewness and kurtosis of trading profits/losses. By that I mean that proprietary trading could be negatively skewed or with fat tails (positive kurtosis) in the negative end of the distribution.

This is an example of a positively skewed, positive kurtosis, positive mean distribution. While it will be on average positive most of the time, the long negative tail creates the possibility for financial ruin when returns end up in that part of the distribution.

I'm not going to take sides on this issue, since I can see the validity in both. That being said, I think that Vinair's argument regarding proprietary trading is mistaken. For example, while fund managers can on average beat the market by holding a portfolio of the S&P 500 and writing puts (insurance against a fall) on the S&P 500, when it does go bad, it's ruinous. This is the "picking up nickels in front of a steamroller" scenario.

While I don't claim to know the distribution of trading profits or the absolute level of risk these firms are taking on (they might not know either), I can understand why Obama and his administration would want to prevent artificially-lowered interest rates from fueling these types of activities. Especially since it's meant to stimulate lending, allowing firms that have access to the Fed's discount window to engage in hedge funds, private equity and proprietary trading could be a usage of funds not best for unfreezing financial markets.

That being said, Obama needs to be careful when he talks about "curbing excessive risk taking". Lending to small businesses is one of the riskiest types of loan a consumer bank can do, and since stimulating the flow of credit to areas of the economy such as this seems to be an Obama administration priority, curbing risk taking shouldn't be his goal.

No comments:

Post a Comment