Eliminating Some, Not All, Risk

Yesterday, I wrote a post about how to attract employees for high quality regulators, or really how to pay for them. In response, a commenter noted that government cannot eliminate risk. I agree with that statement to a point: the government cannot eliminate all risk. Indeed, the government should not eliminate all risk, as an economy where firms don't take risks will have little innovation or growth. But the two aspects of regulation I was referring to really didn't refer to that kind of risk so much as creating a healthy financial system.

In the post, the regulation I addressed was specifically aimed at two targets: fraud and a new resolution authority. Fraud is swindling; it's illegal. So saying the government shouldn't have good regulators try to prevent fraud is like saying it shouldn't have good police to prevent murders. Certainly not all crime can be eradicated, nor can all fraud. That doesn't mean that it isn't a legitimate goal for the government to try.

As for the new resolution authority, recently I've don't a lot of pleading for writing about its possible creation. It would help facilitate the tidy failure of large and complex financial institutions. I've said that such a regulator should acquire the potential failure plans for systemically relevant firms (as suggested by Treasury Secretary Geithner). I further asserted that those firms should pay fees for the resolution authority to audit those plans to be sure they would actually work. They should also pay for any costs upfront that their failure would cause the broader financial system. But this authority isn't really about regulating away all risk.

In fact, this new agency would have no rule making capacity: it wouldn't really regulate. It would simply determine which firms might pose systemic risk, require them to submit plans and evaluate those plans. The risk it would seek to eliminate would be that the economy would face if one or more of these firms fail. In order to accomplish that goal, it would not have to anticipate when a future market shock might happen. Instead, it just has to understand the economy's connectedness. When a given firm fails, it needs to figure out how that one domino can fall but the rest remain standing.

I think that goal is not only possible, but desirable. As I've said before, such an authority would actually strengthen capitalism, because it would prevent the need to ever bail out another firm. Instead, bad companies would fail, as they should. But in doing so, they wouldn't take the entire economy with them.

I hope this makes clear that this authority wouldn't really exist in order to prevent firms from taking risk. In fact, it would allow firms to take as much risk as they want -- at their own peril. If that risk was too much, and the firm fails, then the resolution authority could quickly sweep up the pieces, and the economy would move on. It would eliminate the systemic risk, so that other firms don't suffer as a result of one bad apple.