What is a Systemic Risk Regulation?

Systemic Risk Regulation

If one wants to discuss this issue seriously, it is essential to decide what produces systemic risk. While that seems easy, it is surprising how quickly people can “take sides” rather than be realistic. Perhaps that is the result of a need to find a sinister someone to blame. But, if one wants a sinister force, watch a James Bond movie or Star Wars. We are supposed to be addressing systemic risk; risk in the system as well as risk to the system. Otherwise, if the risk to the system is due to an evil villain, it is evil villain risk: Doctor No, Goldfinger, Specter, or Darth Vader risk. If it is evil villain risk, then we need evil villain regulation.

Here’s a question. How many of the candidate evil villains are already regulated? Not all of them are, but a lot of the current candidates are, for sure. Part of the problem was that we had regulation of the individual players but no one watching the game. The proposed reforms would re-order the entire regulatory framework. That in itself is important. But, re-ordered oversight, even with regulatory oversight added for the players not previously supervised, is not systemic oversight.

To illustrate, starting in 2003 or 4 and up until recently, The Hedged Economist had the good fortune to be talking to regulators of depositary institutions. At one, I presented a sample of some data and advanced the argument that with the data the agency could better measure systemic risk. The response was, “we know the risk isn’t on the books of the banks we supervise.” In other words, their definition of systemic constrained their focus to the internal risk within the institutions they supervised.

At another, the comment was that the risk is out there, broken up and spread around. The natural response was: “who is baring that risk?” The answer, “no one knows.” Since no one knew where the risk was, no one felt that they were responsible for measuring it, much less managing it. It was in the system, not in any individual locale. At two other agencies, the response was that the risk was too big for them to worry about given their limited resources. They just passed on systemic risk.

At another, the response was that they could not act on the risk for political reasons. In short, systemic risk was the politicians’ issue, not their issue. At the agency that was most concerned, the issue was acknowledged as was the realization that they would have to clean up the mess. However, they could not do anything to manage the risk if no one else recognized it. They had no mandate.

Needless to say, many of the firms they all supervised got clobbered, not by what the firms did, but due to the surfacing of risk in the entire system. Most of the agencies did spectacular jobs of minimizing the negative consequences of the risk within the framework of their organization. All of them had bright, motivated people, but none of them focused on the entire scope of the issue. Just re-ordering responsibilities will not eliminate that problem. So, at this abstract level, adding systemic risk oversight makes sense.

But the reader may have noticed that much of the discussion so far focused on what isn’t systemic risk and what isn’t systemic risk oversight. Well, let’s take that approach a step further.

First, a lot of people equate managing systemic risk with producing stability. It is not the same. It shouldn’t be. For starters, perfect stability is an absence of change in other words stagnation. Thus, if stability and systemic risk management were equivalent, the job of managing systemic risk would be to produce stagnation. That clearly should not be the objective.

Second, the purpose of systemic risk oversight is not to protect us from all risk of negative outcomes. Risk is inherent in every activity. Further, one can draw on some behavioral economics to see where this objective could lead. Behavioral economists tell us that most people feel loss more intensely than gain. If the primary task of a systemic risk regulator was to balance negative and positive outcomes, the bias would be to overweight avoiding negative outcomes.

Third, the systemic risk regulator’s job should not be to pick winners and losers. Clearly, effective systemic risk oversight requires making judgments about what activities create systemic risk and reining them in. But, the focus has to be on reining in the activity, not reining in actors. There will be winners and losers as a result, but they can not be the focus without subverting the objective of managing the (as yet ill-defined) phenomena of systemic risk.

Finally, the purpose of systemic risk management can not be just to keep people happy, to win popularity polls, or to get elected. That would preclude the possibility that systemic risk could ever originate from “the will of the people.”