Using Economic Analysis in SEC Rulemaking

Editor’s Note: Elisse B. Walter is a Commissioner at the U.S. Securities and Exchange Commission. This post is based on Commissioner Walter’s recent remarks at the Conference on Current Topics in Financial Regulation, which are available here. The views expressed in the post are those of Commissioner Walter and do not necessarily reflect those of the Securities and Exchange Commission, the other Commissioners, or the Staff.

As you may know, the SEC has recently enhanced its economic firepower, through, for example, significantly increasing the number of PhD economists in the Division of Risk, Strategy, and Financial Innovation. Lately much of the external focus on the role of economic analysis at the SEC has been on cost-benefit analysis – which is certainly an important part of economic analysis. However, it is not the only way that the Commission is using economic analysis in our work. Increasingly, our economists are getting involved earlier and more comprehensively in the rulemaking process, not just to help the Commission weigh the ultimate costs and benefits of our regulatory decisions, but to provide a reasoned framework for making those decisions. Examples include providing up-to-date information about the current state of the markets, and helping us think of alternative ways to meet our regulatory goals.

I believe that these efforts are bearing fruit, and I would like to provide a recent example of a significant regulatory action where, in my view, we used economic analysis effectively to guide our decision-making. This was in our adoption of the rule defining “security-based swap dealer” under Title VII of the Dodd-Frank Act, as part of a joint rulemaking with the CFTC. Further defining the term “security-based swap dealer” was one of the many tasks that Congress assigned to us as part of creating a new regulatory regime for security-based swaps. Congress also mandated that the Commission exempt from the dealer designation an entity that engages in a de minimis quantity of dealing activity. Again, however, Congress left it to us to hammer out the details of what would constitute a de minimis level of dealing activity. Considering that the over the counter derivatives market is still a largely unregulated space, determining an appropriate de minimis level seemed like a daunting task, and the comments we received reflected a diversity of views on what this de minimis level should be.

We did, however, have some data at our disposal – the transactions in single name credit default swaps in 2011 that were submitted to the Depository Trust and Clearing Corporation’s Trade Information Warehouse. I understand from our economists that the relevant data was very complicated, and that it was a significant undertaking for our staff just to begin to understand what the data could be used to show. Moreover, when we received the data, it was not yet in usable form, and therefore, for example, the economists had to go through the data by hand to verify certain information for each of the 1500 market participants.

Although the data covered CDS trades and not the entire security-based swap market, the CDS trades represented the vast majority of security-based swap activity. Based on their understanding of what the data showed, our economists identified several factors that could be indicative of dealing activity, such as the number of counterparties that a participant had, particularly the number of counterparties that were not dealers recognized by the International Swaps and Derivatives Association, known as ISDA. Then, using these factors, our economists analyzed the data with the goal of identifying market-making activity commensurate with that of a dealer. The analysis, which can be found on the SEC’s website as part of the proposing release’s comment file, was quite illuminating and helped me greatly in reaching my decision on the rules.

The results indicated a high degree of concentration of potential dealing activity in the credit default swap market. Of course, the analysis itself did not reveal the precise level at which to set the de minimis threshold. That would be expecting too much. Instead, it showed the possible effects of a range of thresholds, and indicated that setting a de minimis threshold at a level of $3 billion notional over a 12 month time period to even as high as $100 billion notional would likely result in dealers operating above the de minimis threshold being involved in 98 to 99 percent of the single name credit default swap market. In other words, 98-99% of the market likely would be intermediated by dealers that will eventually be subject to regulation. To give you an idea of how illuminating this was, our original proposal, without the benefit of this data, proposed a de minimis threshold of $100 million.

This data gave my fellow Commissioners and me a pretty good picture of the market as it exists today. My years as a regulator have taught me, however, that regulations can have a powerful effect on shaping incentives. This issue, present in every regulation we make, was especially relevant for this rulemaking. This was the first of what will likely be many rules that will eventually convert the security-based swap market from a largely unregulated market to a regulated one. Congress directed us not just to define security-based swap dealer, but also to implement mandatory clearing of many types of security based swaps, require the reporting of data concerning such transactions, and regulate security-based swap dealers and the clearing agencies, data depositories and exchange platforms that will provide the market infrastructure. Many of these changes are intended to increase the transparency of the security-based swap market, as well as lower barriers to entry. Consequently, the market might become less concentrated as a result of the new regulatory regime. The higher we set the de minimis threshold, the more likely it would be that security-based swap activity could take place outside of a regulated dealer because of the appearance new entrants.

Again, the data was a useful tool in illustrating the potential effects of setting the de minimis exception at different levels. Using the same data, staff used an assumption of 15 additional entrants to the dealing market, about double the number of dealers currently recognized by the OTC Derivatives Supervisors Group chaired by the New York Fed. Under this assumption, the data indicated that a $100 billion notional de minimis exception could have allowed as much as 15% of credit default swap dealing activity eventually to fall outside of the regulated market, and a $25 billion notional exception could have resulted in up to 4% of activity being unregulated. Of course, we recognized that 15 new entrants was only an assumption – there could be 5 new entrants, or there could be 25, or there could be more. We also recognized that these observations were based on looking at historical data, and that market dynamics could change these estimates when the market transforms from an unregulated to a regulated one. Likewise, we did not know whether such entrants would or could operate right at the limit of the de minimis thresholds.

Of course, I realized that none of the data was going to reveal one “magic” number – even if we knew how many new entrants there would be, we would still need to make the judgment call about what proportion of the market we thought was appropriately considered de minimis. So the data did not make the decision for us. Rather, the data helped us better understand the potential impact of our judgment calls and provided us with a reasonable framework to make our decisions.

In the end, we chose a de minimis exception of $3 billion notional threshold over 12 months for dealing in credit default swaps under our jurisdiction, and a $150 million notional de minimis exception for dealing in other security-based swaps. The latter determination was based upon estimates that security-based swap activity, other than credit default swap activity, constitutes about one-twentieth of the security-based swap market. Taking into account the extra protections that Congress accorded federal, state, and municipal agencies, employee benefit plans, and endowments, referred to as “special entities” under the Dodd-Frank Act, we set a lower de minimis exception of $25 million notional for any dealing activity with special entities as counterparties.

As I mentioned earlier, the data indicated that an even higher figure likely would have covered 98% to 99% of the marketplace – so one could reasonably consider the $3 billion figure actually to be at the lower, more conservative end of the range of choices. And, once again, the data made me confident that this was a reasonable, and in my view optimal, regulatory choice. That is because the data painted a picture of a market where the likely dealing activity was so heavily concentrated that it was not clear that a $3 billion de minimis would capture many, if any, more market participants, than, say an $8 or $10 billion figure. Given that the potential regulatory cost of a lower de minimis threshold was therefore likely to be small, it seemed to me that the prudent approach would be to increase the likely regulatory benefit by picking a number at the lower end of the range and reducing the potential that future dealing activity would fall outside of the regulated market.

There were still several issues for us to face, and those were how to account for the fact that our data did not cover all security-based swap activity, that none of our factors for identifying potential dealing activity would actually identify all dealers, and, most importantly, that we did not know how this market was going to evolve as it became regulated. Also, since we were selecting a de minimis threshold toward the lower end of the range, it made sense to give market participants some time to conform their security-based swap activity to the rule. So, we agreed upon a phase-in period, during which time the de minimis level would be $8 billion notional for dealing in credit default swaps and $400 million for dealing in other security-based swaps, with no higher phase-in de minimis exception for dealing with special entities.

I have to admit that I was a bit annoyed the morning after we adopted the rule to see headlines saying “SEC adopts $8 billion de minimis for dealers.” I want to make clear that based upon the data available, I believe that $3 billion was the appropriate level to set the de minimis exception for credit default swaps, and $3 billion is what I consider the de minimis exception to be. However, the data helped me get comfortable with the phase-in process and the triggering levels we set. The high market concentration gave me assurance that the vast majority of security-based swap activity in the market as it exists today would still be regulated during the phase-in period.

The Commission also directed the staff to study the data that would be collected on security-based swap transactions once the rest of the regulations concerning security-based swaps were in place. Based on the results of this future study, the Commission will be able to determine whether de minimis levels different than the ones we adopted would be appropriate. I applaud staff for suggesting this construct. I think this was a pragmatic and innovative solution, and represents a model that I hope we will use again when the circumstances allow it. Data will not always be available for our decision-making, and even when we do have data that we can rely on to make a well-reasoned estimation of our regulations’ impacts, it remains only an estimation. Phase-in periods, pilot programs, and similar measures, with a commitment to revisit such decision-making with data gathered once the regulations are in place, can be an effective approach when there is a need to make regulatory decisions in the face of uncertainty.

So that’s my real-life example of how economic analysis continues to be an important influence in financial regulation. If you had asked me before our economists’ analysis whether I would have agreed that $3 billion could be considered a reasonable de minimis level for security-based swap dealing (or just about anything for that matter), I’m pretty sure I would have been skeptical. But the data gave me comfort that setting the de minimis exceptions at the levels we selected was reasonable, responsible regulation.

I will say that I believe exercises such as this are just one important part of the regulatory process. Qualitative analysis is also a vital component for framing the issues and considering the choices that regulators need to make. Indeed, a qualitative analysis of the possible effects – not to mention the costs and benefits – of our rules is a prerequisite for even determining what types of quantitative analysis might be helpful. Only once we have a thoughtful understanding of the universe of possible considerations that we should take into account when regulating can we ask the subsequent question of what quantifiable data is available to help us in our evaluations. I hope we will engage in the kind of thoughtful quantitative analysis I’ve described above where feasible, but understand that there will sometimes be cases where data is not available or quantification is not reasonable. When we cannot engage in this kind of quantified analysis, I believe our releases should transparently – but unapologetically – explain why.

Most importantly, even the finest analysis with perfect data can only inform a regulator’s choices. It cannot make the tough judgment calls for us. But the data analysis certainly was a vital component of my decision-making in this instance, and I look forward to our economists and rigorous economic analysis playing an ever important role in the rulemaking process.

And with that, I will leave you with a heartfelt thanks for the measures you are taking to increase the role of economic analysis in regulatory decision-making, and also ask for your help. First, keep the studies and reports coming. While our Division of Risk, Strategy, and Financial Innovation is full of bright minds, your involvement is important to our mission as well. Second, comment, comment, and then comment again on our proposed rules. Attach your studies; include your data; lay out your views; spell out your analysis. Tell us when we should be considering alternatives; tell us what you think of alternatives and analyses that we or other commenters have provided. Your work has to be in our comment file in order for the Commission to rely on it, and more than anything we need a greater number of disinterested, reasoned views of our proposals. Finally, and equally important, I encourage you to come by to see me, and our Chief Economist Craig Lewis requested that I invite you come see him, as well. In person discussion is so helpful. Tell us what you are working on, and what you think we should be doing to better achieve our mission. Teach me something new (in my view, Craig already knows everything, though he would dispute that) and help us help the nation’s investors. My door is always open, and I look forward to the opportunity to learn from you.

Both comments and trackbacks are currently closed.