Streetwise Professor

September 16, 2016

De Minimis Logic

CFTC Chair Timothy Massad has come out in support of a one year delay of the lowering of the de minimis swap dealer exemption notional amount from $8 billion to $3 billion. I recall Coase  (or maybe it was Stigler) writing somewhere that an economist could pay for his lifetime compensation by delaying implementation of an inefficient law by even a day. By that reckoning, by delaying the step down of the threshold for a year Mr. Massad has paid for the lifetime compensation of his progeny for generations to come, for the de minimis threshold is a classic analysis of an inefficient law. Mr. Massad (and his successors) could create huge amounts of wealth by delaying its implementation until the day after forever.

There are at least two major flaws with the threshold. The first is that there is a large fixed cost to become a swap dealer. Small to medium-sized swap traders who avoid the obligation of becoming swap dealers under the $8 billion threshold will not avoid it under the lower threshold. Rather than incur the fixed cost, many of those who would be caught with the lower threshold will decide to exit the business. This will reduce competition and increase concentration in the swap market. This is perversely ironic, given that one ostensible purpose of Frankendodd (which was trumpeted repeatedly by its backers) was to increase competition and reduce concentration.

The second major flaw is that the rationale for the swap dealer designation, and the associated obligations, is to reduce risk. Big swap dealers mean big risk, and to reduce that risk, they are obligated to clear, to margin non-cleared swaps, and hold more capital. But notional amount is a truly awful measure of risk. $X billion of vanilla interest rate swaps differ in risk from $X billion of CDS index swaps which differ in risk from $X billion of single name CDS which differ in risk from $X billion of oil swaps. Hell, $X billion of 10 year interest rate swaps differ in risk from $X billion of 2 year interest rate swaps. And let’s not even talk about the variation across diversified portfolios of swaps with the same notional values. So notional does not match up with risk in a discriminating way.  Further, turnover doesn’t measure risk very well either.

But hey! We can measure notional! So notional it is! Yet another example of the regulatory drunk looking for his keys under the lamppost because that’s where the light is.

So bully for Chairman Massad. He has delayed implementation of a regulation that will do the opposite of some of the things it is intended to do, and merely fails to do other things it is supposed to do. Other than that, it’s great!

Print Friendly


  1. Hypothetically speaking of course, any business that was going to be caught by Jan 1, 2017 deadline would have already been taking action. It’s not like one could wait until December to start the ball rolling here. The push back is too late.

    Measuring notional is a bad regime, but it’s better than many others regulators could dream up. There’s wildly different risk profiles between the items you mentioned. Two countervailing points: future catastrophic is likely to be most concentrated in products that any risk based scheme created today view as low risk (which is why everyone will get over-extended to begin with) and likely a more nuanced system created by the CFTC would be nightmare with gaming, perverse incentives, and other second order concerns.

    Comment by FTR — September 17, 2016 @ 10:18 am

  2. Same as much rat, just later.

    Comment by Green As Grass — September 19, 2016 @ 11:15 am

  3. @Green. Yeah. Giving everyone time to work up an appetite.

    The ProfessorComment by The Professor — September 19, 2016 @ 12:23 pm

  4. Does SWP or any reader have any hard dollar number estimates (other than the CFTC’s own cost-benefit analysis) on exactly what amount is the “large fixed cost” to become a swap dealer?

    Comment by job — September 19, 2016 @ 12:35 pm

  5. @ job

    Not offhand, but it’s all kinds of things like reporting obligations and regulatory capital requirements, that are cold hard cash or headcount. Look at how many EFS trades there were in 2011 (lots) versus now (none) on a big US market like CME or ICE. D-F has constructively outlawed the OTC swap, something it never ostensibly set out to do.

    An interesting thing about regulation is that you never, ever, ever see it put forth with a proper risk assessment of what bad outcomes may eventuate and whether they obviate the regulation being proposed. All new regulation is assumed to be literally riskless.

    Examples abound. In Europe firms are now required to submit an online report to the regulator of any suspected market abuse. Tardiness in doing so may be sanctioned. The actual abuse almost never. So if you’re a firm, and a market abuse suspicion takes you so long to bottom out that your report’s sure to be late, do you submit late and get sanctioned, or do you just not report? The question answers itself really.

    Comment by Green As Grass — September 20, 2016 @ 4:46 am

  6. In fairness to Frank-Dodd, it’s not like there is any recent precedent for economic problems caused by notionally sound investments for which no risk-weighting was performed… oh no, wait! I think there might have been a minor event in 2007… but perhaps nobody in Congress remembers it… (I guess it is easy to forget such events when you are part of an elite which is not affected by it).

    Comment by Hiberno Frog — September 22, 2016 @ 2:30 am

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress