About six weeks ago I wrote a post on the strains put on clearing by Brexit. This informative post by Clarus’ Tod Skarecky provides some very interesting detail about the mechanics of the LCH’s margining mechanism.
One way to summarize it is to say that the LCH was a liquidity black hole. Not only did it collect intra-day and end-of-day variation margin from losers that was paid out to winners only with a delay, it also collected Market Data Runs, which were effectively intra-day initial margin top-ups. A couple of perverse features. First, a position that initially had a loss that triggered an MDR outflow had to pay out, but if the market turned in its favor intra-day, it didn’t get that money back until the following day. Second, a firm that had a loss that triggered an MDR outflow had to pay out, and if the position incurred a loss on the day, it still had to pay variation margin, and didn’t receive the MDR back until the next day: that is, there was”double dipping.”
Tod puts his figure on the logic (crucially, the logic from LCH’s perspective): “Heck if I managed credit risk at a firm, I’d always choose to be paid now rather than later.” Definitely. That minimizes credit risk. But look at how much liquidity was sucked up in order to do this.
Variation margin is bad enough: despite the (laughable) claim of the BIS some years back, the fact that variation margin is recycled does not mean that it does not create liquidity strains. After all, (a) liquidity demand arises due in large part to differences in timing between the receipt of cash and the payment thereof, and the clearing mechanism (in which the CCP pays out VM some hours after it receives VM) creates such timing differences, and (b) even absent payment timing differences, the VM receivers would have to lend to the VM payers, which is problematic especially during stressed market conditions. But the LCH IM top up exacerbates the problem because the cash is stuck in the clearinghouse overnight, and therefore cannot possibly be recirculated. More liquidity becomes less accessible.
Again, this is understandable from LCH’s microprudential perspective: it reduces the likelihood that it will become insolvent or illiquid. But just because this is sensible from a microprudential perspective does not mean it is macroprudentially sensible. In fact, it is anything but sensible: it greatly adds to liquidity demand, particularly during periods of time when liquidity is likely to be scarce, and when liquidity freezes are a serious risk.
This is a perfect example of the “levee effect” I’ve written about for years: raising the levee around the LCH increases the chances of its survival, but just redirects the stresses to elsewhere in the system.
Note the irony here. Clearing mandates were sold on the idea that there were pervasive externalities in uncleared derivatives markets, due primarily to the potential for default cascades in these markets. But clearing (supersized by mandates, in particular) creates externalities too. Here LCH does things that are in its interest, but which impose costs on others. It has a contractual relationship with some of these (FCMs), so there is some potential that externalities involving these parties can be mitigated through negotiation and changing contracts. But there are myriad parties not in privity of contract with LCH, and which LCH may not even know of, who are impacted, perhaps severely, by a liquidity shock exacerbated by LCH’s self-preserving actions.
In other words, clearing mandates don’t internalize all externalities. They create them too. And given the severe dangers of liquidity crises, the liquidity externality that clearing creates is particularly troubling.
Brexit’s Impact on Clearing Activity
Let’s first look at the impact on clearing activity. It’s important to remember first that clearinghouses mark all products to market every day, and require that participants with market losses post margin every day, sometimes more than once a day. Margin payments must be paid promptly because for every payment made to the clearinghouse, the clearinghouse must make a payment to another participant who has gains. The clearinghouse always has a balanced or “matched” book.
Even though margins were increased in advance of the vote, the volatility resulted in very large margin calls on June 24.
Clearing members paid $27 billion dollars in variation margin across the five largest clearinghouses registered with the CFTC. This was $22 billion dollars greater than the previous 12-month average—over five times larger. The good news is no one missed a payment, no one defaulted.
Supervisory Stress Tests
The results after Brexit confirmed what we recently found in our own internal testing: resilience in the face of stressful conditions. Last month, CFTC staff released a report detailing the results of a series of stress tests we performed on the five largest clearinghouses under our jurisdiction, which are located in the U.S. and the UK. Our tests assessed the impact of stressful market scenarios across these clearinghouses as well as their clearing members, many of whom are affiliates of the world’s largest banks.
We developed a set of 11 extreme but plausible scenarios based on a number of factors, including historical price changes on dates when there was extreme volatility. By comparison, our assumed price shocks were several times larger than what happened after Brexit. We applied these scenarios to actual positions as of a specific date. And we looked at whether the pre-funded resources held by the clearinghouse—in particular, the initial margin and guaranty fund amounts paid by clearing members as well as the clearinghouse’s capital—were sufficient to cover any losses.
Still not getting it. The discussion of stress tests essentially repeats the same mantra as LCH: it is a decidedly microprudential treatment that focuses on credit risk, not liquidity risk. The discussion of margins is perfunctory, despite the fact that this is what gave market participants serious worries on Brexit Day. No discussion of what extraordinary efforts were required to ensure that all payments were made. No discussion of whether this would have been possible during a bigger–and unanticipated–price shock. No discussion of the liquidity externalities. No discussion of what would happen if operational difficulties (e.g., a technology problem in the payments system like the failure of FedWire on 10/19/87) interfered with the completion of payments. (More payments increases the likelihood that such an operational failure will jeopardize the ability of FCMs to complete them. And a failure to meet a call triggers a default.)
This “what? Me worry?” approach sounds so . . . 2006. And it is exactly this kind of complacency that makes me worry. The nature of the liquidity issue still has not penetrated many regulatory skulls.
This is most likely due to a severe case of target fixation. Clearing mandates were motivated by a desire to reduce credit risk, and all efforts have been focused on that. That is the target that regulators are fixated on, and in the pursuit of that target their field of vision has narrowed, with liquidity risk being largely outside it. It is obviously the target that CCPs are focused on. This is why I take little comfort in the belated efforts to make CCPs more resilient. The recipe for resilience is to demand MOAR LIQUIDITY. Which is also the recipe for a broader market crisis.
Analogous to the dangers of high powered incentives with multi-tasking when some activities can be measured more accurately than others, the mandate to reduce derivatives credit risk has led regulators and market participants–particularly market utilities like CCPs–to devote excessive effort to mitigating credit risk, even though it exacerbates liquidity risk.
I doubt the clearing portions of Title VII of Frankendodd will be eliminated altogether, but the incoming administration should seriously consider a major re-evaluation to determine how to address the serious liquidity issues that clearing mandates create.