New York | Q: How do you say “black swan” in Norwegian? A: sort svane. But hold that thought for a moment.
Last week readers of The Institutional Risk Analyst probably did not notice the failure of a small NASDAQ clearing member in Scandinavia. The dollar amount of the default event was relatively small, but the details are disturbing in terms of the use of hidden leverage in the trade and the potential risk to central counterparties (CCPs) such as exchanges.
The NASDAQ made the following statement:
“On Monday September 10, 2018, the markets in Nordic and German power experienced an extreme movement in the spread. One of Nasdaq Clearing´s members had a portfolio containing a large spread position between Nordic and German Power that was negatively impacted by the fluctuations.”
Negative indeed. We hear in the credit channel that the “private trader” leveraged his spread position multiple times, then doubled down again before the trade went sideways. The “relevant clearing member” was unable to meet a margin call, says NASDAQ, so the exchange saw the position closed out and liquidated at a loss. The exchange replenished the default fund via a $107 million euro assessment of the other clearing members. NASDAQ also stated:
“When analyzing the data of the event, we concluded that the market movement leading to the extraordinary fluctuations in the defaulted portfolio´s spread was 17 times larger than the normal observed daily spread changes. This has also been confirmed by two external parties, and has been characterized as a true ‘Black Swan’ event.”
Somebody call Nassim Taleb. Meanwhile, the aftermath of the NASDAQ default event played out in a lot of overtime for credit managers at a number of major US banks. The speed and suddenness of the event of default, which was apparently caused by excessive leverage in the trader's position, suggested that existing risk tolerances were at fault. But was there any way to even anticipate this market contortion? Did the NASDAQ surveillance function notice the over-leveraged position?
You can of course say that the NASDAQ did their job, closed out the position in good order and passed the hat to the surviving clearing members to replenish the default fund. But the reality of the situation is far more complicated. When NASDAQ liquidated the positions, for example, the exchange had to use its own capital to close the positions. Had multiple clearing members failed at the same time, the thinly capitalized NASDAQ would have been under water.
Not long ago the Research Department of the FRBNY posted a comment in the Liberty Street blog praising the manifold blessings of centralized clearing:
“First, central clearing allows trades to be netted across all CCP members, lowering net settlement exposures and thereby reducing counterparty credit risks. Second, CCPs employ loss-sharing mechanisms that spread the cost of a member default across all members, thereby lowering the burden of default on any one participant. Third, the clearing process at a CCP is transparent and uniform for all CCP members, which reduces the uncertainty over potential losses owing to counterparty default.”
Our colleague Nom de Plumber, a veteran of the world of credit and counterparty risk who appreciates a good bottle, begs to differ. He notes that the socialization of clearing-member counterparty losses “invites both adverse-selection and moral-hazard risks.”
He continues: “Low regulatory credit-risk capital requirements (2% to 4%, per US Basel Rule Section 35) for counterparty exposures to CCPs seem hardly to suffice for or even acknowledge such risks, which gap market movements can materially exacerbate.”
We agree with a number of our colleagues in the world of credit that while the Scandinavian default event last week was relatively small in terms of dollars, the large magnitude move in the position of a single trader forced the NASDAQ to use its own capital to fill the gap until the clearing members kicked in their share the following day. The loss sharing among clearing members is not instantaneous or equitable, thus the exchange must have capital of its own when a clearing member fails, otherwise the supposed benefits of CCP are an empty promise.
One credit manager at a large bank told The Institutional Risk Analyst that the default process “did not work as entirely hoped.” He notes that the NASDAQ is getting a lot of questions from clearing members about the margin process and the auction of the position. One trader says that the NASDAQ did not notify some clearing members of the default for several days.
“The lessons learned from securitizations and CDOs in terms of first loss waterfalls and the sequential failure of obligors do not seem to have translated into the world of centralized clearing," notes the veteran credit portfolio manager. "Are all clearing members equal? Are JPMorgan and the smallest clearing member that trades one product equal in terms of first loss? In this case, first loss was not properly assessed."
Nom de Plumber agrees and notes that the one-size-fits-all process of CCP margin calls may also be a weakness, especially if non-member counterparties inject idiosyncratic counterparty risks---which can then be socialized to member counterparties, perhaps spawning systemic risk. He observes that the recent NASDAQ counterparty default loss may be an example.
The FRBNY says that “CCPs employ loss-sharing mechanisms that spread the cost of a member default across all members, thereby lowering the burden of default on any one participant.” Perhaps. What the default of the NASDAQ member last week seems to illustrate is that a significant market move can take down a clearing member in an afternoon. Had the position been larger, a 10x standard deviation market move could have impaired the capital of other clearing members as well.
Regulators take great comfort in the idea of centralized clearing, but a thinly capitalized exchange and clearing members may not provide much surety against contagion regardless of how the mutual clearing regime is labeled. The bottom line is that there is not much different between how centralized clearing exists today and the mutual exchanges that existed prior to the creation of the Federal Reserve System in 1913. As former Fed Chairman Ben Bernanke has noted in his research, sometimes exchanges die (see below).
Further Reading