There is a battle brewing between the CFTC and DCOs over collateral liquidity. And its impact on repo will potentially be enormous.
CFTC 39.33 (c)(3)(i) is a proposed rule that requires DCOs to hold qualifying liquid resources as collateral. This is defined, according to CPSS-ISOCO Principles for Financial Market Infrastructures (April 2012) as
“…For the purpose of meeting its minimum liquid resource requirement, an FMI’s qualifying liquid resources in each currency include cash at the central bank of issue and at creditworthy commercial banks, committed lines of credit, committed foreign exchange swaps, and committed repos, as well as highly marketable collateral held in custody and investments that are readily available and convertible into cash with prearranged and highly reliable funding arrangements, even in extreme but plausible market conditions…”
The rub is the term “committed repos”. Are these informal agreements in place to repo out paper to banks or formal committed facilities? There is a big difference.
“…we must express our unease about the informal conversations on liquidity that we’ve participated in with the Commission and the Federal Reserve Bank in which doubts have been expressed by both regulators about whether US Treasury securities can be countered as “qualifying” liquid resources where subject to uncommitted repurchase agreements…”
In other words, can CME and other DCOs rely on the availability of the banks to always be there to provide liquidity against US Treasuries or do they have to go out and pay for a committed facility to insure the ability to monetize US Treasuries via a repo? The shenanigans in Washington that may lead to a US government default does give one pause about the need for liquidity facilities for US Treasuries. But lets, for the time being, get beyond the irony and assume that US government debt is and will remain HQLA. We should note that the CME already does have facilities in place to guarantee the ability to monetize corporate bonds they take in as collateral. This is what, we suspect, allows them to widen out acceptable collateral to include corporates. But we digress.
The cost of setting up committed facilities to guaranty liquidity will be high. These facilities absorb capital, drawn or not. The CME, in their response letter to the CFTC noted that obtaining a facility in excess of $10 billion would be “…difficult to obtain…. even if we were to pay above market rates…”. The CME estimated the cost for the facility to be 30bp (pre-UST default jitters and UST CDS protection going through the roof), all of which would be passed along to end-users.
Should increases in capital ratios that would (in Fed Governor’s Stein’s words) “bind” the repo market come into place, those costs could go up from 30 bp to who-knows-what. That will be a lot of balance sheet put to use. If the Fed’s fear is that capacity in the repo market could be severely reduced by shifting the repo market’s throttle to Enhanced Supplementary Leverage Ratio from RWA, then arguing for facilities to insure liquidity might make sense, albeit in a twisted way. It is hard to believe the Fed was connecting those dots.
In yet another bit or irony, those committed facilities would be subject to LCR, causing the banks to hold HQLA against the possible cash outflow. Those HQLA would be…you guessed it…US Treasuries. (Correction: the asset that the facility would provide liquidity against can offset LCR. So if it was a liquidity facility for HQLA, then there is no LCR impact. But if it is for, say, corporate bonds, then it depends on the LCR treatment of that particular asset including haircuts and other limits.)
At stake is retaining a DCOs designation as a Qualified CCP (QCCP). Without that designation, the favorable capital treatment that comes with dealing with CCPs versus bilateral trades will go away. While the CME’s comment letter was the most vociferous about the committed facility issue (some simply said they encourage the CFTC to follow the CPSS-IOSCO rules, others didn’t mention it at all) we have heard others talk about the impact of committed facilities for US Treasuries too.
If the repo market is feeling like the regulators are playing a big game of “pile on”, it is probably with good reason. Banks simply won’t write committed facilities – even if they can get paid a lot of money for them – without thinking about how this will impact their potential repo exposure. Rightly so, credit officers will look at what could happen if the facilities are drawn. But in all fairness, uncommitted is just what is sounds like: there is no legal obligation to do anything.
And to what end will all this serve? Will it actually reduce systemic risk or just push it around? The optics might be nice, but this seems pretty silly.
For more on where CCP liquidity will come from, see our post from last week, “Where will all that required CCP and resolution authority liquidity actually come from?“.