An article in efinancialnews.com by Anish Puaar called “Safety of collateral questioned after US Woes”, published Nov. 4th, makes some interesting points about the knock-on effects of a US government default on the use of UST collateral in OTC derivatives markets. We unpack some of what the article says.
So how did the collateral market react to the government shut-down and prospect for default? The article said,
“…The failure of the US government to come to a timely agreement on a debt package caused the yields paid out on US Treasuries to spike. Major clearing houses responded by slashing the value of US Treasuries held as collateral, leaving practitioners scrambling to cover the resulting shortfall…”
and
“…In the days leading up to the US’ debt ceiling deadline on October 17, the domestic Hong Kong clearing house devalued short-term US Treasuries, by increasing the percentage it subtracts from the market value of assets received as collateral, known as a haircut…”
and
“… the CME Group’s clearing house called for a total 12% increase to the collateral for all interest rate swap portfolios, although this was withdrawn once a deal was reached…”
These are all consistent with reasonable risk management procedures. If the collateral is illiquid or devalued, haircuts need to go up. US Treasuries do not have cross default, so longer dated paper would not automatically default should a T-bill fail to be repaid. Even short dated paper did not suffer price declines of any consequence. Sure, 1-month bills spiked, but the PV01 on a 1-month T-bill is pretty small. A 30bp increase in yield on a 1 month bill is $250 per $1 million. This was not the market saying the Treasury was about to default. Rather the message was there was a timing issue heaped on an optics problem. Investors didn’t want to have to explain holding the paper, so they sold. Had there been a real expectation that the US government was going through anything beyond a technical default, those T-bills would have dropped a lot more and started to trade on price instead of yield.
The article said,
“…To avoid similar situations in future, and potentially next February, market participants could look to rely more on other assets accepted by clearing houses. In preparation for the greater collateral demand, clearing houses have broadened the assets they accept to include gold and, in some cases, equities…”
Does this mean we are headed for a wholesale broadening out of what constitutes acceptable collateral? We think not. Where will the liquidity come from for that collateral, especially in a stressed environment? DCOs like the CME have back-up liquidity lines in place to quickly monetize corporate paper they take as collateral. This protects them against liquidity issues by shifting the risk to the banks. But these facilities are not cheap and there is a limit on their size (to say nothing of potential contagion issues that arise out of CCPs shifting liquidity risk to banks on a broad scale). If it is a facility for non-HQLA, the structure will hurt LCR, making the banks think twice about signing up. In a world with much broader collateral eligibility, those liquidity facilities would have to proliferate. Short of the central banks helping out, we don’t see that happening.
Those committed facilities may not be limited to corporates and other less liquid assets. We wrote about the potential need for committed facilities for US Treasuries in our post “The CFTC and liquidity rules for DCOs: do they need committed repo facilities for their US Treasuries?” This could be really messy.
Does all this end up increasing the risk of collateral shortages above and beyond the usual reasons? That might be a bridge too far for us. Another close encounter with a government default will cause dislocations in short dated paper (as collateral receivers move to avoid it), increase volatility, and certainly raise haircuts on collateral. But it is hard to imagine US Treasuries becoming ineligible.