After so many wildly stupid stand-offs over the years, it’s easy to ignore the latest US debt ceiling brouhaha. But it looks like some people in markets are getting at least a little bit antsy.
FT Alphaville has been keeping a close eye on the cost of US credit-default swaps, which you can use to insure against the danger of the government failing to service its debts. Right now they are trading at about 80 basis points, up from about 10 bps at the start of the year.
This is hardly apocalyptic stuff, but it’s at least a little disconcerting. For context, that’s close to where longer-term Italian CDS prices are trading at right now, according to Refinitiv data. UK five-year CDS prices peaked at about 50 bps at the peak of last autumn’s market turmoil, and are about 22 bps now (the price of a one-year UK CDS is just 6.4 bps).
Moreover, Barclays notes that the amount of outstanding CDS on US debt has climbed from under $500mn at the start of the year to about $4.2bn at pixel time — the sharpest eight-week jump in net notional outstanding since at least 2011.
At the same time, trading volumes have rocketed. In the last eight weeks of 2022 only a total of $7mn of US CDS traded. This year, there has been $1.1bn traded each week on average, making it the fifth-most actively traded single-name CDS market in the world in 2023.
However, Barclays points out that while the price of a one-year US CDS contract is about the same as it was in past debt ceiling stand-offs, the implied probability of an actual US government default is much lower this time.
That’s because the price of the cheapest-to-deliver Treasury bond in any CDS settlement auction is much lower today than it was in previous debt ceiling showdowns. Our emphasis below:
This has implications both for the potential payout ratio and the implied probability of a credit event. Assuming that bond prices remain unchanged in case of a credit event (see the discussion below) and that 1y CDS spreads represent the cost of owning protection (while recognising that the cost could be lower if the contract were to trigger prior to maturity or if there is less than 1y remaining on the contract), we can calculate the payout ratio as (100-CTD)/1y CDS. Currently, the CTD has a price of roughly 56 and the 1y CDS is at 80bp. This gives a payout ratio of 55x: the potential gain is 55x that of the cost to enter into the trade.
In turn, we can calculate the implied probability of a credit event as 1y CDS/ (100-CTD), which currently is 1.8%. Both of these numbers are vastly different from previous episodes, as the CTD has a much lower price.
The implied probability of a credit event in 2011 and 2013 was around 5%, but currently, even though the CDS spread is similar, the implied probability of an event is only 1.8%. As such, the CDS spread is actually reasonably tight after adjusting for the potential payout in case of a credit event.
We still suspect that this is mostly more of a short-term tactical trade — ie, a bet that the US will get dangerously close to the point where it runs out of money before the Republicans blink, and there’s therefore money to be made as CDS prices keep edging up closer to the drop-dead date (sometime in July-September is the CBO’s estimate).
Even if the Republicans refuse to lift the debt ceiling by then, the Treasury will almost certainly still continue to service its existing debts, slash back all other outgoing cheques and make the GOP wear the political fallout. Or they could just #MintTheCoin.
But for fun, let’s assume this really is the point where no one blinks. What are the mechanics around US CDS contracts? Handily, Barclays has also written a great Q&A on the subject:
How would the USA CDS contract be triggered?
Valid credit events for a [Standard Western European Sovereign] contract are Failure to Pay, Repudiation/Moratorium, and Restructuring. For USA CDS, the market has been focused on whether a technical Failure to Pay credit event could occur.
A Failure to Pay means the failure to make any payments on Borrowed Money obligations (bonds or loans) when due beyond any applicable grace period. US Treasuries do not specify a grace period, so in this case we believe that a three-business day grace period would apply as specified by the ISDA Credit Derivatives Definitions.
What about missed payments to Social Security, Medicare, or vendors? Could these trigger CDS?
We believe that only borrowed money obligations as specified above could trigger a Failure to Pay credit event.
Who determines if a credit event has occurred? And how is the final settlement price for CDS calculated?
A regional determinations committee (DC) consisting of up to ten sell-side and five buy-side firms would meet to determine whether a Failure to Pay credit event had occurred in accordance with the ISDA Credit Derivatives Definitions.
If they determined that a credit event had occurred (which would be posted on their website3), they would then set an auction date and publish a list of deliverables (investors would have an opportunity to provide feedback on the list before it is finalised). The DC would also publish the specific details governing the auction process (including the day for picking the FX rate for the auction) in the Auction Settlement Terms (ASTs). The auction process would then be used to determine the final settlement price for CDS.
For more details on the CDS auction process, see CDS Auction Basics.
What types of obligations would be deliverable into USA CDS?
As mentioned above, the DC decides which obligations will be deliverable. While not a legal opinion (and investors should do their own due diligence), based on our interpretation of the Definitions, we believe that Treasury bills, notes, and bonds would be deliverable into USA CDS. We also believe that the current cheapest-to-deliver bond among Treasuries is the 1.25 of 5/15/50 (ISIN US912810SN90), which was trading at $55.7 on March 6, followed by the 1.375 of 8/15/50 (ISIN US912810SP49), which was trading at $57.5.
Would any other debt be deliverable?
In addition to US Treasuries, debt obligations of US government agencies that benefit from a ‘Qualifying Guarantee’ (QG) would be deliverable, as per the ISDA definitions. The QG requirement is generally very restrictive, excluding any guarantee that is capable of being released in any other way than by payment. Entities that carry (implicit or explicit) guarantees but do not satisfy the strict QG requirements would not be deliverable obligations. With that in mind, entities such as Fannie Mae and Freddie Mac, which benefit from implicit guarantees, are unlikely to be deliverable.
What factors could affect the final settlement price for USA CDS?
One consideration for investors who just buy USA CDS protection (and do not own the cheapest-to-deliver obligation) is what might happen to the CDS payoff in the event that rates rally sharply prior to the CDS auction. It is worth remembering that rates rallied sharply following the downgrade of the USA sovereign rating by S&P in August 2011. If a technical default triggered a similar reaction (i.e., a flight to quality), recovery could be higher, in which case the CDS payoff would be lower than implied by the current price of the CTD bond.
What did Treasury and Fed officials discuss heading into prior such episodes?
The minutes of the 11 August, 2011 FOMC conference call showed that Fed officials discussed the procedures that were put in place to handle government payments, if the debt limit was not increased in a timely fashion. 4 These procedures were based on three principles which were reiterated in the October 16, 2013 FOMC conference call. 5
Specifically, the 2013 minutes noted that “Those principles [discussed in 2011] largely continue to apply today based on current discussions with Treasury staff. The first one is that principal and interest payments on Treasury securities would continue to be made on time. The second principle is that the Treasury would decide each day whether to make or delay other government payments. The third principle is that any payments made would settle as usual.”
The minutes added that “In terms of principal and interest payments, principal payments on maturing Treasury securities would be funded by Treasury auctions that roll over the maturing securities into new issues, so the new issues would fund the redemption of the maturing securities. Interest would be paid based on available cash in the Treasury general account. To make a coupon payment, however, the Treasury may need to delay or hold back making other government payments, even if it had sufficient balances on a given day, in order to accumulate sufficient funds to pay a future large coupon payment.”
The minutes further added that “While not expected based on current discussions, if the Treasury decided not to prioritise principal and interest payments and had insufficient balances to redeem maturing securities, it would instruct the Reserve Banks to roll forward in one-day increments the maturity date of maturing securities. The securities would then continue to be transferable on the Fedwire® Securities system. Interest payments would also be held within the system until the Treasury authorised their release. When the Treasury could make its payments, it would pay principal and interest in a manner that market participants have indicated would be the least disruptive to their operations.”
More recently Treasury Secretary Yellen noted that “You should not assume it’s operational and feasible to prioritise . . . Even the prioritisation of interest in debt I think it’s fair to say is not a foolproof way by any stretch of the imagination for avoiding economic and financial bedlam.”
Could a roll forward of the maturity date trigger a restructuring credit event?
We do not think so. In a separate document that was prepared by the NY Fed’s Treasury Market Practices Group in December 2013, it was explicitly stated that “ . . . the process of rolling the operational maturity date forward would not change the actual maturity date specified in the terms and conditions of the Treasury Offering Circular…” As such, we do not believe that this would be considered a restructuring event if it were to come to pass, but we believe that a Failure to Pay determination would be based on the original maturity date (and the three-business day grace period).