Matt Levine / Money Stuff: Ugh! The debt ceiling...

Old_Joe

For many, many, many purposes in financial markets, you have to post collateral. Most importantly, if you want to borrow money in some of the biggest lending markets, you put up some collateral and get the money; when you repay the money you get the collateral back. If you do certain derivatives trades, you have to post collateral to ensure that you are good for your obligations. If you want to sell a stock short, you borrow the stock and post collateral to secure your obligation to return it. If you are a retail brokerage, your customers do trades today but actually hand over the money on Friday, and meanwhile you post collateral with a clearinghouse to guarantee their trades. The plumbing of global finance works through collateral: You and I can agree to do stuff in the future, without necessarily knowing each other well or trusting each other much, because we have posted collateral to ensure that we’re good for our promises.

Often the way this works is that the collateral consists of some securities that you own, and the safer the securities are — the more likely they are to retain their value — the better they are as collateral. The best collateral, for most things in US finance, is short-term US Treasury bills, which have no credit risk and very little interest-rate risk and are pretty much always worth a pretty predictable amount. Longer-term Treasury bonds are also good, though they have more rate risk. Some other kinds of collateral — agency bonds, municipal bonds, highly rated corporate and structured-finance bonds, etc. — are also quite good and acceptable for many purposes, though not all; some sorts of collateral-demanding businesses are very picky and will accept only Treasuries. And then there is a lot of other stuff: Junk bonds, penny stocks, private-company stocks, fractional ownership of racehorses, any sort of financial asset you can name. So much crypto. All of this stuff could be collateral, for some purposes; somebody would probably lend you money against any of it. But if you bring it to the GCF repo market or the Fed’s discount window they will turn up their noses; they don’t take fractional racehorses there.

The point of the collateral, in a lot of the big markets, is to make things easy and efficient, so nobody has to think about risk. Somebody will lend you money against a fractional racehorse or your startup shares, but they will think about it for a while and make a particular underwriting decision; they will be in the specialized business of lending money against weird assets. Financial-markets plumbing is different; it is a big organized system that uses fungible collateral that everybody agrees is good and that nobody has to think about specifically. Huge piles of similar, safe assets — like the trillions of dollars of reliable and well-understood US Treasuries — work best; weird bespoke stuff doesn’t work at all.

And of course if you have some bonds that are in default, where the issuer is not current on its interest and principal payments, they would be quite bad collateral. Not that they’re worthless, necessarily — maybe the issuer will get its act together and start paying again — just that evaluating defaulted debt is a very specific skill. Many systems that demand collateral would not even consider defaulted bonds as collateral; defaulted bonds are not at all the sort of safe, fungible assets that work as collateral in deep and liquid markets

And so of course the question is, if Treasury bills are also defaulted bonds, what happens? Does everything break?

(Continued)

Old_Joe

I don’t know. Bloomberg’s Chris Anstey and Liz McCormick report:

Investment bank clients are peppering Wall Street with questions about what happens if the US Treasury in coming weeks runs out of cash and does the unthinkable — failing to make payments due on Treasury securities, the bedrock of the global financial system. …

One school of thought is that the impact might not be so damaging. After all, since the 2011 debt-limit crisis, market participants have worked out a process for dealing with the Treasury announcing that it couldn’t make an interest or principal payment.

But JPMorgan Chase & Co. Chief Executive Officer Jamie Dimon warned earlier this month that even going to the brink is dangerous, with unpredictable consequences.

“The closer you get to it, you will have panic,” he said in a May 11 interview with Bloomberg Television. “The other thing about markets is that, always remember, panic is the one thing that scares people — they take irrational decisions.”

And even a key group that helped to compile the emergency procedures, the Federal Reserve Bank of New York-sponsored Treasury Market Practices Group, has issued its own caution.

“While the practices contemplated in this document might, at the margin, reduce some of the negative consequences of an untimely payment on Treasury debt for Treasury market functioning, the TMPG believes the consequences of delaying payments would nonetheless be severe,” it said in its December 2021 gameplan.

And:

“We are likely to see localized dislocations in the event of missed payment,” if that were to happen, JPMorgan rates strategists, co-led by Jay Barry, wrote Friday in a Q&A for clients on a technical default.

RBC Capital Markets strategists, also writing Friday, said they “doubt” a downgrade would trigger any forced reallocation by fund managers away from Treasuries.

At the same time, RBC’s Blake Gwinn and Izaac Brook cautioned that the “back-office issues” of delaying payments “could very easily bleed into the front-office, causing disruptions to liquidity and market functioning.”

The TMPG noted that firms holding Treasuries in custody for other financial institutions tend to advance payments scheduled for those securities, and would need to sort how to proceed if those payments weren’t received from the Treasury on time.

Firms that offer pricing on Treasuries could run into challenges, “such as setting the price of a Treasury subject to a delayed payment to $0,” the group said.

Some market participants “might not be able to implement” the contingency plans, “and others could do so only with substantial manual intervention in their trading and settlement processes, which itself would pose significant operational risk,” the TMPG said.

Yesterday FT Alphaville published much of that JPMorgan rates strategy Q&A, which I would say is broadly sanguine about market plumbing. The first point is that, if the US government does default, that will probably cause the prices of Treasury bonds to go up, since a government debt default is a crisis and crises cause a flight to safety and the safest assets are, still, Treasuries:

This is certainly not our modal view, but in the unlikely event of a technical default, we think Treasury yields would decline and the curve would steepen. This seems unusual in the context of a default, but Treasuries have rallied into the latter stages of other serious debt ceiling debates in 2011 and 2013.

From first principles, if a US debt default does not reduce the value of Treasuries, then Treasuries should remain pretty useful for plumbing and collateral purposes. Of course very little about financial plumbing is derived directly from first principles, and if your computer has a switch that is like “IF bond is defaulted THEN don’t accept it as collateral,” then there are problems. But people have had years of debt-ceiling warnings to adjust their switches and one hopes they have things kind of right:

Treasury can, in principle, delay coupon or principal payment dates. If Treasury announces its intention to postpone a payment date in advance (the day before the payment is due), the security will remain in Fedwire, and would therefore still be transferable. …

If Treasury fails to notify investors of its intent to delay a principal payment due the following day by approximately 10:00 PM, the security in question will drop out of Fedwire, and such defaulted security will not be transferable. If (only) a coupon payment is missed, however, the underlying security is still in the system and remains transferable. ...

The status of Treasury collateral depends on the timing of Treasury’s notification of any delays in payments. If done in the timeframe discussed earlier, the security remains in Fedwire and is still transferable. As a result, it could in principal be used as collateral for repo and derivatives transactions, although possibly with higher haircuts.

If notification deadlines are not met, particularly for principal payments, that particular security is dropped out of the system and is no longer transferable, and as a result, cannot be used as collateral. It is possible that an OTC market may develop for securities that drop out of the system, but the likelihood of such an outcome is unclear at the present time.

Since Treasury securities do not have cross-default provisions, other Treasury securities that have not had a delayed/missed payment will remain transferable on Fedwire and can therefore continue to be used as collateral. ...

Under the US non-cleared margin requirements (NCMR) finalized by CFTC and prudential regulators, Treasury securities are considered eligible collateral even in the case of a missed payment. However, this is not the case under the UK and EU NCMR regimes. Thus, for any transactions facing counterparties in those regions, defaulted Treasury securities would be assigned zero collateral value, requiring the swap counterparty to substitute or post additional collateral. …

We believe the Federal Reserve will accept defaulted Treasuries as collateral at the discount window.

And so on. Money market funds, for instance, hold about $1 trillion of short-term Treasuries; “ultimately,” say JPMorgan, “we believe these funds would not be forced to liquidate Treasury securities in a technical default.”

I want to make a couple of points here:

• 1) I assume that they are basically correct not to be too worried about the plumbing. We have been having debt-ceiling crises every few years for ages now, and surely everyone has war-gamed this out over and over again. Financial markets are not full of idiots, and it would be annoying if this extremely predictable and predicted event brings down the global financial system through some technicality.

• 2) That said, I assume that with, like, 85% confidence. There is a lot of stuff out there. Surely the biggest global banks and asset managers have gamed out how they will keep markets going in the event of a US default, but is there some smaller firm whose computers will say “Treasury price = $0” and cause chaos? Maybe!

• 3) If you work in some corner of financial plumbing that you think won’t work in the event of a default, please do let me know! Send me an email. Also, though, fix it? You still have a little bit of time, and you’ve had plenty of warning.

• 4) Wouldn’t it be so tiresome to work in financial plumbing at some big bank and have to go to all the meetings about this stuff? To have to build all the systems to deal with a US government default, just because the US government can never get its act together to get rid of the debt ceiling, and because debt-ceiling negotiations always have to go to the last second? Like imagine pulling the all-nighters at JPMorgan to prepare for this, scrambling to save the US government from the consequences of its own incompetence and malice, and meanwhile the Securities and Exchange Commission is fining because sometimes you text your colleagues about work. Just pay your debts, come on.