There’s just the one Treasury market, isn’t there? The same has been assumed of the money market, too. Singular blocks which operate something like a Medieval marketplace, only here bank dealers line its alleyways and hawk financial products under the watchful, experienced, sometimes vengeful eye of the local lord Maestro Greenspan. It makes for a compelling fairy tale, one to which common sense had surrendered a very long time ago.
Many still pine for those far simpler if fantasy days, to go back to believing that ideal technocrats easily controlled the entire world with the flick of a federal funds target. Not just technocrats but scientists who knew easily how to objectively calculate the exact right maneuvers to produce elongated economic nirvana (the Great “Moderation”).
Yet, even during this presumed Golden Age, underlying that whole time was the dirty, gross complexities of monetary sausage being made by hidden players. These were the dealers who did the actual work attributed to central banker overlords.
Among their primary activities – these are called primary dealers for this reason – was intermediation in the Treasury market. We all want the federal government to be able to dependably borrow on reasonable terms, even if most of us (outside MMT and anarchists) wish, just from time to time, Congress could restrain itself.
The role of primary dealers in this setting is as a storehouse or warehouse of securities. They have to buy what no one else does of whatever the government sells; that’s the obligation to accepting membership as one of this financial subset. Statutorily bound to participate and clean up every last scrap in this primary market.
There’s not a legal obligation in the secondary market, in which primary dealers are also a prime participant, yet a business commitment to very large customers (including foreign governments and central banks) no less binding. If Government Finance Minister A has commanded that its Reserve Manager X needs to liquidate US Treasuries in the nation’s reserve account post haste, the lucrative relationship any primary dealer has enjoyed serving the Reserve Manager suddenly becomes fraught with, shall we say, potential difficulties.
You might already ask, why wouldn’t the Reserve Manager simply liquidate the Treasury asset in “the market?” These are, after all, Treasuries and all we ever hear is that the market for them is the deepest and most liquid in human imagination. Yet, while true, functionally there is only a relatively smaller marketplace for easily, dependably, and efficiently disposing of really any kind of financial asset. The secondary market is no monolith.
In almost every case, what the Reserve Manager wishes to quickly push is an illiquid specific security – even if it is a Treasury note or bond - for which there is no ready quote, let alone a sea of buyers and sellers already lined up and trading the same specific asset. These are called off-the-run(OFR) Treasuries, and since there isn’t much if any regular trading in them our fictional Reserve Manager is essentially tasking its primary dealer to become that market.
Buy this thing from us at a reasonable price, or else!
Unwilling, likely unable, to upset a custodial relationship of such substantial potential revenue, the dealer dutifully obeys and takes the illiquid OFR asset off the Reserve Manager’s hands and crediting its account with some form of “cash”; which does deserve the quotation marks.
Where does our harried primary dealer get such “cash” in order to purchase this foreign-reserve OFR UST? It could, in theory, adopt several postures but for now we’ll assume it would stick to what is very common practice: repos, reverse repos, likely some securities lending/swaps/transformations and a literally insane use of badly outdated accounting provisions.
After taking the ticket from the Reserve Manager, the dealer now has an illiquid OFR UST on its balance sheet, asset side. The good thing going is that it can run a daylight overdraft to “pay” the Reserve Manager immediately, arranging maybe a whole day and beyond (if triparty repo’s the way to go) to figure out where this “cash” might come from.
In this standard practice, further assuming (quite realistically) the primary dealer can’t immediately find another buyer for the OFR asset, then it will arrange a repo transaction whereby the illiquid OFR UST is posted as collateral (plus a haircut) with some other dealer which has obtained excess “cash” from still another counterparty unseen (and for now unimportant).
However, while the cash piece has been taken care of that’s hardly the end of the matter. Since the incoming OFR wasn’t some specific investment of the dealer’s Treasury Desk or otherwise a conscious choice of regular internal business, this means involuntarily storing the asset on its balance sheet contains risk. The US CPI might surge and raise inflation expectations, leading to gross readjustment of Treasury prices especially for those at the yield curve’s longer ends (of course, this will have to remain hypothetical because even large CPI or PCE Deflator spikes don’t and won’t ever amount to much).
To protect itself from the (very remote, in practice) possibility of inflation risk, the primary dealer warehousing the OFR would hedge against it by short selling an on-the-run (OTR) UST security of the same characteristics (similar coupon rate and maturity, within reason). The reason for hedging with an OTR UST is that by hedging, the last thing the dealer wants is to be caught short an illiquid security and being forced to close-out at an unexpectedly unfavorable price (short squeeze).
This works (well) as a hedge simply because the perceived price of the OFR instrument will track closely the liquid price of the OTR hedge; should inflation risks rise and LT UST prices fall, being short the OTR hedge produces a gain (accounting for convexity; a further complication we don’t need to get into here) near enough offsetting any potential price loss in being long the OFR UST being warehoused.
Having sold short the hedge, the dealer now has both the obligation of delivering that security borrowed as well as the benefit of having raised cash after selling short (you can see why intraday timing could be an issue, arranging many moving parts even if I’m really simplifying things here). What to do with the cash, another warehouse item?
Reverse repo.
The term repo stands for repurchase agreement, a collateralized short-term loan viewed from the perspective of the cash borrower (the one posting security, or collateral). The reverse repo is simply, as the name already implies, the same transaction except viewed from the opposite or reverse perspective; the one doing the cash lending (and receiving collateral).
In our case, the dealer arranges a reverse repo with the cash obtained from the short sale accepting back another OTR UST (which might not have be the exact same) to close out this other obligation. The original unwanted OFR UST warehouse obligation which began this series of increasingly intricate events puts the dealer on both sides. Repos as well as reverse repos – and we aren’t done yet.
What would happen should Reserve Manager X need to liquidate a whole bunch of OFR UST’s from that country’s reserve account? Further, what if Reserve Manager Y from another nation as well as Reserve Manager Z from still another does the same? An overseas deluge of illiquid OFR UST’s results in these things being foisted upon primary dealers (plural) wholesale who are at least honor-bound if not contractually obligated (assuming backstop provisions and such) to make this off-market market in them.
Wherefore art thou repo?
As you can hopefully begin to imagine, imagine a clog in repo coming from both sides. To begin with, dealer repo.
Unable to arrange immediate secondary market buyers, our dealer cohort together is becoming stuck with unwanted and unexpected volumes of warehouse paper that, though of the UST variety, isn’t readily liquid. But a rash of dealers put into this same situation trying to access the repo market all at once might quickly dry up funds in the repo market (especially if dealers, many of them the same as in our example here, become skittish given a sudden surge in this demand; as might’ve been previously established in, say, September 2019).
What then? More than a few pocket what repo funding they can find leaving the rest to, gulp, contemplate if not actually liquidating these OFR UST’s into a secondary market which is no regular marketplace. Fire-sale situations develop; or, in most cases, distressed-sales which, like fire-sales, have the effect of depressing OFR prices.
Which, of course, means OFR UST’s become even less desirable for potential repo counterparties, closing the repo funding window that much more. As always, cash lenders who accept collateral (the reverse repo) care only about the liquidity characteristics of the collateral, even if ostensibly UST’s. In a distressed-sale deluge, price volatility drives reasonable terms - even what might have been otherwise offered by willing dealers and sources of funds - out of the marketplace, rendering the collateral increasingly useless as collateral.
As you might have already pieced together, I’m not actually describing a hypothetical situation. You probably recall that, during the worst days of last March’s GFC2, LT Treasury yields, by and large, went up rather than down during them as if some ridiculously powerful, unseen force had unleashed the ability to make safe assets go down in price at the very moment safe assets should’ve been (and were otherwise) in huge demand.
But what of any other dealers, perhaps the majority, those who couldn’t arrange increasingly scarce repo but also didn’t want to book any loss by being forced to sell in a distressed OFR UST market? There’s still the reverse repo option, only with several changes to its structure from the non-stressed version described above.
This still requires, of course, future delivery of those OTR securities which might be arranged instead by a mad scramble to mobilize other collateral options beyond any reverse repo. Perhaps a term collateral swap (transformation) with some non-panicky dealer (good luck finding one of those) who might accept lesser quality collateral in order to borrow the OTR stuff.
And at some point, it might not matter any equivalence; just find any OTR collateral you can get your hands on, however you can get your hands on it, cross your fingers and pray your daylight overdrafts get covered and you survive to yank your hair out tomorrow when you have to do all these things all over again. This changes your view on fundamental operational parameters.
On net: the Treasury market appears to break down when in fact it isn’t theTreasury market what’s failing at all. Because of these repo complications, as well as others not discussed here, massive OFR dysfunction simply herds more and more participants toward the only parts of “the” market still available: OTR. Bills, mostly. Remember what bill yields were doing last March.
It is all these things which were described by, of all groups of people, the Federal Reserve’s FOMC. On April 8, 2020, these policymakers released their sanitized meeting minutes from the earlier March policy meeting held toward the end of GFC2. Here’s the one never-cited passage which truly spills the beans:
“In the Treasury market, following several consecutive days of deteriorating conditions, market participants reported an acute decline in market liquidity. A number of primary dealers found it especially difficult to make markets in off-the-run Treasury securities and reported that this segment of the market had ceased to function effectively. This disruption in intermediation was attributed, in part, to sales of off-the-run Treasury securities and flight-to-quality flows into the most liquid, on-the-run Treasury securities.”
Exactly.
The FOMC, however, used and still uses this episode as an example of monetary policy done right! The conventional view has settled upon massive Treasury purchases under QE6 as the reason for righting the repo ship, for the Fed acting as the market-of-last-resort underneath all those OFR UST’s obligatorily bloating up dealer balance sheets and thereby clogging repo markets.
Of the few who notice such things, none have ever asked the only question worth asking: where the hell did all that selling come from in the first place? We are instead supposed to be dazzled by what the Fed did toward the end, rather than pay any attention to how all this came about – or, most importantly, how the fact that it did happen caused enormous harm to economic and financial functions (for a second time in a dozen years).
The lack of curiosity stems from an alarming and unappreciated change in how central banks think of what it is they’re supposed to be doing.
This began during the first GFC in late 2008; it used to be, should be a central bank was Bagehot-like in that its primary job was to make sure this kind of monetary disruption never happened. To lend freely at high rates and all that. Panic averted before it ever begins, or at least before it reaches its more destructive phases.
This is, by the way, the dogma which had prevailed up until, oh, December 2008. As Ben Bernanke had said to Milton Friedman, we won’t do it again (where “it” was allowing a full-scale monetary meltdown). Within a few years, full-scale monetary meltdown, one failed rescue after another before, toward GFC1’s finale, QE becoming an instrument designed to, they’ve said, keep the damage to as little as possible.
So, the modern central bank no longer prevents panics, it now sees itself as some kind of janitor who has to come by and clean up in its aftermath; the natural response to being forced to recognize how the monetary system doesn’t really have a place for it. You may be puzzled since there weren’t any Congressional hearings, nor public discussions, extolling this change in fundamental character. It slipped quietly underneath all these years covered by absurd fascination with abundant reserves.
Rather than figuring out why so much selling happens, either in 2020 or 2008, monetary policy in any emergency has transformed into buying as much of what is being sold, with all that selling for whatever reasons which will remain completely unknown. This is not what central banks do; the Fed is not a central bank.
The term which answers for all these is “global dollar shortage.” It has been what drives both foreign liquidations of reserve assets as well as creates additional difficulties in dealer balance sheets which contributes to both the original shortage developing as well as how it becomes self-reinforcing.
Balance sheets constrict (perceived volatility, some of which is derived from practical experience of being totally unable to count on central banks); global funding dries up in key parts, including repo; among other things, foreign reserve managers have to liquidate reserves to fund their local dollar short (bank commitments, in US$s); kicking off the OFR/OTR driven collateral bottleneck described above; which then further constricts global funding creating even more repo problems; and on and on.
In one last bit of sensational “coincidence”, a temporary way out of repo/OFR funding jail can be, as I wrote, reverse repo. Short of usable collateral? Can’t find your way out of being short an OTR UST? No securities lending dealers willing to rehypothecate you some bills? Take whatever cash and go to someone who has these things all locked up: the Fed.
And in the Fed’s case, this final leg would be a reverse repo. What this essentially becomes is the SOMA portfolio as an emergency and unwitting source of collateral transformation; the one thing that all this fungibility with collateral absolutely requires, more so during times of stress, to maintain monetary sufficiency.
For all these things, who cares about bank reserves. The Fed buying up OTR UST’s after-the-fact is an abomination and yet another example of monetary dereliction especially given the cultivated obsession with bank reserves as money.
In light of reverse repo usage at the Fed’s RRP window as of yesterday rising to nearly $500 billion, not that this signals imminent GFC3 but rather possible serious collateral strain along these same lines, I’ll leave you with still another overlooked bombshell this one from the FOMC’s last published minutes relating to its April 2021 meeting:
“…pressures in the markets for U.S. Treasury securities and Treasury repo that could spill over to other funding markets and impair the implementation and transmission of monetary policy. In this regard, a number of participants noted the potential for pressures in short-term funding markets to arise from time to time, even with monetary policy operating in an ample-reserves regime.” [emphasis added]
Inflation’s dead since 2007 with the Fed no more than an ineffective janitor coming in only after collateral, the real monetary element, makes such a global mess and keeps dealers harried and on edge forever after for how many times this happens, in big and small ways.
