The roaring dot-com era of the late nineties created all kinds of headaches for the Federal Reserve. While its leader, the “maestro” Alan Greenspan, was referring specifically to confusion over money supply definitions when he uttered “irrational exuberance” in late ’96, a few months later the fruits of that exuberance were causing headaches down in the trenches of his Open Market Desk. On an otherwise nondescript day in April ’97, it seemed as if the central bank was losing control of its key monetary policy lever in federal funds.
More than that, the problem turned out to be the US Treasury.
When it all went down, at the FOMC meeting in May 1997 which followed the event, the episode had Open Market Desk manager Peter Fisher dreaming of overseas dollar swaps in order to get around a “paucity of collateral.”
“MR. FISHER. The first time we had a big miss in propositions, it did occur to me that it might be nice to be able to do foreign exchange swaps to inject reserves when there was a paucity of collateral.”
Originally, fed funds was the place banks with surplus reserves lent overnight to others facing short run deficits. That has always been the case, though participation here expanded greatly during the fifties. By its very name, federal funds, the “cash” being traded wasn’t physical stacks of Federal Reserve notes but rather electronic balances of reserves plied through the central bank’s balance sheet actions.
Thus, the Open Market Desk in New York was the mechanism by which central bankers kept the federal funds market within their specific tolerances; whatever the FOMC in Washington decided was the “best” interest rate in order to achieve the economic objectives of full employment and stable prices.
A marketplace is a messy place, therefore at times it’s not so straight and easy as the textbooks make it sound (the Fed says, the market obeys). To begin with, depository banks (and only depositories; before GFC1 in 2008, there were two separate classes of banking institutions, the other called commercial banks which were securities broker dealers) had and have to manage several possibly competing mandates: from minimum reserve requirements to maintaining sufficient clearing balances.
Reserves in this sense have a functional use as it pertains to the first duty of any money: medium of exchange. In the modern monetary economy, physical cash has been relegated to small-time largely one-off transactions. Most of what goes on instead gets disbursed via large clearing networks, the modern (compared to, say, the 19th century and before) electronic payments system.
It is cumbersome, relatively expensive, and most of all inefficient to move large quantities of cash around any national economy of size. Attempting to do this would result in the need for excessive volumes of currency, much of which would be caught up idly “floating” around in between paying off claims.
Instead, electronic claims can be aggregated together for one bank, balanced against claims on or for others, and then netted out on a daily basis through a central clearing mechanism (one such like Fedwire) crediting and debiting ledgers at need. Participation in this clearing structure, however, requires only the maintenance of necessary clearing balances held in reserve – and for Fedwire, obviously bank reserves held at Federal Reserve branches qualify – for protection against overdrafts.
Banks have to balance between the possibilities of customers withdrawing “funds”, in an electronic sense, with the opposite of being deposited with more than anticipated (or lower withdrawals than expected). Should the former happen, the bank might then become impaired for its credit creation duties or even in how it might be able to further contribute to general commerce being slowed down for lack of liquid electronic dollars available to it.
That’s the ultimate point of money; to be a medium of free and efficient exchange so that commerce can go on unhindered by the lack of that medium. Payments processing is at the very center of that fundamental function.
In that sense, the Federal Reserve’s technical role is to ensure there is no shortage in this one clearing space. Should the banking system begin to run low on its liquid margins, too many customer withdrawals or something, then the central bank taps the Open Market Desk to supply bank reserves in Open Market Operations (OMO) by buying bonds from primary dealers, creating bank reserves in the process available first for specific use.
Presumably, given this potential dearth of electronic clearing margin (or insufficient levels for reserve requirements), primary dealers would become more willing to lend their OMO-created surplus reserves to other depositories in need of it so as to avoid any monetary shortage which might impede the banking system’s ability to keep the wheels of commerce fully greased.
The way in which the Fed and the dealers interact is a complex series of predictions and calculations, balance-checking during clearing periods, as well as, at certain times, just plain guessing.
One reason for unpredictable reserve behavior is the United States federal government. The bigger it gets, the larger its monetary footprint. When an American citizen pays taxes, mostly this becomes another electronic claim settled by Fedwire; the bank which houses the citizen’s deposit balance is deducted that tax payment which is then transferred into one or another deposit account held on behalf of Uncle Sam.
As discussed last week, the Treasury Department has been using commercial banks for holding much of its “cash” since World War I; the War Loan Deposit Accounts of the 1910’s and 1920’s were later transformed into something called the Treasury Tax & Loan program (TT&L). Like the old War Loan Deposits, the TT&L was designed so that the government wouldn’t strip the banking system of cash and reserves especially at points on the seasonal calendar when tax payments (received by the Treasury) were heaviest.
Any dollar, electronic or otherwise, paid to Treasury would be a dollar deducted from the banking system unless Treasury left that dollar with somewhere still inside it. In this case, the dollar rather than being cleared out of the depository system merely switches beneficial owner of that dollar and at which depository that ownership is custodied.
But most of the government’s day-to-day business runs through its main account managed by the Federal Reserve, something called the Treasury General Account (TGA). As such, the Treasury Department – like depository banks – seeks to maintain a minimum clearing balance from which ensure it doesn’t bounce any of its checks or electronic payments (including entitlements like Social Security).
Unlike the TT&L option, a dollar of taxes paid and directed instead to the TGA is one dollar that is removed from the depository system; it gets deducted from the taxpayer, then the taxpayer’s bank’s reserves and credited at the Fed to Treasury’s TGA; now a dollar outside this one specific ecosystem.
As a result, going back a century, the government has sought to maintain only what it might need, a minimal clearing balance, in the TGA while storing most of its liquid “cash” claims in certain depository banks, thereby minimizing the potential drain on systemic reserves as much as possible. The Federal Reserve, obviously, must take into account what the Treasury is doing in TGA and TT&L’s when factoring its management of system reserves.
Uncle Sam, though, has always demanded security from private banks holding its deposits. Going back again to the War Loan Deposit Accounts, depositories had always been required to put up sufficient eligible collateral in order to be and stay eligible to receive these Treasury credits. This means that there is a functional cost and constraint to participation.
April 15, 1997, was, of course, the unholiest of holidays: Income Tax Day. Unlike many – perhaps any – that had come before this one, suddenly the IRS had been flooded with imbursements due to previous unwithheld income tax payments (subsequently shown to be mostly related to individual capital gains tax liabilities bursting from all the paper dot-com profits). Both the Fed and Treasury were expecting a lot, but a lot more showed up.
As the flood of tax payments came in, commercial banks began turning the government away. Why? Banks simply didn’t have enough spare, unencumbered collateral to absorb these tens of billions in unanticipated tax receipts. As a necessary consequence, Treasury had no other option but transfer its windfall to the TGA.
Government policy had been to stick to a clearing minimum in the TGA of around $5 to $10 billion; only a few times had the balance swelled a small amount more than that upper estimate. On April 15, 1997, however, the TGA bulged out to $17.88 billion – before being drawn down as depositories found enough collateral over the coming days to take some of it back.
The payments kept coming, however, as mailed tax returns, and the checks or wire instructions included with them, were processed. By April 23, the TGA was back up to nearly $16 billion – and the federal funds effective rate (EFF) surged to 5.95%, 45 bps above the 5.50% target set by the FOMC (apart from quarter-end, EFF had generally been 5 to 10 bps below target in the weeks leading up to all this).
Alarmed, Peter Fisher’s Open Market Desk ramped up repos with primary dealers - on two occasions doubling up daily OMO’s - to keep the system supplied with reserves as the TGA built higher and higher. It would eventually peak at $52.22 billion on April 30, meaning that tens of billions had been unexpectedly drained from the depository system’s Fedwire electronic reserves.
The Open Market Desk’s cumulative “RP” OMO’s during this two-week stretch reached an astonishingly similar $52 billion by April’s end, too.
Having EFF rise as much as it did as quickly as it did even with OMO’s ongoing was a minor black mark against the Fed’s technical proficiencies; fortunately for the “maestro”, hardly anyone noticed having already become even more irrationally fixated on so much exuberant conditioning elsewhere.
Searching for answers, Peter Fisher found out that reserve forecasts may not have been too far off as initially thought, and that the answer to the “April surprise” in fed funds lay elsewhere. Sure, taxes had come in by the bushelful, and that had drained reserves, but in comparison to April ’96 it wasn’t so egregious as first thought.
What seems to have happened instead was the depository system reacted more intensely to the drain in ’97 than it had the year before (TGA got as high as $15.67 billion on April 17, 1996). In the accompanying briefing notes to the May 1997 FOMC policy meeting, Fisher wrote:
“Thus, our initial conclusion is that the market is somewhat more sensitive this year, compared to last year, which may be owing to the lower operating balances.”
The Fed’s Fedwire clearing system with its central bank-created bank reserves was not, and is not, the only monetary clearing system that banks use. In fact, going back to the earliest days of interbank clearing, the American correspondent system arose with no central bank around at all; a completely private payment network which, at times, had created and used its own form of interbank quasi-money, clearinghouse certificates, that functioned the same way as Federal Reserve bank reserves did in transferring and netting payment claims.
In fact, when we look at how the banking system operates, including foreign and commercial banks, not just domestic depositories, most of the time it does so outside of the Federal Reserve framework. Using the SWIFT system globally for US$ payment processing, for example, global banks get together creating temporary, ad hoc networks, called corridors, to process batches of international payment claims.
These are not settled by the transfer of bank reserves, and indeed it seems that the increasing scope of international eurodollars during these decades (eighties, nineties, right up to around 2011) is one factor we can identify for why the “lower operating balances” Fisher wrote up as a reason for “more sensitive” banks over bank reserves.
Having to maintain two (or more) minimum clearing balances separate from each other is an expensive proposition; thus, any firm would seek to most efficiently manage each one, meaning it would emphasize being able to hold the least amount possible in both segments. Doing so, however, would make anyone “more sensitive” to the potential for any interruption or shortfall.
This raises the immediate question as to what banks operating outside the official, domestic payment system might have been using for “reserves”; and it’s an inquiry that central bankers stopped trying to answer decades before the dot-com bonus started temporarily filling up Uncle Sam’s coffers.
Though the Asian Financial Crisis (otherwise the first regionwide dollar shortage testing the eurodollar system) came about later in ’97 and stuck around destructively through most of ’98, it wasn’t until 2007 that reserve hoarding would truly strike – both systems.
This has become a source of so much confusion when explaining what really went on during the first Global Financial Crisis. To begin with, during its worst parts this gross, global monetary panic saw federal funds rates plummet and remain low; EFF, beginning August 10, 2007, dropped way below the target rate.
In the aftermath of Lehman Brothers the following September, effective fed funds fell so low that the Treasury Department began moving nearly all its deposits out from TT&L accounts placing them into the TGA (the government was paid fed funds minus 25 bps, so as EFF came down closer to 25 bps Treasury wasn’t going to keep cash in depository banks only to end up paying the banks interest on its own funds). Of course, as these reserves were moved to the TGA, it drained the depository system of almost $100 billion by the end of October 2008.
EFF remained way, way below the target anyway.
Economists and central bankers (same thing) have tried to reconcile this fact by claiming the Federal Reserve had swelled its balance sheet with emergency liquidity provisions to the point that the depository system had become flush, or “abundant”, with bank reserves. Therefore, even though TGA took away that much, banks were left with so much more – and mostly via overseas dollar swaps (I don’t think Peter Fisher deserves the “credit”, though).
Yet, global panic came about and persisted anyway for a further five months.
In other words, domestic bank reserves were plentiful enough (along with exhibited counterparty risks favoring fewer unsecured borrowers) to keep federal funds down close to zero when the FOMC had targeted upwards of 1.75% (as of October 8) and higher (2.25% before then). Clearly, bank reserves and domestic fed funds, the collective reach of the Federal Reserve and the federal government Treasury, are not the exhaustive factors they are presented to be.
They explain federal funds, at times, but little else that might matter. Even Peter Fisher had conceded more than two decades ago, on the narrowest terms of his own technical domain, how collateral, for one thing, had been an extremely important constraint.
Over the coming months of 2021, the Treasury Department will be forced to drain the TGA (for several reasons) now swelled upwards of $1.6 trillion, thus reversing the direction when compared to what happened in April ’97. Any dollar pulled out (via spending or refunding debt) of this account then becomes available as reserves for the banking system, at least so far as the domestic Fed’s electronic clearing system might be concerned.
This has been decreed as highly inflationary. But, first, why hadn’t the unusual shortage of April 1997 been more deflationary than the trivial footnote it became? The brief event caused those inside the Fed more than a little consternation, but the wider global marketplace barely noticed apart from a couple days of high fed funds spreads; a little touchy about bank reserves, nothing at all for the rest of the monetary system.
Fisher, patting himself and his crew on the back, credited well-timed OMO’s and skillful reserve management.
Yet, the global marketplace, at least one key region of it, would instead become destructively preoccupied later in the same year by dollar conditions way outside of federal funds and bank reserves. And since those things occupy much larger space well beyond the Fed’s purview and authority, these were viewed solely as someone else’s problem (even when they hit home hard in ’07).
Bank reserves may not be the whole monetary story, indeed maybe not even enough of it. And while we admit incredible difficulty spelling out and defining what the offshore version of bank reserves might fit, in more than one category, we can always say for certain collateral sits right at the base of the whole global money pyramid. As always, a good place to start is T-bills.
