JPMorgan: We Are Fast Approaching The Point Where Banks Run Out Of Liquidity

Last week we first noted that something unexpected has been going on in overnight funding markets: ever since March 20, the Effective Fed Funds rate has been trading above the IOER. This was unexpected for the simple reason that it is not supposed to happen by definition.

As a reminder, ever since the financial crisis, in order to push the effective fed funds rate above zero at a time of trillions in excess reserves, the Fed was compelled to create a corridor system for the fed funds rate which was bound on the bottom and top by two specific rates controlled by the Federal Reserve: the corridor "floor" was the overnight reverse repurchase rate (ON-RRP) which usually coincides with the lower bound of the fed funds rate, while on top, the effective fed funds rate is bound by the rate the Fed pays on Excess Reserves (IOER), i.e., the corridor "ceiling."

Or at least that's the theory. In practice, the effective FF tends to occasionally diverge from this corridor, and when it does, it prompts fears that the Fed is losing control over the most important instrument available to it: the price of money, which is set via the fed funds rate. And ever since March 20, this fear is front and center because as shown in the chart below, starting on March 20, the effective Fed Funds rate rose above the IOER first by just 1 basis point, and then, last Friday spiked as much as 4 bps above IOER.

To explain this bizarre phenomenon in which the EFF has been trading well above IOER in defiance of all of the Fed's monetary orthodoxy, we laid out several possible explanations, including that i) money market outflows around the April 15 tax deadline date and elevated GC repo rates; ii) the continued decline in excess reserves, and most ominously iii) another acute dollar shortage  developing across the US banking system.

One day later, PrismFP picked up on this topic and elaborated on the third, and most notable point, concluding that "there has been a dollar funding/shortage issue brewing under our nose for months; it is just coming to fruition now because we are noticing the DXY breaking out higher. In other words, with FHLB’s selling less FF’s, participants are forced to pay a higher rate to find funds, and that drives rate differentials towards the Dollar." Some other notable observations from his latest note which we laid out last week:

The next question becomes what will the Fed’s reaction function will be? There has to be some hand wringing inside 33 Liberty Street these days. The brewing speculation in the market is the Fed could cut the IOER rate in an effort to bring down the FF rate as well, something the Fed has indicated it could do previously (Nov minutes).

Now this would simply be a “technical adjustment” in the mind of Fed officials but in reality the signaling here is important. Is the market, who sees this developing Dollar funding/shortage issue developing, really going to take this only as a one-off adjustment? Or is the market going to assume this small cut would be the first in the Fed’s rate cut cycle?

Now it's the turn of everyone's JPMorgan contrarian, Nick Panigirtzoglou - the author of the popular "Flows and Liquidity" report, and a lone skeptical voice amid an otherwise permabullish landscape dominated by Marko Kolanovic - to warn that despite some $1.4 trillion in excess reserves sloshing around, "the liquidity conditions in the US banking system are perhaps close to decade lows" which in turn is manifesting itself in the breakout of the effective Fed Funds rate above the IOER, which as Morgan Stanley suggested last week, the Fed may have no choice but to cut by another 5 bps at the next Fed meeting just to "normalize" the fed funds rate and restore some temporary control to the most important interest rate in the world.

In his latest weekly note, JPM's flows strategist recaps what we said previously, noting that "the liquidity effects from the Fed’s  balance sheet resurfaced over the past week following a spike in overnight interbank rates. While a previous spike in both the median and 75th volume-weighted percentiles of the Fed funds rate to 3bp above the Interest rate on Excess Reserves (IOER) around quarter-end was quickly unwound, these rates spiked again over the past week to 4bp and 5bp above IOER creating a more persistent and concerning up move."

In turn, this persistent move higher in is raising questions about whether reserves - which as a reminder have been shrinking rapidly every since the Fed started to rolloff its massive balance sheet in late 2017 - "are close to or in tight territory", which also explains why the Fed recently reversed aggressively on its prior "autopilot" posture vis-a-vis the balance sheet and unexpectedly informed markets that the shrinkage would end in September.

But before we get into the implications, first the mechanics and why the most important plumbing in the US financial system appears to be clogged.

As Panigirtzoglou explains, echoing the observations from both BofA and Barclays, this spike in interbank rates in April was  accompanied by a drop in the reserve balances held by banks at the Federal Reserve, i.e. the "Tax day" excuse. April tax collections boosted the General Account of the US Treasury at the Fed by $124bn in the week ending April 17th and by another $30bn in the week ending April 24th. The mirror image of that was that the reserve balances at the Fed declined sharply by $160bn in April to below $1.5tr.

Yet despite the resultant sharp decline in reserves in April offsetting the size of the increase in the Treasury General Account at the Fed, this only brought down reserve balances back to their downtrend since Quantitative Tightening (QT) started in the fourth quarter of 2017.

More ominously, assuming no more "temporary "swings in the Treasury General Account, more declines in reserve balances likely lie ahead as Fed’s QT continues until September.

But to JPM, the news in April was not the sharp decline in reserve balances - an observation various other analysts had made previously - but rather the rise in interbank rates relative to IOER that accompanied it, the same rise we suggested indicates the Fed has lost control of rates; a rise which incidentally was not only confined to the Fed funds rate which represents a modest component of interbank markets with a YTD volume of around $70bn per day. Other interbank rates, such as SOFR - a broad measure of the cost of borrowing cash overnight collateralized by Treasury securities that includes all trades in the Broad General  Collateral Rate plus bilateral Treasury repos - also increased.

Which brings us to the next point: why is this move in IOER notable, and what does it have to do with liquidity in the banking sector?

The answer is that the ongoing creep higher in the delta between the IOER and the effective fed funds rate is a direct function of excess reserves in the system. One can observe the sensitivity of the Fed funds rate to the level of reserves by plotting the difference of the effective Fed funds rate over the Fed’s target range lower bound to the level of reserves. The steepening of this relationship over the past month as the level of reserves moved below $1.5tr is consistent with the idea that the level of reserves are closer to entering tight territory. This is shown in Figure 5.

The tightening in the reserves space is also evident when comparing the current notional level of reserve balances to the size of the US banking system. This is shown in Figure 6 which depicts the reserve balances at the Federal Reserve divided by the total assets of US banks. Recently, this ratio fell below 9% for the first time since the beginning of 2011, and stood at a similar level as far back as the end of 2009. In other words, as JPM puts it, the liquidity conditions in the US banking system are perhaps close to their tightest in a decade.

Given that the reserve regime the Fed currently operates is different from previous cycles, there is a great deal of  uncertainty over the point where we enter this tight territory, and JPMorgan's interpretation of the above figures is that "we are approaching that inflection point."

Which then brings us to the next key point: what is the relationship between the amount of reserves in the system and the Fed Funds rate? In theory, when the level of reserves in the system is more than generous, i.e. above $2 trillion or so, there is no direct relationship. However, as the quantity of reserves declines, e.g. to somewhere between $1.5tr-$2tr based on Figure 5, regulatory constraints are less binding for some banks due to smaller balance sheets.

It gets even trickier when the quantity of reserves enters "tight" territory - which is used very loosely here, considering that before the financial crisis, there was virtually no excess reserves in the system, as banks largely resorted to the Discount Window for emergency liquidity, JPM cautions that some banks might need to resort to the Fed funds market to maintain their reserve balances and meet their liquidity needs as reserves are not uniformly distributed across the banking system. It is these banks might need to pay rates that are even higher than the IOER, which they clearly are doing now with the EFF at 2.44% while the IOER remains stuck at 2.40% for now.

This brings up an interesting tangent: prior and during the financial crisis, one of the alarming bank run catalysts to emerge was that usage of the Discount Window had become an effective death sentence of a bank: if and when the market sniffed out that a given financial institution was scrambling for liquidity, the shorts would slam the bank's securities while counterparties would either significantly tighten overnight funding terms, or withdraw them altogether, creating a toxic death spiral of liquidity that eventually led to insolvency. One can argue that it was the sudden collapse in peer liquidity provisioning that brought down both Bear and Lehman in the matter of days.

To be sure, the Fed learned its lesson, and perhaps the main reason that the US central bank flooded the entire banking system with trillions in excess reserves (besides paying banks interest on money the Fed had created out of thin air of course) was to ensure that not one bank would emerge as the weakest link in the financial system - as the failure of the weakest link would promptly unravel the entire financial system - and that nobody would be forced to resort to the Fed's traditional emergency funding mechanism, borrowing from the Discount Window.

Of course, once excess reserves shrink to the point where marginal liquidity demands by the least capitalized banks results in tremors in the effective funds rate, which has now been trading above its Fed-mandated corridor for the past month, the next catalyst in the downward spiral of scarce liquidity is one or more banks tapping the Discount Window. If, and when, that happens, hit every bid as the deja vu from the financial crisis, whose underlying causes still remain to this day as nothing has been fixed but merely all the symptoms were drowned in an ocean of excess Fed liquidity which is no longer sufficient, will slam the US financial system with the power of a financial neutron bomb.

Perhaps we are exaggerating, but then again perhaps not: as JPMorgan writes, its interpretation of the sharp moves in interbank rates that accompanied the reserve reduction in April "provides further evidence that we are approaching a point where reserves enter tight territory" and if this is correct, the additional balance sheet shrinkage by the Fed until the premature end of QT in September could cause further tightening, and lead to another "market event."

JPMorgan's solution: to avoid further tightening the Fed might be forced to either bring forward the termination date of its QT from September to June, or cut interest rates by lowering the IOER.

The third, and most troubling "solution" would be for the Fed to inject liquidity via open market operations... even though as JPMorgan admit, "some market participants might see such operations as equivalent to restarting QE."

In short, there is still some $1.4 trillion in excess reserve sloshing in the financial system as a remnant from the trillions in liquidity generated by the Fed but, drumroll, it is no longer enough to ensure that the US financial system won't suffer some "unexpected event."

In short: we are slowly but surely approaching that moment when the US financial system will yet again be the catalyst for the next crisis. Until then, keep a very close eye on total borrowings under the Fed's discount window: once it spikes from its 8 year slumber, all bets are off.