Don't Blame the Symptom

In the latest witch hunt for who to blame for the banking fiasco, we are seeing money market funds (MMF) being blamed for "draining deposits" from the banks.

So why are MMF parking their money with the Fed via the Reverse Repo Facility?

The answer (if we're willing to be honest) lies at the intersection of Government actions and regulatory changes.

First, we must understand that Regulatory changes post Dodd-Frank have set hard restrictions (caps) on the size of bank balance sheets. The largest banks are restricted by something called the Supplemental Leverage Ratio (SLR). The below graphic describes the SLR in detail, but the main point is that SLR looks at Tier 1 capital vs all balance sheet items with no risk adjustment. The important takeaway here is: both Treasuries and reserves are included in the denominator.

In other words, the more reserves a bank holds the lower their SLR (the higher the SLR the "better").

Second, we must understand what Quantitative Easing (QE) is and isn't. QE does not "print" money. What it does is change the composition of money in the market AND influences where that money is held.

Here's a video clip of me explaining QE:

The Fed swaps bonds (puts bonds/bills on its balance sheet) and places reserves into the banking system. This swap takes bonds from all sorts of balance sheets and puts reserves into one balance sheet (banks).

The key takeaway here is that the Fed balance sheet expansion forces bank balance sheet expansion. The problem? Reserves count towards SLR.

Therefore, while there is no limit in terms of Fed balance sheet expansion (in theory) there is a limit to bank balance sheet expansion.

This is why the Fed suspended including Treasuries and reserves in the SLR calculation on April 1, 2020.

And "unsuspended" it on March 31, 2021.

The below graph shows the effects on deposits at both commercial banks [blue line] and money market mutual funds (MMF) [white line].

Note, as balance sheet constraints appeared, MMF deposits grew and then shrank as the SLR calculation suspended Treasury and Reserves from the calculation.

So the SLR calculation puts a restriction on how many deposits a commercial bank can hold. This pushes deposits to smaller banks (not subject to SLR) or to the shadow banking sector (MMF).

The second major factor that influences where MMF put their funds is a function of supply of available collateral (mostly T-bills), repo opportunities and bank deposit rates.

For more details on how Government regulatory changes impacted MMF and why 80%+ MMF deposits are in Government funds see here:

The main factor affecting bill supply is what the Treasury decides to do. Currently the Treasury is limited in issuance thanks to the debt ceiling. This means the account the Treasury keeps funds in (TGA) is being drained (adds reserves to banking system) BUT there is little investment opportunity for MMF to acquire bills. Since Government regulations pushed most money market deposits into Government MMF, this leaves very few alternatives.

If enough collateral is available (T-bills) MMF will park money in bills (and so will consumers directly). If enough is not available they (MMF) can redistribute funds into the commercial banking system. In other words, if banks pay a high enough market rate on deposits, MMF will re-deposit funds into the banking sector (which keeps reserves in the banking system) and so will consumers. If however, banks are balance sheet constrained then MMF will seek out other places to park cash.

Enter the Fed's Reverse Repo facility (RRP). For MMF needing a place to park cash, the Fed will borrow from entities via the RRP at a fixed rate of return (currently 4.8% note this rate is always below the interest on reserve balances IORB the Fed pays banks, which is currently 4.9%. So the banks always have an advantage vs RRP). MMF have the ability to tap the RRP and they will do so when no better risk adjusted return is available (Note consumers can't tap the RRP directly, so they use MMF as an intermediary).

Below is a graph showing how the RRP usage fluctuates relative to SLR. Note the spike in Apr 2020 (SLR suspension begins) and then the spike again in Apr 2021 (SLR suspension ends).

We can also layer on T-bill supply to see RRP behavior:

Note, as T-bill supply shrinks (green line above) and SLR suspension ends (APR 2021), there is a large spike in RRP usage.

Also note, many industry trade groups and Wall Street analysts called for an indefinite suspension of Treasury and Reserves from the SLR calculation, predicting this exact behavior. For example, here is an article from SIFMA:

Note that as MMF decides to park money at the RRP, it essentially mimics quantitative tightening (QT) by draining reserves from the banking system.

While some may fret that MMF are influencing QT, please note that the behavior of MMF is 100% predictable and preventable IF market forces are allowed to work. However, if QE causes balance sheets to reach maximum capacity (SLR) and small banks reach capacity via other leverage ratios, then the natural behavior of MMF will be to seek out alternative places to park cash. If the Treasury is not issuing bills (or not enough bills), then MMF will seek out the RRP.

I should also note that Basel accounting rules heavily penalize Repo transactions (where MMF lend via banks to end borrowers) and, therefore, balance sheet constraints also limit this mechanism of converting bonds/bills into cash (a major problem with our current financial plumbing and a factor in the 2019 Repo Crisis).

Therefore, the MMF behavior is a symptom of the current issues facing the banking sector. We should not blame the symptom (or shoot the messenger) for behaving exactly how the regulatory system was constructed.

Since we now understand why MMF's are behaving this way, here is how we can unwind the RRP usage, if desired.

The Richmond Fed ran regression analysis on the factors most influencing RRP behavior, the below table summarized their conclusions:

The clear path to reducing the RRP usage is addressing the debt ceiling and having the Treasury issue more bills. This would effectively reduce RRP usage by 96% (extending SLR relief would reduce MMF usage of RRP by 33%).

In conclusion, QE causes the Fed's balance sheet to swell which, in turn, causes bank balance sheets to swell. Unlike the Fed, bank's have a finite limit on balance sheet size (imposed by the regulators). This limit forces reserves and deposits to flow in the direction of the least constrained entities (small banks and MMF). If bank's are balance sheet constrained, they can't attract the MMF deposits back and can't facilitate borrowing via repo. Therefore, MMF will seek out alternatives. The first alternative will be T-bills. If the Treasury is not issuing/cannot issue bills, then the last remaining alternative is to find the only balance sheet that is not constrained: the Fed's via RRP.

The most effective solution to unwinding the RRP is to refill the TGA by Treasury issuing securities. While issuing securities drains reserves, if the money is already at the RRP, it is already outside the commercial banking system. The one potential issue with the Treasury issuance, is anything outside bills (that the MMF community can't take down) will require bank balance sheets (primary dealers). Therefore, it may be required that the SLR relief be extended, Basel repo rules are improved, and/or that the Fed will be forced to do QE (since it is the only non-constrained balance sheet), which will add reserves, which may push deposits to MMF, and in the absence of bank balance sheets will cause RRP usage to spike.

With the debt ceiling "X date" approaching soon, we will certainly get a chance to see all of the above in action. The only question is, have we learned anything this time?