Amid the prevailing market narrative of looming inflation concerns, the shortest part of the US Treasury Curve has been dominated by zero and negative rates. Surprisingly, the Federal Reserve’s actions suggest they are comfortable with these rates, at least for now. Below we explore the drama surrounding the expiration of a temporary Supplementary Leverage Ratio (SLR) exemption, how funding markets have shifted from being dominated by demand for reserves to demand for collateral, and how rates have turned negative in funding markets.
The New Year Brought Many Changes: More Reserves, Less Treasury Cash at the Fed, and Warnings of Negative Rates
The first quarter of 2021 brought many changes to the Federal Reserve’s balance sheet and the state of funding markets in the United States. Amid these changes, some of the lesser-known inner workings of the banking system and regulatory controls garnered attention and became fodder for media headlines. Chief among them was the Supplementary Leverage Ratio (SLR), a leverage requirement imposed on US Banks to satisfy the Basel III Leverage Ratio (LINK). The SLR applies to the eight US Bank holding companies that are deemed Global Systemically Important Banks (G-SIBs) and requires these banks to maintain a minimum leverage ratio. This ratio, the SLR, is calculated as the ratio of Tier 1 Capital to Total Leverage Exposure where Tier 1 Capital is core capital, mainly common stock and retained earnings, and Total Leverage Exposure, which takes into account both on- and off-balance sheet asset exposures, derivative exposures, etc. (LINK). In May 2020, in response to the onset of the COVID-19 pandemic and stresses in funding and credit markets, the Federal Reserve temporarily modified SLR requirements, excluding US Treasury securities and deposits at the Fed from SLR calculations. While this was done under the guise of better enabling banks to “provide credit to households and businesses in light of the challenges arising from the coronavirus response” (LINK), the lack of bank lending activity and ballooning reserves on G-SIB balance sheets suggest it may have also been to facilitate the massive stimulus outlays from the Federal Government. Regardless, the temporary SLR exemption was allowed to expire at the end of March 2021, and there were concerns—raised by Zoltan Pozsar at Credit Suisse and others (LINK, LINK, LINK, LINK)—that without a continuation of the exemption, G-SIBs would be forced to dump US Treasuries on the open market to meet SLR requirements. As shown in Figure 1, traditionally there has been an inverse relationship between primary dealer (mainly G-SIBs) holdings and reserves. The impact of the SLR exemption, shown in gray, is a clear deviation from trend.
Figure 1: Historically, G-SIB Treasury Holdings Inversely Track Reserves
Concerns of primary dealers selling US Treasuries were only enhanced by guidance from the US Treasury Department. At the beginning of February, the Treasury Borrowing Advisory Committee (TBAC) published their quarterly refunding report, in which they disclosed that the cash balance in the Treasury General Account (TGA) at the Fed, which ballooned following the onset of the COVID-19 pandemic, is forecast to drop substantially. The TBAC projected that the TGA would likely decline to ~$800B from the current level of ~$1.6T, which would result in further increases in reserve balances in order to balance the Fed’s balance sheet, as shown in Figure 2 below (LINK). Assuming the debt ceiling will likely come into effect in August 2021, it is likely the TGA will need to be reduced relative to current predictions. (For context, the TGA contained roughly $350B in December 2019.) Combined with the ongoing $120B of monthly asset purchases from the Fed, reserves are forecast to expand substantially.
Figure 2: Reduction of TGA and Ongoing Fed Asset Purchases Driving Reserves Higher
The problem with the continued, dramatic increases in reserves is absorption; the financial system must house them somewhere. G-SIB banks are usually the largest warehouses of reserves (see Figure 3).
Figure 3: Large Banks Absorb the Majority of Reserves
Banks are a primary mechanism relied upon by the Fed to stimulate economic growth. Reserves created by the Fed (money printing) are “sent” to commercial banks that theoretically lend the majority of these reserves, putting the newly created money into the economy. This lent money cycles through the economy and, in theory, a large part is eventually returned to the banking system, enabling additional lending. This cycle, known as the money multiplier effect, is a hallmark of the fractional reserve banking system. Prior to the impending expiration of the SLR exemption and other binding constraints including G-SIB scores, banks resuming share repurchases, and the ongoing Wells Fargo asset-growth ban, it appeared in early spring that G-SIB banks were quickly running out of balance sheet capacity to absorb the new reserves.
As Zoltan Pozsar at Credit Suisse clearly articulated (LINK), there was concern if banks were unable to absorb the new reserves, they could turn away non-operating deposits, sending the funds instead to money market funds. Money market funds would likely then try to park the funds at the Overnight Reverse Repurchase (ON RRP) facility at the Federal Reserve. However, this “temporary” facility offered by the Fed is capped, and at the time only allowed roughly $30B per fund. Amid rumors of G-SIBs already starting to turn away deposits, JP Morgan warned during their fourth quarter 2020 earnings call that there was a risk of implementing negative rates on deposits, which would effectively mean the bank would turn away deposits (see Figure 4). JP Morgan’s CFO stated, “We could simply shy away from taking new deposits, redirecting them elsewhere in the system, or we can issue or retain additional capital and pass on some of that cost, which is certainly something we wouldn’t want to do in this environment.” While this was ominous at the time, in hindsight it may have been partly a political maneuver. Senators Elizabeth Warren and Sherrod Brown were vocal in their opposition to extending the SLR exemption (LINK) and warned of the risks to the financial system; they pointed out that bank lending has been declining as a percent of assets, and expressed their shared view that banks should retain more capital, rather than provide returns to shareholders, in order to meet leverage ratio requirements. Ultimately the SLR exemption was not extended.
Figure 4: JP Morgan’s Warnings of Failure to Extend SLR Exemption
If G-SIB banks were to turn away operating deposits, a likely alternative for these deposits would be money market funds. But, as Credit Suisse argues (LINK, LINK, LINK, LINK), money market funds, which would normally direct these flows to the overnight reserve repurchase (ON RRP) facilities, would quickly hit the cap of $30B allowed per fund. This would likely force the funds into alternative short-term transactions like Repurchase Agreements (repos), where one party sells a security and promises to buy the security back at a higher price. However, given ongoing quantitative easing purchases by the Fed to the tune of $120B per month, there is less collateral available as the Fed is actively buying Treasury securities. As a result, money market funds would likely face collateral shortages. This is the exact opposite of what occurred in the fall of 2019, when funding markets froze due to a shortage of reserves and oversupply of collateral.
Negative Rates Arrived in the Repo Market and the Fed Targeted the ON RRP
Negative rates did arrive but not in the way JP Morgan warned. In the first week of March 2021, 10-year repo rates dropped from trading barely special to near-record lows (see Figure 5), trading as low as (4.25 percent)—a remarkable figure because repo markets feature a “fail charge” penalty of (3 percent), which is intended to prevent “failure-to-delivers” (i.e., instances when the repo counterparty fails to deliver securities). In short, trading below (3.0 percent) means it’s cheaper to pay the fail charge, and some interpreted the fact that 10-year repos were trading below the level as indication of the magnitude of short positions on Treasuries amid the dominating market narrative of inflation. However, given the dearth of liquidity in Treasury markets and the likelihood for shortage of collateral, the dramatically low repo rates could have been a result of market structure, or a combination of short interest and market structure. Regardless, it is clear something was amiss as repo market participants were effectively paying to lend.
Figure 5: Repo Rates Plunge to Near-Record Lows
The SLR Exemption Expired, the Fed Lifted the Cap on the ON RRP, JP Morgan Issued Preferreds, and the Treasury Quadrupled Its Borrowing Estimate but Still Expects Dramatic Reductions to TGA
Ultimately, the Federal Reserve decided at its March meeting to lift the cap on the ON RRP to $80B from the prior cap of $30B and a few days later announced it was not renewing the SLR exemption. Since then, volumes at the ON RRP facility have exploded—it recently topped $430B to reach the third highest level in history (see Figure 6). This is far higher than the temporary facility’s use at the start of the pandemic and is conceptually laughable; the Fed is taking Treasury securities out of the market via ongoing quantitative easing purchase and turning right around and putting them back into the market via the ON RRP!
Figure 6: Overnight Reverse Repurchase Agreement (ON RRP) Volumes Have ExplodedDespite JP Morgan’s threats of negative rates on deposits if the SLR exemption were to expire, the bank has so far been able to manage the expiration. Its SLR stood at 5.5 percent at the end of the first quarter, just 50 basis points above the 5.0 percent minimum required, and the company had issued $1.5B of preferred stock in the quarter to help pad its capital base (numerator in the SLR formula). Of course, JP Morgan included the following warning to regulators in its quarterly earnings call, “As we told you last quarter, we have levers to manage SLR and we will. However, raising capital against deposits and/or turning away deposits are unnatural actions for banks and cannot be good for the system in the long run.” Going forward, JP Morgan’s three primary levers to manage SLR constraints are (1) to issue more preferred equity; (2) to retain more equity; and (3) to turn away some deposits, either by offering negative rates or simply refusing to accept them. However, JP Morgan’s preference so far has been clear—to issue preferred equity. (For those interested in digging deeper, Credit Suisse explores the recent and massive G-SIB debt issuance in response to operating subsidiary SLR and TLAC constraints LINK).
On May 3, 2021, the Treasury released its latest marketable borrowing estimates (LINK) and disclosed that the pace of drawdown of the TGA has been slower than previously estimated; the Treasury has borrowed more, and expects to borrow more, than previously expected and therefore expects to have a higher cash balance. However, the Treasury is expecting to have “a cash balance of approximately $450 billion at the expiration of the debt limit suspension on July 31.” In the quarter following, the Treasury plans to raise a whopping $821B of debt and projects the cash balance of the TGA will balloon back to roughly $750B. Driving this dramatic increase in borrowing projections is the most recent $1.9T fiscal stimulus bill. While some have interpreted these newest projections as indications the “liquidity tsunami” has ended, it’s important to remember that all borrowing by the Treasury is deficit borrowing. The US government is so far extended over tax receipts that every dollar raised needs to be printed by the Fed, which keeps driving more liquidity. While the Fed will taper at some point, fiscal spending is ensuring that easy money will stick around for some time.
The Current State of Short-Term and Funding Markets
Despite tangible concerns, the expiration of the SLR exemption hasn’t yet sparked the dire consequences forecast as potential. Longer-term Treasury rates have remained fairly constant since the expiration, which suggests the feared liquidation of Treasury securities has not occurred (there are indications it may never occur). As the market narrative has shifted to inflation concerns and the market continues to be flooded with ample reserves from the Fed, demand for collateral has skyrocketed and the G-SIB banks are uniquely positioned to meet this demand. As Polzar details in Credit Suisse’s Global Money Dispatch (LINK, LINK), unlike bond investors, the banks often own Treasuries via transactions—not as outright investments; in other words, they are somewhat price agnostic and willing to lend to short sellers. As Credit Suisse explains, though best enabled to “backstop a bond bull regime” G-SIBs can flip the script and serve as lenders of Treasuries, providing support to the US rate complex and helping firm up Treasury auctions. This seems consistent with what we are seeing in the market. At the end of April, rates in repo markets were trading around zero and often trading negative, Treasury bills were recently issued at a 0 percent rate, the Secured Overnight Financing Rate (SOFR) was 0.01 percent, the Effective Federal Funds Rate (EFFR) was 0.05 percent, and as predicted the ON RRP facility has been flooded with cash.
With a rate of 0 percent at the ON RRP, it is clear this surge of cash is not investors seeking yield, but investors seeking to avoid negative rates found in other short-term investments. We can also assume ample liquidity will force some market participants farther out on the risk curve, as any yield earned is dramatically better than zero—or negative—rates. This has the potential to make money market funds and other short-term investment funds much riskier than would normally be assumed. Following the Fed meeting on April 28, the market is now pricing in over a 90 percent chance of a rate hike by the end of 2022, well ahead of the Fed’s forecast of its first rate hike occurring in 2024. When the Fed does hike and begins to taper its quantitative easing program, other risks could manifest as reserves are removed from the system, such as the “freezing” that occurred just prior to the pandemic. While it is unlikely money market funds currently pose significant risk, we continue to monitor the ongoing tumult in short-term and funding markets.