Two of the legacies of the financial crisis were new financial regulation that made the banking system safer and a major increase in the size of the Federal Reserve's balance that arguably helped stabilize the economy and added trillions of dollars of reserves to the U.S. banking system.

Notwithstanding the benefits of these twin legacies, they have introduced new complexities to monetary policy and funding markets, as demonstrated partly by the recent spike in repurchase agreement (repo) rates.

That spike, likely triggered by certain factors that temporarily drained liquidity from the U.S. banking system, demonstrates that regulatory financial requirements and the shrinking of the Fed's balance sheet that has occurred in the last two years effectively may be limiting banks' willingness and ability to support market financing during times of stress more than previously thought.

Positively, we do not believe the stress in the market, which has abated, reflects credit concerns, especially since the spike occurred on repos collateralized mostly by low-risk government bonds (the general collateral financing repo market). We do not expect the volatility to cause us to lower ratings on any banks.

Still, smooth functioning of the $2 trillion-plus repo market may require continued support from the Federal Reserve, which, after the spike in repo rates in mid-September 2019, opted to extend funding to the market to lower repo rates.

A Confluence Of Technical Factors, Regulatory Limitations, And A Smaller Fed Balance Sheet

During the week of Sept. 16, the daily Secured Overnight Financing Rate (SOFR), a volume-weighted median for general collateral repo trades, rose as high as 5.25%, with transactions priced as high as 9.00%. We believe corporate tax payments, the settlement of two Treasury bill auctions, and perhaps other factors drained a significant amount of cash temporarily from the banking system, making less liquidity available to the repo market. We believe money market funds also saw roughly $30 billion of outflows related to corporate tax payments.

Chart 1

The spike in repo rates occurred after several months of the Fed shrinking its balance sheet to about $3.8 trillion from $4.5 trillion, as a result of which bank reserves fell by a roughly equivalent amount. The excess reserves of U.S. banks, measured as the reserves banks maintain less their required reserves, dropped to about $1.4 trillion from $2.1 trillion at the end of 2017. Reserve requirements are the amount of funds that a depository institution must hold in reserve against specified deposit liabilities, and the Fed has historically conducted monetary policy in large part by adding to or subtracting from reserves, referred to as "open market operations."

Chart 2

While the $1.4 trillion figure still seems elevated as a result of the Fed's large-scale asset purchases after the 2008-2009 financial crisis (compared with near zero before the financial crisis), we believe other external and internal financial requirements put into place in recent years have introduced additional and often tighter constraints on how banks use capital and liquidity and may affect how the Fed conducts monetary policy. For instance, particularly at the largest banks, the liquidity coverage ratio (LCR) and certain liquidity measurements associated with resolution planning have significantly raised the amount of liquidity the large banks under enhanced regulation must maintain on a consolidated basis and at their key subsidiaries. We believe banks also now aim to hold larger cash balances on an intraday basis rather than paying to borrow to meet intraday needs--as they may have done more frequently before the financial crisis.

The reduction in the Fed's balance sheet as a result of quantitative tightening has likely slimmed banks' cushion above some of those new external and internal requirements. As a result, when liquidity in the banking system dropped and repo rates spiked, dealer banks seemingly were less willing and able to extend additional liquidity to the market to bring those rates down. Lending cash in a repo transaction collateralized by low-risk, highly liquid securities (like Treasuries) should not have a material impact on a bank's consolidated LCR. However, it could require the movement of cash within the organization as well as a depletion of cash below levels the bank may target to stay in compliance with a variety of other liquidity measurements. Even for internal requirements, banks may feel more comfortable holding cash than repos. In addition, banks have other regulatory constraints such as the leverage ratio, and the global systemically important bank buffer, which limit their willingness to grow their balance sheets, particularly at or near quarter-end.

Market participants have also speculated that demand for U.S. dollar repos from emerging market countries may have contributed to the spike in repo rates.

The Future Of Fed Intervention

Given those constraints at the banks, the Fed may need to continue to support the market in some form to keep the federal funds rate within its target range for monetary policy. Since repo rates spiked, the Fed has offered three 14-day term repo operations for an aggregate amount of at least $30 billion each. It is also offering at least $75 billion in daily overnight repos until Oct. 10, 2019.

Whether the Fed continues with such open market operations beyond Oct. 10 likely depends on the persistence of technical factors that propelled the scarcity of repo in mid-September--for example, temporary factors like corporate tax payments and Treasury issuances. Over time, if the U.S. government's deficits were to rise, the repo market will need to absorb greater Treasury issuances, meaning it will need to absorb temporary cash drainage from the money supply.

September's repo spike was contained and did not appear to show signs of spillover to adjacent markets. Nevertheless, we cannot rule out the risk that a prolonged period of uncertainty or stress in repo markets could prompt more contagion to other funding markets. Also, given the size and importance of the repo market to market funding, an extended period of elevated rates would be detrimental to a variety of banks and nonbank financial institutions that rely significantly on it to finance their activities.

We believe the Fed could decide to maintain a somewhat higher buffer of excess reserves in order to lower the risk of future stress in repo markets. Doing so would permit banks to have sufficient spare capacity in terms of excess reserves to facilitate their role as dealers in the repo market. Some market participants have also floated the idea of a standing repo facility similar to what had been in place during the 2008-2009 financial crisis.

The Transition From LIBOR To SOFR

While we do not believe the recent spike will jeopardize the industry's planned transition away from the London Inter Bank Offered Rate (LIBOR; one of the more popular reference rates to calculate interest payments on financial instruments) and toward the SOFR (a repo financing rate), the mid-September jump shows that even SOFR can demonstrate significant volatility and spike higher than LIBOR.

Some banks have worried that SOFR could contract during periods of stress, given it is a risk-free rate, at a time that LIBOR (and bank funding costs) probably would widen. That contraction could be disadvantageous to banks because it could lead to lower yields on floating-rate assets at the same time that their liability costs widen. In contrast, with a spike, there may be a different concern from bank borrowers that they may have to pay more on borrowings tied to SOFR.

Chart 3

Still, we don't see temporary and infrequent spikes in SOFR as a major risk in and of themselves, largely because asset yields tied to SOFR will likely be averages over longer periods such as three months. For instance, the jump to 5.25% in the SOFR overnight rate on Sept. 17, 2019, would only result in an approximately 15-bps increase in one-month SOFR and a 5-bps increase in three-month SOFR.

Related Research

Farewell (Regional Bank) LCR, We Hardly Knew Ye: What The Liquidity Coverage Ratio Tells Us And What Could Be Lost, June 17, 2019