1. Why did the Fed expand its balance sheet by $300 billion?
Since last Friday, some regional banks in the United States have suffered severe runs due to the SVB incident, with depositors transferring large amounts of deposits to large banks. Some small and medium-sized banks facing liquidity pressures have had to turn to the Fed for loans by pledging collateral.
Currently, the Fed has provided liquidity through three channels. The first is the traditional Discount Window (DW), which provided approximately $148 billion in liquidity. The second is the newly established liquidity tool BTFP, which provided approximately $12 billion in liquidity. The third is the $143 billion provided to the new entities of SVB and Signature, which were taken over by the FDIC.
2. What are the relationships and differences among these three types of loans?
All three are collateralized loans, but there are differences in borrowers, collateral content, loan terms, collateral valuation, interest rates, and disclosure details.
The Discount Window (DW) is the most traditional tool used by the Fed to provide liquidity support to banks facing liquidity problems. Borrowers are all sound savings institutions (including banks and thrifts); collateral includes not only risk-free US Treasuries and Agency MBS, but also various corporate loans, commercial loans, and household loans. Loan terms range from overnight to three months. The loan amount provided is equivalent to the market value (or market value estimate) of the collateral minus a certain discount (haircut, depending on the collateral rating). The interest rate is the upper bound of the federal funds rate range (currently 4.75%), and borrower information is disclosed after two years.
BTFP is a new tool launched by the Fed this week. Borrowers are also all savings institutions, and collateral only includes risk-free US Treasuries and Agency debt (including MBS). The loan term is up to one year, and the loan amount provided is equal to the par value of the collateral. The interest rate is OIS plus 10bps, and borrower information is disclosed in March 2025.
FDIC loans are essentially loans provided to the new entities of “Silicon Valley Bridge Bank” (SVBB) and “Signature Bridge Bank” (SBB) established after the FDIC took over SVB and Signature, respectively, to meet the needs of these banks to complete the withdrawal of deposits. These loans are backed by guarantees provided by the Treasury Department (FDIC attributed to the Treasury Department for management). The interest rate and term of these loans are unknown.
3. Why is the usage of the discount window much higher than the BTFP?
There are several possible reasons, listed from high to low likelihood.
The main reason is probably that the collateral requirements for BTFP are too narrow, as it only accepts safe assets. Many of the banks that experienced liquidity shocks are regional banks with a primary focus on lending business, and they don’t have many government bonds or MBS assets on their balance sheets, such as FRC. Therefore, they cannot use BTFP and can only use the discount window.
The second possible reason is that BTFP is a new tool and banks are not yet familiar with it.
The third possible reason is that the stigma effect of BTFP is stronger than that of the discount window. However, I personally believe that this argument is not convincing because the collateral requirements for BTFP are narrower. Although investors can see a bank’s unrealized losses from using BTFP, this information has already been disclosed in the bank’s annual report, and it does not constitute a new signal for listed banks. Conversely, the collateral requirements for the discount window are broader and have always been seen as having a stigma effect. Therefore, the stigma effect may not be the main reason for the significant difference in usage between the two tools.
4. Should this total of 300 billion be understood as the Fed restarting quantitative easing (QE)?
Absolutely not. For central banks, monetary policy and financial stability are two main goals, each with its own set of tools. The former includes interest rate tools and balance sheet tools, while the latter includes macroprudential and various temporary liquidity tools, which can be used without conflict. Just like the BOE raising interest rates while temporarily implementing QE to save local pension funds last September. In fact, the ECB’s meeting on Thursday will raise interest rates as expected, and Lagarde has also expressed the same view that price stability and financial stability are not in conflict.
Back to the present, QE is a monetary policy tool aimed at price stability and full employment, while this 300 billion is temporary liquidity support aimed at maintaining financial stability. Mechanically, QE involves the Fed actively buying assets from the market and creating reserves, while this 300 billion in liquidity involves commercial banks borrowing money from the central bank using qualified collateral to address their liquidity shortage. The Fed is passively injecting reserves. In terms of effects, the reserves created by QE will indirectly or directly flow into the bond and stock markets, depressing risk-free interest rates and boosting the prices of risk-free and risky assets. However, the reserves created by this emergency liquidity tool are liquidity reserves that banks hold to cope with liability-side runs or prevent future liability-side outflows, and are unlikely to flow into the bond and stock markets. The recent decline in risk-free asset rates is more likely due to the escalation of market risk aversion than the impact of the 300 billion flow.
5. Why is the usage of the $300 billion discount window higher than during the 2008 GFC, does it indicate greater liquidity pressure on US banks than in 2008?
If we only compare the usage of the Discount Window (DW), then the usage last week did indeed exceed the usage during the 2008 GFC by around $100 billion. However, if we look at the total amount of reserves created by “all liquidity tools,” then the $300 billion usage this week is far lower than the peak of nearly $1.7 trillion during 2008. The reason is that the DW was not the main tool for creating liquidity in 2008.
One reason is that banks at that time were very afraid of the stigma associated with the DW and tried to avoid using it as much as possible, which led the Federal Reserve to rely more on active auctions of liquidity (such as the Term Auction Facility) to inject liquidity. Another reason is that the financial institutions facing liquidity difficulties at that time were not limited to banks, but also included primary dealers (PDs, which were not regulated by the Federal Reserve and could not directly borrow from it like banks), money market funds, corporate financing departments, asset-backed securities (ABS), non-US financial institutions, and so on. Therefore, the Federal Reserve established a large number of targeted tools to support the financing needs of different sectors. For example, the Money Market Investor Funding Facility (MMIFF) for money market funds, the Primary Dealer Credit Facility (PDCF) for primary dealers, the Term Asset-Backed Securities Loan Facility (TALF) for the ABS market, the swap lines (with foreign central banks as counterparties) for non-US institutions, and the Commercial Paper Funding Facility (CPFF) for corporate short-term financing activities. These specific tools were the main providers of liquidity at that time, and the DW played a relatively minor role.
Therefore, if we look at the sum of liquidity tools, the liquidity pressure facing the current financial system is far lower than during the 2008 period.
6. What are the forward-looking implications of such large usage of liquidity tools on US monetary policy?
As mentioned earlier, central banks usually distinguish financial stability from price stability and use two sets of tools to address them separately. Therefore, as long as financial stability tools are playing their corresponding roles, the central bank usually has more confidence in continuing its original policy stance in terms of monetary policy. Since DW and BTFP are currently being used and are effective, it actually reduces the need for the Federal Reserve to use interest rate tools to maintain financial stability — at least in the short term. After all, trying to alleviate the liquidity pressure or solvency crisis of financial institutions through adjusting interest rates is like telling a seriously ill person to drink more hot water; it is probably not very effective.
7. How did JPM come up with the estimate of $2 trillion for the potential usage of the BTFP?
The JPM report actually stated that the theoretical upper limit for the usage of the BTFP tool could reach $2 trillion. The calculation method is simple: assuming that the entire banking system uses all of its holdings of US Treasuries and Agency MBS as collateral, it could exchange them for about $2 trillion.
However, most of the banks that experienced problems this time were regional banks, and the four largest banks in the United States (JPM, Citi, BoA, and Wells Fargo) not only did not face a liquidity crisis but also faced excess liquidity due to deposit inflows. In contrast, the distribution of US Treasuries and MBS is mainly concentrated in these four major banks, whose holdings of US Treasuries + MBS total $2 trillion, accounting for 70% of the total US banking system’s holdings of US Treasuries + MBS. Since the four major banks are unlikely to use the BTFP, the banks that truly need to use it can only offer about $700 billion of eligible collateral at most. Furthermore, considering that the FHLBs and the four major banks will also provide some liquidity to banks experiencing runs, the final usage of the BTFP is likely to be at most in the hundreds of billions of dollars.