Events over the past 15 years have shown how important it is for major financial institutions to remain highly liquid. That lesson was encapsulated in the regulatory response to the 2008 financial crisis. The foremost new requirement was the Basel Committee on Banking Supervision’s liquidity coverage ratio (LCR), which requires banks to hold a minimum amount of high-quality liquid assets.
In 2014, the definition of HQLAs was expanded to encompass borrowing capacity at a committed liquidity facility. Although CLFs – contractual arrangements whereby a central bank promises to provide a commercial bank with funds on demand – have been adopted in a couple of jurisdictions, major economies, including the US, have proved reluctant to follow suit. This is a pity, as the facilities would serve numerous policy goals and solve important problems faced by the Federal Reserve.
CLFs would benefit Main Street as well as Wall Street by allowing banks to lend more to US businesses and consumers and make fewer loans to the federal government. Currently, banks primarily satisfy their liquidity requirements by owning Treasury securities (loans to the federal government) and reserve balances (loans to the Fed that are, in turn, invested in Treasuries). If the Fed created a CLF, and if any line of credit extended under the facility were recognised as an HQLA, a bank could instead lend to a small business, a student or a farmer and use the loan as collateral to back its line of credit with the US central bank.
According to the most recent available data, banks have pledged collateral with a lendable value of $1.6 trillion to the Fed. If US banks used part of this to establish CLFs with a capacity of $1 trillion, they could reduce their lending to the government by a similar amount. Lending to non-financial businesses and households would increase by about 12%. Based on a recent review by the Bank for International Settlements of the impacts of liquidity and capital requirements, this would translate into a permanent increase in GDP of between 0.25% and 0.75% – or about 370,000 to 1 million new jobs. CLF capacity and reserve balances provide the same amounts of actual liquidity, so the shift would in no way degrade the liquidity condition of the banks.
CLFs would also help to avert the risk of congestion in the market for HQLAs when banks need liquidity. The LCR offers only a short list of acceptable options as HQLAs. This means that if many banks seek to raise liquidity at the same time as using their HQLAs, markets for these securities could be overwhelmed. At the start of the pandemic in early 2020, many banks sought to sell the same types of assets simultaneously, with the result that broker-dealers were sometimes unable to fulfil orders. If banks under stress could instead draw down CLFs, this would leave greater room for securities markets to handle liquidity demands from other institutions.
Commercial banks are often unwilling to use the discount window because of the stigma it carries, but there could be significantly less stigma associated with a CLF. Banks would be paying for the privilege of maintaining CLF capacity and could view use of the facility as a right. CLFs would be knitted into liquidity planning by bank management, bank regulation and bank examiners, and would look and feel different to the discount window.
Ultimately, though, success in keeping CLFs stigma-free would depend on the Fed leadership educating the public and Congress about the beneficial purpose of the facilities. During periods of financial turmoil, the Fed will lower the discount rate and extend the terms on related loans, and it could do the same with the terms on a CLF. Moreover, since the Fed would already have committed to providing the facility, and the liquidity takes the form of a line of credit rather than a deposit, it would be instantly expanding its supply of liquidity to the financial system without expanding its balance sheet. The Fed could also lower the fee for the service or the haircuts applied to HQLAs, or expand any limit it placed on the share of HQLAs that could form part of the CLF capacity. If CLFs were largely stigma-free, such actions could be considerably more effective than a similar easing of terms on the discount window.
Because CLFs would involve the Fed charging banks for the liquidity it is currently paying them to maintain, this would raise money for taxpayers. A committed line of credit is economically the same as a reserve balance held at the Fed. By contrast, the Fed would earn a fee on the liquidity provided by the CLF. If banks were to establish the aforementioned $1 trillion in credit lines via CLFs and the Fed were to charge 15 basis points for the new lines of credit, the US central bank would earn $1.5 billion annually in fees.
The lower demand from banks for reserve balances would also help the Fed to reduce the portfolio of Treasury securities and agency mortgage-backed securities it acquired in response to the Covid crisis. When reserve balances reach the minimum level banks can use for liquidity management, money market interest rates will rise above the rate the Fed pays on the reserves; this is because banks will require an interest rate premium to part with their reserves. At that point, the Fed will need to stop shrinking its portfolio or significantly adjust the way it implements policy. Because banks could satisfy more of their liquidity requirements with a line of credit via a CLF than with deposits at a branch of the Fed, the central bank’s portfolio could become smaller than would otherwise be the case.
CLFs would also improve the functioning of the Treasury market, which has deteriorated in recent years as banks have become bound by risk-insensitive leverage ratio capital requirements. These have discouraged banks from engaging in low-risk, balance sheet-intensive activities such as Treasury market intermediation and financing. Half of the assets held at large banks are at institutions bound by leverage ratios. If banks could satisfy some of the liquidity requirements using CLF capacity rather than reserve balances or Treasury securities, they would be less constrained.
Lastly, and by no means least, creating and recognising CLFs as HQLAs would make liquidity assessments, including the LCR, more accurate. A bank with greater capacity to borrow – either from the discount window or via a CLF – is more liquid than an otherwise identical bank with less capacity. Because that difference is not recognised in regulatory liquidity assessments, banks do not currently have the appropriate incentive to establish and maintain the capacity.
Was Basel faulty?
CLFs were initially proposed by the Basel Committee as a tool for jurisdictions with insufficient level 1 HQLAs, which consist of reserve balances and Treasury securities. Under a CLF, a bank must maintain unencumbered collateral pledged to the central bank to cover the lines of credit provided. A line is irrevocable unless the bank becomes insolvent.
The LCR requires each bank to hold HQLAs equal to or greater than its projected net cash outflows over 30 days of idiosyncratic and systemic stress. The outflows are not “projected” but are instead determined by multiplying the different categories of bank assets and liabilities by fixed constants that are meant to approximate inflow and outflow behaviour under stress.
However, some jurisdictions do not run large budget deficits and do not issue sufficient government debt for banks to purchase to meet their HQLA needs. From the start, the LCR standard allowed such jurisdictions to count collateralised lines of credit provided under CLFs as HQLAs.
To be compliant with the LCR, CLFs must also be provided for a fee that is intended to make the facilities approximately as costly to the bank as holding regular HQLAs. In particular, the net yield on the assets backing the CLFs, after subtracting the fee, should not be higher than a representative yield on a representative portfolio of level 1 and level 2 assets. In the US, level 2 assets comprise agency debt, agency mortgage-backed securities and other less liquid securities.
In 2014, the Basel Committee allowed all jurisdictions to count CLFs as HQLAs. However, in jurisdictions that were not systemically short of such assets, the facilities can only be offered at extremely elevated fees and can only count as a small fraction of HQLAs.
Yet despite the Basel requirements, there is still a case for the Fed to offer a CLF at a market rate and in which lines of credit provided by the facility would count as level 1 HQLAs. The term sheet below shows how such a facility might work.