Download PDF Motivated by the perception that liquidity problems caused the recent financial crisis, the Basel Committee on Banking Supervision included minimum liquidity requirements for banks in the new Basel III framework. Although many observers, including the Shadow Financial Regulatory Committee, would argue that most of the liquidity problems were a symptom of deeper underlying insolvency problems and uncertainty about how the losses would be allocated, it is clear that markets and regulators need much better information on the liquidity positions of banks than has been collected to date. To be sure, regulatory standards for more liquid balance sheets and less reliance on short-term funding will promote system stability, but will do little good to staunch destabilizing runs and credit contractions if banks are too thinly capitalized to absorb losses in an economic downturn.
The Basel Committee envisioned two complementary liquidity ratios: one focused on short-term funding requirements (the Liquidity Coverage Ratio (LCR)) and the other focused on the mismatch between the maturity of assets and the maturity of liabilities (the Net Stable Funding Ratio (NSFR)). The NSFR has not yet reached the proposal stage, but the LCR has been proposed by the Basel Committee and revised several times. In late November 2013, the US regulatory authorities have just released their proposal for a “super equivalent” version of the LCR to be applied in the United States.
The original Basel framework described a very clear concept. It set a requirement that would assure that all large, internationally active banks could survive a 30-day period of stress without recourse to the lender of last resort. The proposed measure divided the stock of ‘High Quality Liquid Assets’ (HQLA) by the cumulative 30-day net cash outflow under conditions of stress (NCOS) and required that HQLA be at least 100% of NCOS. This proposal was adopted in December 2010. But after substantial resistance from European regulators and banks, the ratio was weakened in three ways during January 2013.
First, HQLA, the numerator, was expanded to include many kinds of assets that depend for their liquidity on well-functioning secondary markets. Eligible assets were divided into Level 1, Level 2A and Level 2B assets to reflect differences in liquidity through haircuts and limits. Unfortunately, these have no clear basis in experience during the crisis.
Second, the denominator was weakened by lowering the run-off assumptions regarding assets and liabilities --the primary driver of net cash flows-- by roughly half. Third, the phase-in period was lengthened considerably with a delayed start date of 2016 at an initial ratio of 60% and annual increases of 10% in the minimum until the LCR reaches the full 100% in 2019.
The US regulatory authorities promised to implement a “super equivalent” version of the LCR, which was published in the Federal Register in November 2013. They tightened the definition of HQLA, by excluding several assets that are permissible under the Basel proposal and they made the denominator more rigorous by basing it on the maximum cumulative net outflow on any one day during the 30-day period and tightening the definitions of outflows and inflows to be included. They also shortened the phase-in period proposing a start date of January 2015 at 80% reaching the 100% ratio by 2017. Finally, the US authorities put greater sanctions on institutions that fall short of the ratio for more than two consecutive days. If an institution is below the 100% minimum for three consecutive days, it must submit a liquidity compliance plan to its primary supervisor.
The US regulators deserve credit for delivering on their promise to strengthen the Basel LCR. But the Shadow Committee has a number of concerns about whether the approach is the most effective way to deal with liquidity problems. First, the numerator includes assets that depend on well-functioning secondary markets for their liquidity. Haircuts and caps attempt to capture differences in the quality of HQLAs, but these haircuts and caps have little basis in theory or empirical evidence and make it very complicated to compare HQLA across banks or within the same bank over time. The denominator is based on a series of politically negotiated run-off rates that do not appear to be based on actual experience during the crisis. Thus, even though the supervisory authorities will provide a stress scenario, the denominator will not necessarily reflect a coherent view about the impact on the LCR.
In October 2013, the six international Shadow Financial Regulatory Committees met in Tokyo. The Committees identified four problems: (1) Excessive complexity in monitoring, administering and complying with the LCR; (2) Unanticipated interactions with the plethora of new regulatory requirements and the potential unwinding of unconventional monetary policies that have pumped liquidity into the economy at an unparalleled rate; (3) the lack of transparency of the ratio, which perpetuates the asymmetric information that ignites most runs and makes it impossible to compare the liquidity position of one bank over time or of several banks at the same time; (4) allowance for difference in business models across financial institutions.
The Shadow Committees felt that the objective of liquidity standards should be to gauge the ability of a financial institution to survive a crisis when it is prohibitively expensive to attract new funding. The HQLAs should be confined to assets that do not depend on a well-functioning secondary market for liquidity or on extraordinary central support.
The denominator should reflect liquidity needs during a specific worst-case stress test. Each institution should be asked to recalculate its net cash flows over the preceding 30 days based on the assumption that the institution would not have been able to roll-over its uninsured liabilities. The largest cumulative net cash outflow on any of the 30 days would be the denominator.
Ideally, it would be preferable to use a forward-looking measure that reflects best estimates of net cash outflows under a specified stress scenario. Unfortunately, this would require such a large number of subjective assumptions that it would be an administrative and compliance nightmare. Moreover, the complexity would make it relatively easy for an institution to disguise a deteriorating liquidity position.
The Shadow Committees recommend a ‘Simple Liquidity Indicator’ (SLI) instead of a required minimum. This reflects the concern that a required minimum might interact with other regulations and unconventional monetary policy in ways that disrupt financial markets and the real economy. Moreover, different ratios will be appropriate for different business models and this information will allow the market to form peer groups for comparisons among institutions that pursue similar lines of business.
This approach is much easier to measure, verify and administer. It is based on cash flows rather than a complicated combination of haircuts, run-off assumptions and limits that still fail to capture changes in the liquidity of assets and markets over time. This approach is based on a well-specified degree of stress that is easy to understand, not a combination of subjective haircut and run-off assumptions constrained by arbitrary caps. Moreover this will greatly reduce the compliance costs for banks.
The SLI can be easily interpreted. It is the number of days that an institution could meet its net cash outflows (if it were unable to roll-over its liabilities) without selling illiquid assets or depending on central bank assistance.
The SLI would facilitate comparisons over time and across banks at a moment in time, which is not possible for the LCR. This greater transparency will foster market discipline, which may be a better way of dealing with this problem at a time when there is considerable uncertainty about what the regulatory minimum should be.