HomeMembersRecent StatementsPrevious StatementsJoint StatementsIn The PressPrevious NewsWorldwide ShadowsLinksRecent Member Working Papers

Shadow Financial Regulatory Committees of Asia, Australia-New Zealand, Europe, Japan, Latin America and the United States 
“Misdiagnosis of Crisis has led to Botched Liquidity Regulation”
Joint Statement 10
Tokyo, October 28, 2013

Overview
In this statement the Shadow Committees recommend mandatory disclosure of a simple liquidity indicator (SLI) as an alternative to the Basel required minimum liquidity coverage ratio. The SLI is easier to measure, verify and administer by financial market regulators, imposes lower compliance costs, and can be easily applied to estimate the impact of stress. In addition it facilitates comparison across banks and fosters market discipline.
 

Introduction

 The Basel Committee’s response to the 2008-2009 financial crisis is based on a misdiagnosis. The lack of recognition of losses from the original subprime crisis delayed the recovery and economies worldwide are still suffering the consequences. The fundamental cause of this recent crisis was insolvency problems in the U.S. and Europe rather than lack of liquidity. For example, it was uncertainty about which banks had large exposure to bad assets that led to a collapse in the functioning of the interbank money markets. We also observed a significant widening in the credit risk spreads between bank-issued paper and the government issued alternative.

Banking regulatory agencies have recently put forward quantitative liquidity requirements based on the continued misunderstanding of the role that liquidity problems, relative to solvency problems, played in the recent financial crisis. Indeed the U.S. regulatory authority just announced the final implementing regulation for the liquidity coverage ratio last Thursday for large US banks. The Shadow Committees believe that liquidity problems were symptomatic of underlying solvency problems. Hence, what is needed is not a regulation of minimum required liquidity but a clear indicator of the liquidity deficiency of an institution. The Committees instead have an alternative proposal for measurement and disclosure of liquidity that would rely on an indicator that encourages transparency of an institution's short term funding needs and its ability to meet those needs from its liquid assets without extraordinary reliance on central bank funding.
 
Imposing binding liquidity requirements as a means of buying time to deal with a troubled institution facing insolvency may increase the problem, which ultimately may spill over to the rest of the banking sector. Central banks still need to fulfill the traditional role of the lender of last resort for all solvent institutions that need temporary liquidity assistance. In the global financial crisis, such temporary facilities became protracted because the underlying solvency problems were not recognized and addressed.

 Key Attributes for an Effective Liquidity Indicator

The current Basel proposal for required minimum liquidity ratios involves implementing two measures: (i) Liquidity Coverage Ratio (LCR) that will be phased in by 2019 and (ii) Net Stable Funding Ratio that is planned to be implemented by 2018. The latter has been subject to massive criticism and remains under development. The LCR requirement is the focus of this statement.

 Good liquidity management by banks is critical to their ability to survive stressful situations and it is therefore important that banks measure their liquidity positions appropriately and manage them carefully. This is relevant not just for internal bank management, but also for supervisors and for counterparties who ultimately may bear some of the risks arising from poor liquidity management. Indeed, a crucial feature of this statement is that liquidity positions should be publicly disclosed in a form that is easy for outsiders to understand and monitor. The resulting market discipline can help to ensure better management of liquidity risk by individual banks. This is at variance with longstanding bank and regulatory instincts to hide and disguise banks’ financial conditions.

A liquidity indicator should have the following attributes. First, the indicator should be simple to calculate and verify, and second, the measure should be easy to interpret for both supervisors and the market relative to some specified degree of stress. Third, the regulation based on the indicator should be simple to administer from the supervisor's point of view and not onerous to comply with from the bank’s point of view. Fourth, a reliable liquidity monitoring regime would also reduce the problem of asymmetric information that underlies most liquidity problems. Finally, the liquidity indicator should be comparable across banks but allow for differences in business models.

Under the approach we propose, institutions would be free to disclose any additional data that they believe would help the market better understand their liquidity position, but the proposed indicator would ensure that at least one measure could be readily compared across institutions. Moreover, our approach would be less likely to disrupt markets. Imposition of a new required LCR may interact with the large number of other regulatory changes in unpredictable ways imposing yet another constraint on a bank that is already subject to multiple new constraints. Although the authorities allow for a lengthy phase in period that may be used to adjust the ratio to new information about these interactions, experience has shown that the lengthy phase-in may simply give bank lobbyists more time to weaken the ratio and make it still less meaningful.

Basel Proposal for a Liquidity Coverage Ratio

 The Liquidity Coverage Ratio (LCR) regulation proposed by the Basel Committee starts with a similar concept, but it gets complicated quickly. The LCR is calculated by dividing high- quality liquid assets (HQLA) held by the bank by the assumed run-off of funds occurring in a stressed scenario over a 30-day period. A ratio of 100 per cent or greater would imply that the bank is able to meet the run off by selling off HQLA.

From the initial consultation document through to the current version, the LCR regulation increased the complexity in both numerator (HQLA) and denominator (what happens in the stress period).

First, the concept of HQLA became broader. It began with mainly assets that were liquid in their own right because they were claims on very high-quality borrowers including cash, central bank reserves and claims on governments. Over time, the revisions permitted introduction of assets that were less clearly high in quality and less certainly liquid. Eventually, the LCR allowed so-called Level 2 HQLA that could be as much 40% of the total HQLA.2
The fundamental problem is that the numerator of LCR now includes assets that depend on the secondary markets for liquidity. The haircuts attempt to reflect liquidity quality differences, and can be interpreted as implying a liquidity weight for each type of asset. There is no empirical or theoretical basis for such "weighting" and the level 2 assets are most likely to become illiquid in a crisis when markets cease to function well.3
The minimum LCR, currently set at 100%, is intended to be met at all times during benign financial conditions. It must be reported monthly and can be required to be reported more frequently in times of stress. Since assets, which are required to be held, may no longer be liquid, the LCR ratio may fall short of 100% during periods of financial stress. However, the definition of financial stress involved, as well as the period over which the bank may fall short, are unclear. 

Merits of SLI relative to LCR

With regard to the key attributes of an effective liquidity measure, we believe that the SLI dominates the Basel LCR. First, it is easier to measure and verify because it is much less complicated. It is based on basic cash-flow liquidity management 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.
Second, the SLI is based on a well-defined degree of stress while the LCR has no clear relationship to any particular degree of stress. The haircut and run-off assumptions in the LCR seem to differ from the actual experience during the crisis.
Third, the SLI is much easier to interpret than the LCR. It shows the number of days a bank could meet net outflows under a well-defined stress situation without selling illiquid assets or depending on central bank assistance. The Basel LCR was originally intended to provide assurance that a bank had sufficient high quality liquid assets to meet a 30-day outflow under conditions of stress. But since the concept was originally proposed, the numerator was expanded to include lower quality, less liquid assets and the denominator was reduced by applying less onerous haircuts and run-off assumptions, so that it is no longer clear how to interpret the ratio.
Fourth, our approach relies primarily on market discipline for its effectiveness rather than a specified required minimum ratio imposed by regulators. This has several advantages.
Because the measure is consistent across institutions and over time for the same institution, the market is likely to rely on peer-group comparisons of institutions pursuing similar business models. The level of the SLI that the market deems appropriate may vary across peer groups (and across countries). In contrast the Basel LCR imposes a minimum ratio that makes no allowances for differences in business models and is very difficult to compare across institutions or over time because it relies on a complicated set of haircuts, run-off assumptions and limits that will vary in opaque ways as an institution’s portfolio changes.

The SLI is also more likely to reduce the asymmetry of information between financial market participants because it is easier to interpret and compare across institutions. Lack of clarity about the stress scenario implicit in the Basel LCR and the changing composition of both the numerator and denominator impede comparisons across institutions. When markets are uncertain about the credit worthiness of institutions, interbank markets tend to break down exacerbating liquidity problems for all institutions.

The success of our approach depends on whether some significant groups of creditors believe that they might incur some cost if the institution experiences a shortage of liquidity. Currently the authorities lack credibility in this regard. In almost every country, the authorities protected all creditors against loss despite having the power to impose such losses. But many of the other regulatory reforms, including more and better quality capital, higher risk-weighted capital ratios, a new leverage ratio, and more effective resolution policy, may persuade the market that the authorities will be less likely to bailout all creditors in the future.

Admittedly, greater reliance on market discipline may increase the vulnerability of institutions to runs. But to the extent that such runs occur, they are more likely to be focused appropriately on institutions with weak liquidity positions. This may also prompt the authorities to take corrective actions before losses become large. Moreover, the improved disclosure will provide an incentive for banks to place greater emphasis on liquidity management and thereby reduce their vulnerability. 

Summary

 In summary, the SLI dominates the Basel LCR. The SLI is easier to measure, verify and administer by financial market regulators. From the perspective of all financial market participants, it imposes lower compliance costs. The SLI may easily be applied to estimate the impact of stress, while its simplicity in estimation and clarity reduce asymmetries in information between market participants. It also facilitates comparison across peer groups of banks and for the same bank over time. In addition, it avoids the unintended consequences of imposing a regulatory minimum that may interact in complex ways with changes in asset prices in an already complicated financial regulatory structure. Finally, the SLI offers the advantage of greater reliance on market discipline, rather than a uniform regulatory minimum, and it facilitates comparisons between differences in business models across banks now and over time.

 


1 It may be upwardly biased because it omits other net cash outflows that are likely to occur in a crisis, but it is a useful first-order approximation. Ideally, the measure would be forward-looking and reflect the best estimates of net cash outflows under a stress scenario. Unfortunately, such a measure involves numerous subjective assumptions, is difficult to monitor, and presents multiple opportunities for an institution to disguise its deteriorating liquidity condition.
2 Within Level 2 HQLA there are two sub-categories. Level 2A assets include certain corporate debt, covered bonds, and claims on certain foreign sovereigns, which can be included with a haircut of 15% and count up to a maximum of 40% of the total HQLA. Level 2B assets include lower-rated corporate bonds (with ratings A+ to BBB- and given).

haircuts of 50%), residential mortgage-backed securities (25% haircut), and equities (50% haircut) subject to maximum limit of 15% of the total HQLA. 3Overtime,theassumedrun-offratesthatareusedtocalculatethedenominatorhavebecomesmaller. For example, the assumed runoff of uninsured deposits was reduced from 40% to 20%, the run off of insured deposits was reduced from 7.5% to 3%, run off of non-financial corporate deposits from 70% to 40%, and usage of committed liquidity facilities to non-financial corporates reduced from 100% to 30%. They were not based on actual experience during the crisis. They have been negotiated to take account of the special concerns expressed by banks in various countries and this has generally meant a weakening (an approximate halving) of the run-off assumptions. Some of the relaxations were undoubtedly intended to better approximate reality, but the outcome is that the denominator reflects no coherent view of any particular degree of stress. 

Enter supporting content here