Banks' Liquidity

Liquidity risk is an integral part of banks' activities. A bank's complex liquidity profile can be summarised as liquidity indicators for use in the bank's internal risk management and in the supervision of liquidity.

Based on available data, the liquidity of Danish banks is judged to be more than adequate. The stock of liquidity is comparable to that in other EU member states with comparatively sophisticated financial systems and by far exceeds the statutory requirement, even though the excess liquidity cover has declined somewhat in recent years. Denmark's statutory liquidity requirement is more quantitative and stock-based than equivalent requirements in other EU member states, but it has functioned as an efficient check. The Danish Financial Supervisory Authority supervises individual banks and ensures that they have adequate liquidity and contingency liquidity plans.

INTRODUCTION

Liquidity has various meanings in different contexts, e.g. market liquidity of financial instruments, liquidity in payment, clearing and settlement systems, monetary-policy liquidity and the financial liquidity of individual entities. This chapter focuses on the liquidity risks in banks as these risks have attracted growing interest from central banks and supervisory authorities in recent years. Firstly, liquidity problems in banks have played a significant role in financial crises in many countries. This is one reason why explicit regulation and supervision of the banks' liquidity form part of most countries' supervisory regimes.

Secondly, supplying the banking system with sufficient liquidity is primarily seen as a central-bank responsibility. This is the case for the general implementation of monetary policy and for the design of robust payment systems, as well as for the provision of liquidity support in the event of a financial crisis.

Thirdly, liquidity supervision and regulation is one element of the framework for the EU banking sector that so far has not been harmonised. One argument in favour of national liquidity regulations is that banks' liquidity needs depend not only on each bank's business profile, but also on the national financial system, and particularly on the scope for selling or borrowing against various assets (e.g. securities) in different countries.

With increased international banking, the lack of harmonisation of national liquidity requirements can potentially distort competition and encourage regulatory arbitrage. It may also impede international banks' liquidity management, e.g. by tying up liquidity in certain countries, business units or currencies. The growing international interest could in due course lead to harmonisation of liquidity supervision.

Banks' liquidity risk
Liquidity risk can be roughly defined as the risk of not being able to raise cash to make payments as they fall due. For ordinary individuals this e.g. means the risk of unexpected expenses or loss of income, which may be hedged and managed to varying degrees, e.g. via insurance, extra cash in the bank, an overdraft facility, credit cards or other financial products. Bank customers regard their demand deposits and overdraft facilities as liquidity that can be converted into cash at ATMs in their home country or abroad, or used as ready means of payment (i.e. money) via payment cards or bank transfers.

Banks supply liquidity to individuals and companies, and take on liquidity risks as an integral part of their lending and deposit-taking activities. Banks take deposits, typically demand deposits that can be withdrawn again immediately, and relend the funds, e.g. as a 7-year car loan. This creates a mismatch between the payment profiles for deposits and lending.

For the Danish banking sector as a whole, liabilities payable on demand or at less than one month's notice make up more than two thirds of total liabilities, while cash in hand accounts for less than 1 per cent of total assets, and liquid assets – as defined in section 152 of the Danish Financial Business Act, cf. Box 14 – constitute approximately 30 per cent of the total assets, cf. Chart 46.

This balance-sheet structure, plus the potential liquidity drain from guarantees and other off-balance-sheet items, may at first sight appear to entail a high level of potential liquidity risk. However, banks monitor their liquidity closely, subject to the supervision of the Danish Financial Supervisory Authority.

Section 152 of the Danish Financial Business Act, cf. Box 14, requires banks to have "appropriate liquidity" (broadly defined in subsection (2)) to cover 15 per cent of short-term liabilities and 10 per cent of total debt liabilities (excluding capital), cf. Chart 46. There are several reasons that liquid assets covering only 15 per cent of a bank's short-term liabilities, can be considered to be "appropriate liquidity":

• Portfolio effects: the bank pools the liquidity risks associated with the individual customers – the uncertainty as to when the individual customers will need or want to spend the money in their accounts or on their overdraft facilities – as a diversified portfolio of deposits from and loans to a wide range of customers. This provides for a more stable and predictable payment profile for the overall portfolio than for each individual account.
• IT systems and historical data make it easier for banks to get an overview of their liquidity profile and risks.
• Confidence in the banking sector, strengthened by deposit insurance, reduces the risk of sudden panic among the depositors.
• The relatively advanced financial system in Denmark, including extensive use of electronic payment systems, Danmarks Nationalbank's certificates of deposit and liquidity facilities via the extended automatic collateralisation arrangement at VP Securities Services, etc., allows banks to obtain liquidity immediately at a reasonable price on the basis of a wider range of assets.
• Banks smooth their individual liquidity effects via interbank deposits and lending, and banks with easy and secure access to market liquidity sell liquidity insurance (via backup lines^[1] ) to other banks. In normal times, these instruments contribute to redistributing liquidity in the sector. In the event of bank-specific liquidity problems, where the bank's solvency is not a concern, such instruments can form an important part of the contingency liquidity plans of the individual bank. In the event of a systemic banking crisis, however, where all banks seek to draw on their interbank liquidity, these instruments – that normally have a stabilising effect – can become potential channels of contagion, especially in the absence of timely central-bank liquidity support.

BANKS' LIQUIDITY RISK MANAGEMENT

Liquidity problems can threaten the very existence of a bank. The risk of liquidity problems can be minimised by maintaining substantial holdings of cash and fully liquid assets. A careful strategy like this will be inefficient, however, entailing considerable opportunity costs in the form of lower interest income. A milder form of liquidity risk is that inefficient liquidity management can increase the cost and earnings risks. On days with a liquidity shortfall, there is a need to borrow in the market at perhaps higher-than-normal interest costs, and on days with a surplus this liquidity will need to be deposited or invested, perhaps at a lower interest rate or a higher credit risk. Thus, efficient liquidity management can minimise these liquidity-related cost and earnings risks.

Efficient liquidity management requires a comprehensive overview of the bank's total actual and potential payment obligations, as well as a good contingency liquidity plan with a view to adapting the liquidity position to the actual development. The overview of the bank's liquidity position is complicated by two aspects: (i) the time dimension, i.e. the payments in the next few hours^[2] , days, weeks, months and years; and (ii) the uncertainty regarding the liquidity effect of many items with a non-deterministic (i.e. not precisely defined) payment profile. This applies to liabilities (e.g. demand deposits), assets (e.g. overdraft facilities) and off-balance-sheet items (e.g. guarantees and derivative contracts).

A bank must be able continuously to calculate, project and monitor the liquidity need for its entire portfolio. In the calculation and projection of the liquidity need, liquidity risk management systems will typically disaggregate the portfolio into items with a deterministic (i.e. precisely defined) payment profile and items with a non-deterministic (not precisely defined) payment profile, cf. Chart 47. Items with a non-deterministic payment profile can be broken down further into items with payments related directly to specific market rates, e.g. currency swaps or variable-rate loans, and items with payments related to apparently random human behaviour, e.g. demand deposits. Market-related payment profiles can be integrated into the bank's market-risk models, while random payments can be projected (with confidence intervals for the risk in the liquidity development) on data for historical behaviour.

Chart 48 shows the liquidity profile given by such a liquidity management model for a fictive bank.

The liquidity curve shows the expected development in the bank's liquidity position over the time horizon, i.e. the most probable development in liquidity if the banking institution stays on its charted course without raising additional liquidity. For each point on the time horizon, the liquidity curve shows the sum of the deterministic payments plus the expected value of market-related and random payments.

The Liquidity-at-Risk (LaR) curve illustrates the highest risk (with an x per cent confidence interval) in the projected development in the bank's liquidity position, i.e. the worst case for the liquidity development, assuming that the banking institution stays on its charted course without raising additional liquidity^[3].

The red curve illustrates the bank's contingency liquidity, i.e. how much money (for meeting payment obligations) can be raised, and how soon, from sale of, borrowing against or repos on liquid assets, or from reliable funding sources. The contingency liquidity does not include the possibility of new bond issues or other sources of funding that would require continued unhindered access to the financial markets. The contingency liquidity reflects only the bank's resilience to liquidity shocks, or how fast the bank can restore liquidity if it decides to change course.

The liquidity profile in Chart 48 is complicated by both the time dimension and the uncertainty, as mentioned above. The liquidity profile is a snapshot of the liquidity situation and outlook at a given time on a given day. It may change relatively quickly, depending on the bank's business profile, since the liquidity profile is continuously updated with new transactions and changes in the payment profiles of existing items. The liquidity profile does not imply proceeding along the liquidity curve, but is a management tool used by the bank to determine how much liquidity should be raised or placed over the time horizon.

Liquidity indicators
An overview of a bank's liquidity position can be gleaned from liquidity indicators that give a simple summary of the information in the complex liquidity profile from various angles. Liquidity indicators can be shown as time series and can help the bank's senior management and supervisors to monitor the development in the bank's liquidity in relation to supervisory requirements or internal limits.

Mismatch-based liquidity indicators focus on the liquidity amounts, or the mismatch between ingoing and outgoing payments over a given time horizon. The relevant time horizon depends on the bank's business profile. Many liquidity indicators focus on the 1-month horizon, e.g. A in Chart 48. Other mismatch-based liquidity indicators focus on the time horizon associated with the greatest negative liquidity risk. The liquidity indicator can be specified in various ways, e.g. to measure the liquidity capacity in relation to the liquidity curve or LaR.

The simplest stock-based liquidity indicators ignore the time dimension and measure e.g. liquid assets as a ratio of total assets. Other types of liquidity indicators concentrate solely on the time dimension, e.g. on how far the bank is from having to draw on its contingency liquidity. In Chart 48, B, for example, shows a "day-count to default" of 2 months between the lowest point on the LaR curve (3 months) and the corresponding point on the liquidity-capacity curve (1 month).

LIQUIDITY SUPERVISION

In contrast to the internationally harmonised capital-adequacy requirements and accounting standards, the regimes for supervision of banks' liquidity still vary considerably across countries. Liquidity regimes can be roughly classified according to two main dimensions: by type of approach (quantitative/qualitative) and by type of indicator (stock/mismatch), cf. Chart 49.

Quantitative liquidity regimes focus on compliance with liquidity indicators in relation to statutory thresholds, while qualitative liquidity regimes are based on dialogue concerning the banks' individual procedures and liquidity management systems. Many EU member states employ a hybrid approach that to varying degrees emphasises both compliance with quantitative indicators and a qualitative supervisory dialogue about the banks' own liquidity management.

Liquidity regimes applying a quantitative or a hybrid approach use very different liquidity indicators. These indicators can be classified on a spectrum from stock-based to mismatch-based liquidity indicators.

Quantitative liquidity supervision regimes with stock-based liquidity indicators were more widespread a few years ago. Several countries have in recent years switched to mismatch-based or hybrid indicators, or towards a more qualitative approach, particularly since the Basel Committee's revised recommendations were issued, cf. Box 15.

DATA FOR BANKS' LIQUIDITY

The Danish Financial Supervisory Authority publishes indicators for the banks' excess liquidity cover relative to the statutory liquidity requirement, cf. Chart 50.

The median and the 25th percentile in Chart 50 show that most banks have far more liquidity than required by section 152 of the Danish Financial Business Act. The excess liquidity cover has diminished somewhat in the last few years, but most banks still have ample excess cover. The stock-based specification of liquidity requirement in section 152 – including the requirement to promptly report and remedy any breach of the liquidity requirement – seems to function as a check which the banks seek to stay clear of.

The Danish Financial Supervisory Authority's data for the sector also shows which types of liquid assets are included in the banks' liquidity, cf. Chart 51. Liquid interbank loans account for only a modest proportion of the banks' liquidity. It can be argued that interbank claims are internal to the banking sector and should therefore be excluded from the measure of the overall liquidity in the banking sector, since they would not provide much resilience against a systemic liquidity crisis. The low proportion of interbank loans in total liquidity, though somewhat higher for small banks, can therefore from a systemic stability perspective be regarded as reassuring.

Listed and eligible securities account for a large share of banks' liquidity, particularly for large banks. These securities are highly liquid in the Danish financial system since they can be used easily and quickly to raise means of payment, particularly after the extension of the automatic collateralisation arrangement in 2002. The banking sector's stock of securities is shown in the monthly MFI data, cf. Chart 52.

The banks' stock of securities has accounted for around 20-30 per cent of total assets and has grown since 2000.

International comparison of liquidity cover
The Banking Supervision Committee of the ECB has for some years collected consolidated banking data from member states for the calculation of stability indicators for the banking sectors of the EU and the euro area. Indicators for the banking sectors of individual member states were published in the annual EU Banking Sector Stability Report 2005. These indicators include three different definitions of liquid assets as a ratio of total assets. The broadest definition, Liquidity3, has a high degree of overlap with the definition of liquidity in section 152. Thus, the liquidity cover of Danish banks can – subject to national data differences – be roughly compared with corresponding figures for other EU member states, cf. Chart 53. The data for the Danish banking sector is provided by the Danish Financial Supervisory Authority and covers banks only, whereas the data from the ECB report for other countries covers all credit institutions. Danish mortgage-credit institutes are excluded from this comparison because they manage their liquidity risks tightly according to the balance principle, which minimises liquidity risks and the need for liquid assets.

The Danish banking sector's liquidity is comparable to that of other EU member states. The typically higher levels of liquidity in new member states – particularly as regards Liquidity1, i.e. cash in hand and other ultra-liquid funds – may reflect their less developed financial systems with a higher degree of uncertainty concerning the liquidity profile and fewer liquidity management instruments, making more substantial liquidity holdings necessary.

The banks' structural funding and liquidity risks
The banks' balance sheets have grown in recent years, with a stronger increase in lending to than deposits from the non-financial sector. This has eroded the deposit surplus, obliging the banks to resort to other sources of funding to make up the difference, e.g. by issuing bonds and borrowing from other credit institutions, cf. Chart 54.

The Danish deposit guarantee scheme covers deposits up to kr. 300,000 so that ordinary depositors need not fear for a bank's solvency. This makes the banks less vulnerable to panic among retail depositors. Retail deposits are therefore – despite the short formal maturity, often on a demand basis – regarded as one of the banks' most stable and reliable sources of funding.

Bond issues constitute an important funding instrument with long maturities and fixed payment profiles. The maturity is not infinite, however, and bond debt requires refinancing on maturity.

Net debt from credit institutions includes liquidity transfers from foreign banks to branches and subsidiaries in Denmark, which can be regarded as a stable source of funding (unless the parent bank gets into problems). However, it also includes some borrowing from foreign banks, which could turn out to be a less stable source of funding and liquidity, especially in the event of a financial crisis.

A common feature of bond debt and net debt to credit institutions is that the bank must be able to maintain market access and that these sources of funding can make the bank vulnerable to changes in its own credit spreads. A case in point is that of the Icelandic banks in the spring of 2006 when a pronounced shift in market sentiment concerning the Icelandic currency and banking sector widened the credit spreads for Icelandic banks in the international markets.

[1] Back-up lines reportedly play a very small part in the liquidity management of Danish banks.

[2] Real-time liquidity can be important for the banks' participation in payment, clearing and settlement systems, particularly for international banks operating in various currencies in different time zones.

[3] In relation to Chart 47 the LaR curve shows, for each point on the time horizon, the sum of the deterministic outflow tables plus e.g. the 95th percentile value from the market-risk models and the statistical models for market-related and random liquidity effects, respectively.