Whither liquidity? Developments, policies and challenges

Speech by José Manuel González-Páramo, Member of the Executive Board of the ECB28th Nomura Central Bankers Seminar Tokyo, 14 April 2008

Ladies and Gentlemen,

It is a great pleasure for me to be here in Tokyo. I would like to thank the organisers of this important event, the 28th Nomura Central Bankers Seminar, for giving me the opportunity to share with such a distinguished audience some considerations about the role of liquidity in the recent market turmoil.

The current turmoil in financial markets has been to a very large extent about “liquidity”. Since August 2007 the different phases of the turmoil have been marked by repeated disruptions to the traditional mechanisms of liquidity creation and transmission, both at the aggregate and at the individual level.

From the start of the turmoil we have experienced several episodes of contraction in the degree of liquidity of a variety of money and credit markets. Some of these episodes have been experienced in the market for inter-bank unsecured loans, a key component of the money market and the starting point of the monetary transmission mechanism. Moreover, liquidity has also dried up in the market for mortgage-backed securities and commercial paper as well as for structured credit products. And, occasionally, poor liquidity conditions have been reported in markets for securities historically regarded as very liquid and safe, such as the market for banks’ covered bonds or, even, the government bonds of some developed economies.

Indeed, at some points in time over the last few months we have observed rises in order-based measures of liquidity risk for a variety of money and credit markets and declines in traded volumes that could be explained only by the existence of severe stress.

At the individual level, liquidity shortages have taken a toll on a limited number of financial institutions.

In my intervention, I would like to first recall some of the main issues related to liquidity and banks’ liquidity management that have played a central role during the ongoing market turmoil. Second, I will briefly review recent developments in liquidity in credit and money markets. Third, I will describe some of the actions that central banks - and, particularly, the European Central Bank (ECB) ­- have undertaken to mitigate disruptions to the traditional channels of liquidity provision. Finally, I will attempt to draw some preliminary lessons on liquidity management and present some of the public and private initiatives that are currently under discussion in order to reinforce the industry and regulatory framework for liquidity risk management.

The multi-faceted nature of liquidity

Before moving on, it may be worth agreeing on a working definition of the term “liquidity”. Indeed, this term is often used to indicate many different things. This is a problem that has been pointed out by several contributors to a recent volume of Banque de France’s Financial Stability Review on liquidity issues. ^[2] For instance, Charles Goodhart notes in his contribution that the difficulty to define liquidity clearly is due to the multi-faceted nature of liquidity itself. Less charitably, Jean Tirole blames the problem on the “unfortunate habit of economists” to use this term to cover different concepts.

While some economists may argue that the specific meaning of the term becomes clear once it is put into context (particularly whether it is used in relation to macroeconomic developments, financial markets or single agents), it is worth striving for more clarity in the use of the notion of liquidity. Since I realise that this may sound like a criticism to the profession’s practice, let me come clean by admitting to the same unfortunate habit of using the same term for different purposes. Indeed, I have to acknowledge that at the ECB we refer to liquidity to describe conditions and notions at different levels (individual, market and economy-wide) that are distinct from each other. So, as a sign of repentance, let me attempt to clarify these different notions and make the firm resolution to stick to them, at least until the end of this speech.

In the context of the properties of an individual financial instrument, the notion of liquidity typically refers to the ability of trading such instrument (i.e. purchasing or selling it) without unduly affecting its price. In particular, an asset is described as highly liquid, thereby entailing limited liquidity risk, if it is reasonable to assume that it can be easily sold without incurring a loss of value.

The notion can also be used to characterise an entire market, also to reflect co-movement in liquidity across assets. In this case, we refer to the ability to trade or liquidate positions in a market without significantly affecting the prevailing market prices. The degree of liquidity of the market for a financial asset is a function of a variety of factors, including: its outstanding size; the relative size, frequency and modalities of transactions; the number and quality of market participants; the relevance of transaction costs; the amount and quality of information on prices, traded volumes and, more fundamentally, on the degree of security of the asset and of credit-worthiness of counterparties. Thus, a market in which trades can be smoothly executed thanks to the stable presence of numerous buyers and sellers willing to undertake large transactions at narrow bid-offer spreads is generally regarded as highly liquid.

Going back to the issue of the different notions of liquidity, from the start of the turmoil we have also heard reports about liquidity (or, to be more precise, lack of liquidity) of specific individual financial institutions. In these reports, liquidity problems at specific individual banks are usually referred to as equivalent to difficulties in funding their business activities. Indeed, an additional notion of liquidity refers to the ability of a firm (not necessarily from the financial sector):

The degree of funding liquidity of a firm may depend on a number of factors, notably the amount of liquid assets it holds and the liquidity of the credit and securities markets it relies on for its external financing.

The last observation points to the fact that, while the different notions of liquidity (notably, market versus liquidity funding) are conceptually distinct, they are in practice interrelated. Indeed, some authors have developed models in which market and funding liquidity reinforce each other, leading to the emergence of “liquidity spirals” that may account for some of the declines in the liquidity of markets and individual institutions observed in recent months. ^[3]

Aggregate liquidity

Before closing the subject of how to define liquidity, let me briefly recall that there is a more aggregate notion - that should be fairly familiar to this audience of experts on the Japanese economy and Bank of Japan’s Quantitative Easing Policy - that is often used to describe monetary developments at the macroeconomic level. Indeed, liquidity is often used as a synonymous for money supply. In particular, at the ECB we use the term “excess liquidity” to indicate the existence of a stock of money relative to income in excess of the level needed to finance non-inflationary growth in the medium term (as in the quantity theory of money).

Of course, even this notion of liquidity is not entirely unrelated to those mentioned above, since money supply measures are derived from the consolidated balance sheet of the banking sector. However, it is important not to confuse references to liquidity in the context of liquidity management (that refer to the overall funding needs of the banking sector) with the aggregate money holding behaviour of agents other than banks.

Market and funding liquidity risks

As I mentioned above, the liquidity of a financial instrument may vary over time, and may become severely impaired under conditions of stress. Uncertainty about liquidity implies that market participants face the risk that it may not be possible to execute a transaction when they wish to do so due to the lack of counterparties. For instance, a market participant may not be able to sell a security or liquidate a position due to lack of buyers.

Similarly, firms (but also households) face the risk of being unable to obtain enough cash to meet their obligations. This funding liquidity risk is particularly relevant for banks since their core business involves the maturity transformation of funds. Indeed, the traditional activity of banks consists of providing liquidity transformation services, by collecting short-term deposits in order to finance the provision of loans with longer maturities. As a result, banks intrinsically face a maturity mismatch in their balance sheets that exposes them to funding liquidity risks.

Under supposedly extreme circumstances (that the experience of Northern Rock suggests may not be that extreme), the maturity mismatch may even imply the risk of a bank run. This may happen if large numbers of depositors perceive liquidity conditions (or solvency, for that matter) at a specific institution as untenable and rush to withdraw their deposits in waves. Several mechanisms and policies have been developed over time for the prevention and resolution of bank runs. These include deposit insurance schemes (whether privately- or publicly-funded), regulatory and supervisory policies (e.g. on reserve and capital requirements) and the possible recourse to emergency liquidity assistance from the central bank (the “lender of last resort” classic policy prescription).

The “originate and hold” vs. “originate to distribute” models

The characterisation of the liquidity risks faced by banks I have made so far conforms to the so-called “originate and hold” business model, according to which banks create loans to firms and households and hold them to maturity in their balance sheets. However, in recent decades, many banks have taken advantage of financial innovation and of developments in securitisation to progressively shift towards the so-called “originate to distribute” model. Under this new model, banks originate loans but sell them to structured investment vehicles (often set up by themselves) to be repackaged and subsequently sold as asset-backed securities (ABSs).

In principle, this business model is attractive for banks since it provides them with a new source of financing to expand lending (thereby mitigating their funding liquidity risks), while also allowing them to economise on costly capital requirements. In fact, it has been argued that regulatory capital arbitrage has been one of the main reasons why banks have adopted the “originate to distribute” model.

Also from the point of view of the economy as a whole, the “originate to distribute” model could be viewed as having many advantages. First, being capital costly, the ability to sustain a given level of credit supply with a lower volume of capital enables the banking sector to reduce the costs of financing for borrowers and favours financial development. Moreover, it may be seen as representing a step towards more complete credit markets, thereby contributing to enhancing the efficiency of the economic system. In addition, the securitisation of loans in principle could reduce a secular source of vulnerability of the economies, by taking risk concentrations associated with loan portfolios away from the banking sector and spreading them more broadly across other sectors. As a result, the “originate to distribute” model may potentially diminish the likelihood of the credit busts and banking crises that have historically been a major source of macroeconomic and financial instability in many economies.

It should be noted though that, even before the outbreak of financial market turmoil, some commentators have voiced doubts about some of the welfare-enhancing effects of the “originate to distribute” model. For instance it has been argued that, while securitisation certainly spreads existing risks, it may in fact encourage the creation of further risks, particularly by relaxing the incentives for banks to screen and monitor borrowers in order to alleviate the informational asymmetries associated with credit contracts. More generally, it has been questioned whether the securitisation of loans may not perversely shift risk concentrations to sectors potentially more vulnerable than those that originated them.

In this regard, I remember reading a few years ago in the Bank of Japan’s Financial Stability Review that in Japan credit derivatives based on loans originated by large and relatively well-capitalised commercial banks had been acquired by small regional financial institutions with significantly weaker balance sheets. Also in the course of the current turmoil, we noticed that some of the individual European banks affected by exposures to the sub-prime mortgage market were relatively small institutions without the necessary technical expertise or balance sheet strength to withstand them. In addition, for many banks all over the world, risks do not seem to have travelled farther away than the balance sheets of their own conduits, and have been internalised again once the market turmoil erupted.

While this certainly does not mean that securitisation is not potentially welfare-enhancing, it suggests that - in the environment of historically low returns on traditional securities and an implied imprudent hunt for yield that preceded the turmoil – the markets for asset-backed securities and credit derivatives may have failed to spread risks as effectively as expected and have, by contrast, exacerbated information asymmetries, probably as a result of their opacity and the complexity of the underlying contracts.

More generally, the events of the last few months have revealed that, while the adoption of the “originate to distribute” model relieves banks from traditional funding liquidity risks and “liquefies” their lending-related assets, it increases their exposure to the risk of market illiquidity. Indeed, this model relies on the assumption that banks are able to dispose of mortgage loans in the form of repackaged asset-backed securities, which presupposes the existence of highly liquid wholesale markets. As it is by now well known, over the last few months liquidity has become severely limited in a number of markets playing a key role for the execution of the liquidity risk management strategies of the banking sector. This has particularly affected banks pursuing very aggressive strategies of maturity transformation that rely on the continuing issuance of asset-backed commercial paper at short maturities.

To illustrate this point, we only need to recall the start of the current market turmoil. The deterioration of the US sub-prime mortgage market led to the drying up of liquidity in the markets for mortgage-backed securities. These developments led to liquidity funding difficulties at the structured investment vehicles (SIVs) specialised in purchasing mortgage loans from banks and transforming them into mortgage-backed securities for their subsequent sale to investors. Liquidity funding problems at SIVs translated into analogous problems at banks as they were no longer able to dispose of their stocks of mortgage loans, while facing unexpected obligations arising from contingent credit lines and, in some cases, also from the prospects of being forced to incorporate struggling associated off-balance sheet SIVs into their own balance sheets. Uncertainty about future obligations among individual banks led to significantly increased precautionary demand for liquidity, which – together with asymmetric information – resulted in severely diminished liquidity in the money market.

I have anticipated already some of the developments in liquidity from the start of the current market turmoil. Let me now review them more in detail in the remainder of this section.

As you are well aware, in July and early August a series of events led to an intensification of the tensions in the US sub-prime mortgage market and a sharp decline in the degree of risk appetite of global investors. Market volatility increased across almost all financial asset classes. Stock prices declined as investors sold equities and moved funds into safe-haven investment assets, such as government bonds. In this context, several investment funds holding asset-backed securities suspended withdrawals from their clients. At the same time, a number of European banks made their direct or indirect exposures to the US mortgage market, particularly to its sub-prime component, public.

In this environment, market liquidity declined for a number of assets. This was most obvious for the markets directly related to the core of the current turmoil, namely the market for sub-prime mortgage-backed securities (MBSs). However, during the turmoil, the problem of illiquidity spread much further, first affecting other ABSs and asset-backed commercial paper (ABCP), then mortgage-backed securities and basically all structured credit instruments, such as collateralised debt obligations (CDOs).

The market for government bonds is the only fixed income market in which liquidity has remained fairly good. It should, however, be noted that within the euro area there has certainly been a differentiation among different government bonds: while German Bunds have generally remained quite liquid, the liquidity of bonds issued by countries with weaker public finances, such as Italy and Greece, has at certain points in time become heavily impaired. This has also had an impact on the yield spreads between these government bonds and the benchmark Bunds, which in some cases have widened to their highest levels since the introduction of the euro.

In money markets the impact of the turmoil was initially felt mainly in the longer-dated unsecured inter-bank market. It can be said that liquidity in the unsecured deposit markets almost completely dried-up in maturities beyond 1-month (although it should be noted that these were not the most liquid ones even in “normal” times). These frictions eventually spilled over to the very short-term money markets (i.e. below one-week), at first in the US dollar market, where banks – particularly, from Europe – encountered difficulties in raising short-term liquidity.

Early in the morning of Thursday 9 August, the tensions spread to the short-term euro money market (and also to other money markets such as those for the British pound and the Swiss franc) triggering an exceptional liquidity-injecting operation of the ECB. Important factors driving these adverse liquidity developments in the unsecured market were general uncertainty about banks’ sub-prime exposures and the scope for related write-downs resulting in declining lack of mutual confidence within the banking community. Indeed, rising adverse selection together with increased precautionary demand for liquidity, nearly led in august 2007 to the euro money market providing a real-life illustration of Akerlof’s “market for lemons” prediction of how information asymmetries in a market with uncertainty about the quality of goods (in this case of credit claims) may lead to a decline in trading and, potentially, to the collapse of the market.

Interestingly, also the liquidity in the secured non-government repurchase agreement (repo) markets was heavily impacted, as many banks no longer wanted to accept the types of securities mentioned above (MBSs, ABSs at large, and CDOs) as underlying collateral in repo transactions. As a result, only the repo markets based on government bonds remained fairly liquid, but even in these markets the degree of price differentiation across repos secured with different types of government bonds was larger than normal.

Another market segment that saw a partial (temporarily, even severe) deterioration of liquidity conditions is the foreign exchange swap market, which is very important for banks managing liquidity in different currencies. In particular at times when many banks were keen on obtaining USD liquidity, most market participants tried to swap other currencies into dollars, while hardly any bank was willing to provide the USD funds to the extent that the market became dysfunctional for maturities beyond 1 week.

From the start of the market turmoil, the ECB has undertaken a variety of liquidity management operations - in some cases in coordination with other major central banks - in order to mitigate the disruptions to those segments of the money market that are under its direct influence.

The ECB’s liquidity interventions have been accompanied by an intensification of its communication policy, with two main purposes. First, in order to maintain inflation expectations firmly anchored at a time when risks to price stability have been clearly tilted to the upside, it has proven essential to stress the separation between monetary policy and liquidity management policy. In addition, given the rather technical nature of issues related to liquidity management and the relative lack of familiarity of the general public with them, more intense communication has been needed to prevent and correct occasional misunderstandings in the financial press about the scale, modalities and objectives of our liquidity operations. In fact, observers often fail to notice that the magnitude and nature of a central bank’s liquidity management interventions is very much conditional on the structural features of its operational framework. Therefore, impressionistic comparisons among interventions from different central banks, abstracting from the structural differences in their frameworks, do not necessarily serve well the public’s interest in assessing policy responses to the market turmoil.

The Eurosystem’s operational framework

Thus, before explaining what the ECB has done, let me briefly outline some elements of the Eurosystem’s operational framework that have proved crucial to enable it to continue influencing short-term money market rates in spite of the difficult situation in financial markets.

The operational framework of the Eurosystem has been chosen with great care to function both in normal and stressed times. ^[4] It aims at steering short-term money market rates close to the key policy rate - the minimum bid rate in the main refinancing operations (MROs) - in a smooth and efficient manner, with the primary aim of implementing the monetary policy decisions of the Governing Council of the ECB. ^[5] It consists of elements that are regularly employed and those that can be activated should the need arise. In particular, this framework comprises three categories of instruments: 1) open market operations, 2) minimum reserve requirements and 3) standing facilities. ^[6]

Through these instruments, the Eurosystem manages liquidity conditions in the euro area money market so as to steer very short-term interbank money market rates as close as possible to the level of the policy rate decided by the Governing Council.

Eurosystem’s open market operations during the financial market turmoil

In the context of the current market turmoil, the primary aims of the ECB’s liquidity management policy have been: (1) as in normal times, to keep the overnight rate as close as possible to the minimum bid rate; (2) in addition, to ensure the smooth functioning of the money market, also at term maturities, by facilitating access to liquidity to solvent credit institutions, thereby helping them to meet their liquidity needs in a smooth manner. By doing so, the ECB has intended to contribute to re-establishing confidence among market participants and to safeguard financial stability and the appropriate functioning of the monetary policy transmission mechanism.

In line with these aims, during the recent period of financial market volatility, the Eurosystem has implemented a variety of measures, which can be summarized as follows:

First it has made some relatively small adjustments to the modalities of its supply of euro liquidity by changing:

Second , in the context of coordinated central banks’ actions, the ECB has agreed with the Federal Reserve System to grant loans in US dollars to euro area banks. The Eurosystem’s loans have been financed through a currency arrangement (swap line) between the FED and the ECB in connection with the FED’s Term Auction Facility.

Boxes 1 and 2 explain in greater detail these measures. Let me underscore that, in fact, they are all only rather minor adjustments to the existing operational framework for monetary policy implementation, which, has so far, proved to be sufficiently flexible to cope with the recent tensions in the market.

Third, throughout the period of financial market turmoil, the ECB has promptly communicated to the market its liquidity policy orientation and explained its actions via press releases, statements on newswire services and other communications channels (e.g. speeches, articles in the Monthly Bulletin, etc.). These communication efforts have supported the monetary policy implementation operations by reassuring the market on the readiness of the ECB to take adequate measures when necessary.

Eurosystem’s collateral framework

The ability of the Eurosystem to respond to the challenges for monetary policy implementation posed by the recent period of financial market volatility, that represents its most difficult test from the start of Stage Three of Monetary Union in 1999, has to a large extent depended on its collateral framework.

Indeed, the collateral framework is considered to have played an important role in supporting the functioning of the money market and effectively addressing the asset refinancing needs of counterparties, without requiring ad-hoc changes.

In order to understand the role of the collateral framework in the current turmoil, let me recall a key principle of the Eurosystem’s operational framework stating that all credit operations need to be based on adequate collateral. ^[7] The concept of “adequacy” implies, first, by statute that the Eurosystem should be protected from incurring losses in its credit operations and, second, as a policy consideration that sufficient collateral should be available to a wide set of counterparties, so that the ability of the Eurosystem to provide the amount of liquidity it deems necessary for both its monetary policy and payment systems operations should not be impeded by a shortage of collateral in the banking sector as a whole, or even in a significant part of the banking sector. To ensure that this is not the case, the Eurosystem accepts a broad range of debt instruments as collateral, ranging from government bonds, bonds issued by supranational institutions, covered and uncovered bank bonds, corporate bonds and ABSs to non-marketable instruments, such as credit claims.

This wide list of eligible assets - together with the sizeable number of counterparties and the large size of its refinancing operations - has helped the Eurosystem to implement its liquidity management policies during the recent period of decline in market and funding liquidity.

More specifically, by allowing the refinancing of ABSs on a large scale at rates close to ECB policy rates, the collateral framework may have also contributed to financial stability, at least in the short term, by avoiding forced sales of ABSs, which would have intensified the liquidity crisis in the market for these securities.

Two significant developments have been observed during the turmoil with respect to the Eurosystem collateral framework.

Actions of other central banks

As regards other central banks, in general their primary response to the market turmoil has been to intervene to keep short-term money market rates near target rates, through more active liquidity management. Moreover, central banks have aimed to ease the increasing pressure in term money markets by: (1) increasing the viability of standing loan facilities as backstop funding options; (2) boosting the share of longer-term operations in total refinancing provided to banks; and (3) expanding the range of acceptable collateral and eligible counterparties, in those countries where it was needed.

Central banks have also gradually increased their mutual cooperation, initially through enhanced communication and collective monitoring of market developments, and later on by undertaking coordinated actions to provide funds over the year-end. These actions have included the establishment of the earlier-mentioned foreign exchange swaps.

While the current turmoil’s ultimate cause are the substantial losses of financial assets linked to the US housing market, liquidity problems have driven its fever curve ever since its outbreak in the summer of 2007. Diminished liquidity in the markets that play a key role in enabling banks to implement their risk management strategies have led to a decline in funding liquidity at individual banks and given rise to adverse liquidity spirals.

In response to these tensions, the ECB and other central banks have undertaken a variety of liquidity management operations in order to support the interbank money market. However, some observers have argued that market tensions are partly the symptoms of major failures in liquidity risk management by private financial institutions in the past and that, by supporting banks when they get into trouble, central banks may be encouraging moral hazard. For instance, Goodhart has warned that “liquidity management has been passed from the commercial banks to the central bank”. ^[9]

There is no question that moral hazard issues need to be taken very seriously into account when central banks decide on measures to support the banking system in a situation of financial turmoil. Of course, any such support policy, if anticipated, will be taken into account when banks establish their risk management frameworks for “tail” events. Still, one should refrain from generalising and concluding from this observation that central banks should not re-act and support the financial system in the case of severe liquidity disruptions to money and credit markets. In fact, nobody would dispute that the entire economy can be seriously affected if we allow financial market turmoil to develop into a credit crunch or a banking crisis.

In the context of the current turmoil, the primary objectives of the ECB’s liquidity management policy have been from the beginning clearly identified: (1) to steer the overnight rate close to its key policy rate so as implement the monetary policy stance determined by the Governing Council; and (2) to support the proper functioning of the money market as a whole by helping credit institutions to meet their liquidity needs in a smooth manner. By doing so, the ECB has contributed to safeguarding financial stability and supported the effective functioning of the mechanisms through which monetary policy is transmitted in the euro area.

I am fully confident that, once the current turmoil subsides, we will have learned much about how to render the operational frameworks of central banks even more effective in providing access to liquidity in both normal times and under stress. At the same time, I think it is legitimate to expect that banks will make substantial progress in addressing the weaknesses in their liquidity risk management practices that the current turmoil has revealed and that, by doing so, they should make it less likely that history will repeat itself.