Ladies and Gentlemen:
It is a pleasure for me to attend the SEACEN-CEMLA Conference, as it offers me a new opportunity to describe and explain to my esteemed colleagues from other parts of the world the experience of the ECB’s monetary policy over the past four years of crisis. I would also like to take the opportunity to discuss with you what I consider to be some of the main challenges that modern central banking will have to face in the future.
With the benefit of hindsight, I should say that the implementation of monetary policy was relatively straightforward during normal times. The Governing Council of the Eurosystem decided on the level of the key ECB interest rates, following a comprehensive assessment of the outlook for price stability and the associated risks, based on its economic and monetary analyses. The Executive Board of the ECB then organised the refinancing operations with eligible counterparts in order to implement the decision of the Governing Council. And it did so in a relatively “standard” fashion, employing a relatively small set of instruments in a rather predictable way.
Through its refinancing operations, the ECB aims at supplying the amount of liquidity that is necessary for a smoothly operating banking system, influencing short-term money market interest rates to align them with the policy stance signalled by the Governing Council. Through the money market yield curve, the monetary policy stance is transmitted to the pricing of financial instruments, and credit conditions more generally. These conditions in turn influence saving and investment decisions and monetary dynamics in the economy which finally also influence price developments in the euro area.
The operational framework is the initial link between the key ECB policy rates and money market rates. It supports the fulfilment of the ECB’s mandate for price stability by letting monetary policy decisions feed through as quickly and precisely as possible to short-term money market rates.
Let me now explain in greater detail the many challenges that the ECB had to face when implementing its monetary policy during the financial crisis.
The assessment of the monetary policy stance is essential in order to be able to make the right interest rate decisions at the appropriate time. In the case of the ECB, this assessment has two elements: i) the formation of a view on the medium-term inflation outlook and, in particular, the risks to price stability, based on the interaction of supply and demand in various markets (economic analysis); and ii) the identification of the monetary impulses from current and past monetary policy decisions taking into account that these impulses are always transmitted to the economy with a certain lag (monetary analysis).
Needless to say that assessing the monetary policy stance in real time is always a challenging exercise. This is even more so in turbulent periods when there is a high degree of uncertainty surrounding current and future economic conditions and the functioning of the economy, including the transmission of monetary policy itself. The assessment of the appropriate monetary policy stance must therefore be broad-based and forward-looking, encompassing all the information relevant to the formation of a view on the risks to price stability over the medium term, including the potential impact on price developments of risks to the growth outlook.
Monetary policy decisions may only affect the developments of economic variables in the intended way if financial markets efficiently transmit the initial monetary impulses along the yield curve. This requires that money markets, as well as broader financial markets, are functioning properly. If this condition is not met, it may become necessary for central banks to take action to ensure that the monetary policy transmission channel continues to work effectively, while at the same time continuing to assess the appropriateness of the monetary policy stance given changes in the economic environment. If the central bank did not act to prevent markets from becoming dysfunctional, its decisions over monetary policy rates would remain purely theoretical. Its decisions would have little or no impact on the real economy and the impact itself would become increasingly difficult to predict. This is precisely what led the ECB to implement various ‘unconventional measures’ during the different stages of the financial crisis. I will elaborate more on this, as well as on the new measures that the ECB announced on 6 October, in a moment.
By construction, the monetary policy framework of the ECB (i.e. the decision on the appropriate level of interest rates and the management of inter-bank liquidity conditions needed to implement this decision) implies a separation between monetary policy decisions, on the one hand, and the implementation of the desired monetary policy stance through open market operations, on the other. A clear distinction therefore has to be maintained between interest rate decisions taken to preserve price stability (i.e. the determination of the monetary policy stance) and liquidity decisions taken in the course of implementing this stance.
This distinction is useful both under normal market conditions and in periods of financial turbulence, albeit for somewhat different reasons. In normal times, this so-called ‘separation principle’ allows the central bank to fine-tune operational procedures in order to steer short-term interest rates very close to the main policy rate, without risking that those adjustments are regarded by observers as changes to the monetary policy stance. In exceptional circumstances, when severe distortions in financial markets emerge (reflected, for instance, in volatile and rising money market spreads), a more non-standard liquidity management by the central bank may be needed without jeopardising the central bank’s main monetary policy objective. Against this background, the so-called ‘unconventional measures’ implemented by the ECB during the crisis still aim mainly at implementing the monetary policy stance by contributing to restoring the smooth distribution of liquidity in the money market. It is along those lines that the ECB’s actions during the ongoing financial crisis must be understood.
It is the view of the ECB that the role of a central bank under any circumstances, and in crisis times in particular, is to inflexibly pursue its main objective, which in the ECB’s case is price-stability, and to perform as a key anchor of stability. This implies very careful analyses of any sudden unexpected shocks before considering a change to its monetary policy stance. In the same vein, the ECB also must assess in real time the effectiveness of its actions in order to be able to decide at the right time whether or not to phase out the exceptional measures implemented to cope with exceptional events. This is the main reason why the ECB never pre-commits to any particular future policy stance or measure. Given the high level of uncertainty that exists in a crisis environment, any pre-commitment would risk persisting with a particular policy stance even when this stance is judged to no longer be appropriate by the ECB on the basis of its assessment of the most up to date information.
As regards the additional liquidity providing operations introduced by the ECB during the crisis, let me mention that these non-standard measures have been enacted at positive interest rates throughout the financial turmoil, even though the level of policy rates was close to a lower bound. In normal times, remunerating reserves allowed banks to have convenient buffers to face liquidity shocks at low costs. In crisis times, it proved to be insufficient. Nevertheless, the link between the policy rate and the remuneration rate on required reserves allowed us to conduct non-standard measures with positive (non zero) policy rates, and hence without affecting the costs of keeping high level of reserves at the ECB. In addition, the rate that the ECB pays on cash placed overnight on our deposit facility in effect sets a floor for short term money market rates. Throughout the crisis, the EONIA, the index for euro-area overnight money market interest rates, always fluctuated at levels above the deposit facility rate. Especially in periods with a very high degree of liquidity provision, EONIA tended to be quite close to the deposit facility rate, but the difference remained always positive. Consequently, it was possible in the euro area to conduct non-standard measures at low, but positive short-term interest rates consistent with a reasonably functioning inter-bank money market.
Let me now turn to the details of the measures taken by the Eurosystem during the past four years of financial market turmoil.
The implementation of the ECB’s monetary policy during the crisis can be divided into four phases in the period between 9 August 2007 and October 2011. 
The first phase – which we can call the ‘Market Turmoil’ phase - runs from 9 August 2007 - the day the French bank BNP Paribas announced that it suspended the redemption of shares in its funds that had invested in sub-prime mortgages - until 15 September 2008 - the day Lehman Brothers failed. The main challenges faced by the ECB during this period were to calibrate adequately the "right" amount and timing of frontloading of liquidity, to be able to continue steering the overnight interest rate without relying on "too" frequent interventions. This was done in an environment where the reserve averaging mechanism – which under normal market conditions ensures a stable interest rate path – was no longer sufficient to stabilize money market conditions. There were various reasons for this, including the hoarding of liquidity by banks, heightened aversion by banks to counterparty credit risk and increased cross border market segmentation. In addition to these factors, liquidity in the inter-bank market was also negatively affected both by the idiosyncratic liquidity uncertainty faced by individual banks and the aggregate liquidity uncertainty faced by the banking system as a whole. During this phase, the ECB also needed to determine the appropriate split in the provision of liquidity between short- and longer-term refinancing operations subject to market participants’ preferences.
The second phase - which we can call the ‘Financial Crisis’ phase - covers the last quarter of 2008 and the first three quarters of 2009. This period was marked by a sharp contraction in world output and trade followed by a sluggish recovery. In fact, in the autumn and winter of 2008/09, macroeconomic developments tracked very closely those observed at the start of the Great Depression of the 1930s. During this period the Eurosystem implemented ‘non-standard’ monetary policy measures. Against the background of a sharp contraction in World output and trade and in reaction to increasingly dysfunctional money and financial markets, the ECB took bold steps to support the banking system which is at the core of the monetary policy transmission mechanism in the euro area. In terms of conventional monetary policy the ECB cut the minimum bid rate (MBR) in a sequence of steps from 4.25% to a record low of 1%. In terms of non-conventional measures the ECB announced that it would conduct all of its refinancing operations at a fixed rate (the MBR rate) with full allotment. At the same time the ECB decided to narrow the interest rate corridor of its standing facilities. The eligibility criteria for collateral was temporarily extended and the credit threshold for eligibility was lowered from A- to BBB- for marketable assets (except ABS) and non-marketable assets (with an additional haircut). In July 2009 the ECB also launched its first Covered Bond Purchase Programme (CBPP) with the intention of supporting a financial market segment that is particularly important for the longer-term funding of banks and the financing of the real economy in the euro area and which had become dysfunctional during the crisis. The ECB also strengthened its international coordination with other central banks through initiatives like re-activating US dollar liquidity providing operations. These type of operations had already been implemented in December 2007, following the establishment of bi-lateral currency swap agreements with the Fed, but in 2008 they were offered under the fixed-rate full-allotment tender modality and with longer maturities.
These measures were implemented with a view to re-establishing market functioning so that the transmission of monetary policy would continue to work. The fear was that the malfunctioning of financial markets would lead to a deterioration of credit conditions and spill over to the real economy. Indeed, as recent ECB research shows, banks with weaker core capital positions, and greater dependence on market funding and on non-interest sources of income, restricted their loan supply more strongly than other banks during the crisis (see Marquez and Gambacorta, 2011).
The third phase, which we can call the ‘Phasing-Out’ phase, spanned the period between December 2009 to April 2010. During this phase the Eurosystem announced and initiated the gradual phasing-out from the ‘non-standard’ policy measures. The ECB began this process by not renewing the one-year longer-term refinancing operations and other supplementary refinancing operations with maturities of 3- and 6-months, which reduced the liquidity surplus in the interbank market as the old operation began to expire. On the collateral side, eligibility requirements for ABSs were tightened (in December 2009) and the expiry date for the other temporary collateral measures, like the enlargement of the list of collateral assets and the reduction of the rating threshold, was initially set for the end of 2010. In addition, the supply of USD liquidity providing operations was discontinued and variable rate tender procedures were reintroduced for regular 3-month refinancing operations.
The intensification of the sovereign debt crisis, which had started in late 2009 and peaked for the first time in the spring of 2010, however, led to a re-assessment by the ECB of the appropriateness of embarking on the phasing-out process. This re-assessment marks the beginning of what one could call ‘ the fourth phase’, or the ‘sovereign debt crisis’ phase, in which we are still in. In early 2010, the sovereign bond spreads of several euro area countries, relative to German bonds, widened sharply. Even though these developments could be interpreted as the result of growing market concerns about the sustainability of public finances in Greece, clear signs of contagion to other euro area bond markets emerged towards the spring of 2010. This contagion affected Ireland and Portugal and, eventually, also Spain and Italy. Sovereign bond markets in an increasing number of euro area countries began to become dysfunctional. This weakened the transmission channel of monetary policy decisions to financial markets and broader financing conditions in the economy. In order to address this problem, on 10 May 2010, the ECB launched the Securities Market Programme (SMP) under which it conducts outright purchases of euro area debt securities in the secondary market. Although the SMP remained inactive for a period of around four months between April and August 2011, following renewed tensions in euro area sovereign bond markets, the ECB announced on 7 August 2011 that it would resume purchases under this programme. This took place after the announcement of new fiscal and structural policy measures by the Italian and Spanish governments.
A key distinguishing feature of asset purchases made under the SMP is that their liquidity impact has been sterilised through the conduct of weekly liquidity absorbing operations. Overall, there has been no net injection of central bank liquidity to the market as a consequence of these operations. These measures and their objectives are therefore fundamentally different from quantitative easing.
On 6 October the ECB announced a number of important additional non-standard measures. First, the Governing Council decided to conduct two new longer-term refinancing operations (LTROs), with maturities of 12 and 13 months, with the fixed-rate full-allotment tender procedures. The rate in both operations will be fixed at the average rate of the main refinancing operations over the life of these respective LTROs.
Second, main refinancing operations (MROs) will continue to be conducted with fixed-rate full-allotment tender procedures for as long as necessary, and at least until 10 July 2012. This procedure will also remain in use for the special-term refinancing operations with a maturity of one maintenance period, which will continue to be conducted for as long as needed, and at least until the end of the second quarter of 2012. The fixed rate in these special-term refinancing operations will be the same as the MRO rate prevailing at the time.
The six LTROs with maturities of three-months that will be conducted in the first half of 2012 will also be conducted with fixed-rate full-allotment tender procedures. The rates in these three-month operations will be fixed at the average rate of the MROs over the life of the respective LTRO.
Third, a new covered bond purchase programme (CBPP2) will be launched. Under the CBPP2 the Eurosystem will purchase, in both primary and secondary markets, covered bonds for a total value of EUR 40 billion over a period on one year beginning in November 2011. Further details on the modalities of CBPP2 will be announced after the Governing Council meeting of 3 November 2011.
After four years of financial market turmoil, market conditions in many countries, including the euro area, remain exceptionally fragile. In response, the Eurosystem continues to provide support to euro area money markets in the ways I have just described. Such support should prove beneficial if three principles are followed: it should be timely, targeted and temporary.
The ECB’s interventions have been timely since the outbreak of the crisis, in the sense that on the day in which tensions from the US sub-prime mortgage market first spilled over to euro money markets, the ECB reacted quickly by temporarily offering unlimited amounts of liquidity to the banking sector, to avoid that liquidity tensions would spill over to other segments of the financial markets and to maintain control over short-term unsecured interest rates. More recently, the decision to phase out, or to introduce new non-standard measures, was also taken in accordance with what was judged to be appropriate under the market conditions that prevailed at the time.
The Eurosystem’s response has also been targeted. Our response has, for example, focused on the banking sector, because this sector continues to play a major role in euro area financial markets, in particular for the transmission of monetary policy.
Finally, the ECB has repeatedly emphasized the temporary nature of its non-conventional measures. Indeed, maintaining measures which have outlived their usefulness for too long could trigger other types of imbalances which could contribute to a new crisis in the future. For example, if support for the banking sector is provided for too long, banks’ incentives to resume market-based funding activities may suffer and markets may become increasingly dependent on central bank funding. Incentives for banks to monitor each other may also be lost, and this could result in further market inefficiencies. For these reasons, the Eurosystem tries to keep its interventions in financial markets at a low level that is still commensurate with its main objective. This is the so-called market principle.
The Eurosystem will eventually also have to face the challenge of timing the process of phasing out its non-conventional policies appropriately. When the time comes, it is important, in my view, that our approach to monetary policy implementation is adapted in a gradual manner to normalizing market conditions. This will allow the central bank to better fine-tune its exit steps, and to ensure that the withdrawal of its support measures does not lead to renewed market tensions.
Once market conditions have improved to the point at which the monetary policy transmission mechanism can function without additional central bank support, monetary policy implementation will be able to revert to a more conventional format. I would like to emphasize, however, that the decision to scale back the non-standard measures can be taken separately from any decisions on the level of interest rates that are set by the ECB.
In a gradual phasing-out, the ordering of different steps will need to be defined carefully, and in line with market developments. When the time comes, the Eurosystem will also need to define clearly what the appropriate design of the operational framework in a post-crisis world should be.
In the years, before 2007, the framework performed very well, as it was able to achieve stable short-term money market rates close to the minimum bid rate of our main refinancing operations. This was achieved with a relatively rules-based implementation framework that did not require frequent ad-hoc interventions by the Eurosystem.
Would the same framework achieve equally good results in a post-crisis environment? This will depend on whether the financial landscape will change in ways that are important for the transmission of monetary policy. I would like to mention a few possible developments.
First, as the annual ECB money market surveys show, wholesale interbank unsecured lending activity has declined during the financial crisis. It is possible, that unsecured lending will continue to account for a low share of overall lending in the interbank market, also in the years to come. This could affect monetary policy implementation in two ways: first, the steering of short-term unsecured rates constitutes the first step of the monetary policy transmission mechanism. This mechanism might be weakened, should unsecured interest rates become less relevant in determining overall access to credit. Second, it could lead to more volatility in this market and thus make the steering of interest rates less precise. One possible hypothetical response to such developments could be that central banks give greater prominence to secured and repo market rates.
As a second possible development, market participants may retain a higher level of risk aversion than was the case before the crisis. This could lead to higher risk premia and continued market tiering. Lending rates in the money markets may also become more dispersed, depending on the credit quality of the bank that is borrowing. This would render the task of the central bank more complex.
Finally, the interaction between the new Basel III liquidity standards and monetary policy operations is likely to be significant and complex. Central banks’ policies can influence the ability of banks to comply with the new liquidity risk regulation, because this regulation is likely to affect banks’ recourse to the central bank’s operational framework. For instance, in jurisdictions where the set of assets eligible for central bank operations is broader than the one qualifying for the fulfilment of the liquidity standards, one may witness an increased reliance on central bank refinancing using less liquid assets as collateral.
The question of whether central bank operational frameworks should support, or not, the fulfilment of banks regulatory liquidity standards is certainly not straightforward. The interaction between liquidity regulation and central bank operations will certainly be pervasive. Indeed, central banks are by definition the only agents which are not affected by liquidity risk and which generally continue to be considered free from default-risk during a financial crisis. For these reasons, and as they fulfil a public mandate, central banks act as lenders of last resort in financial crises, providing liquidity to the banking system in their jurisdiction. Liquidity regulation is about ensuring that banks remain liquid also in a stress situation to avoid the negative externalities of liquidity hoarding and the fire sale of assets. Therefore, the lender of last resort function and liquidity regulations in some way share the same goal, although none of them can by themselves be considered sufficient to address the issue of financial instability caused by liquidity stress. They will also interact in the sense that banks will factor in both liquidity regulations and the last resort function in their optimisation behaviour. The two must therefore not be treated in isolation.
The possible impact of the new liquidity standards on central bank operations and related relevant market segments (including those relevant for the transmission of monetary policy) is currently being analysed carefully by the ECB.
To sum up, the financial crisis has brought many challenges for central banks in general, and for the ECB in particular. More specifically, the conduct of monetary policy had to take into account an increasingly large number of uncertainties while liquidity management by the central banks had to adjust flexibly to implement monetary policy decisions in an environment of distorted money market segments.
The flexibility of the Eurosystem’s operational framework has proved invaluable during the crisis, and it will allow the Eurosystem to adjust, if necessary, to a possibly change in the financial landscape.
Of course, the recent sovereign debt crisis in some euro area counties poses additional challenges, also for the Eurosystem. In this regard, it is essential that prudent fiscal and economic policies are employed in all euro area countries. This will reduce bank’s current over-reliance on central bank funding and facilitate the ECB’s task of delivering price stability. In this context, I would like to emphasize that the central bank must remain vigilant in order to ensure that its enhanced intermediation role during crisis times is not abused by fiscal authorities and market participants, as it remains essential that they strictly fulfil their own responsibilities. This approach means that market participants must also adopt a pro-active liquidity management policy and accept trading with each other again. It also implies that different public authorities, like regulators and governments, should also take the necessary measures to stabilise the financial system, so that it can play its part in the recovery of activity and employment.
Thank you for your attention.
Cassola, N., A. Durré and C. Holthausen (2010): “Monetary Policy Operations – Experiences durint the crisis and lessons learnt”, in Approaches to Monetary Policy Revisited – Lessons from the Crisis, Proceedings of the Sixth ECB Central Banking Conference, 18-19 November 2010, p 280-321.
Eisenschmidt, E. and C. Holthausen (2010): “The minimum liquidity deficit and the maturity structure of central banks' open market operations: lessons from the financial crisis”, European Central Bank Working Paper 1282.
Gray, S. (2011): “Central bank balances and reserve requirements”, IMF Working Paper WP/11//36.
Marqués-Ibáñez D. and L. Gambacorta (2011), “The bank lending channel: lessons from the crisis”, Economic Policy 26, 66, 137-182.
I would like to thank Conception Alonso, Alain Durré and Cornelia Holthausen for their inputs to this speech.
See, for example, Cassola, Durré and Holthausen (2010).
Reproduction is permitted provided that the source is acknowledged.