Yves Mersch: Central bank risk management in times of monetary policy normalisation
The views expressed in this speech are those of the speaker and not the view of the BIS.
Risk management has come a long way since Edward Altman introduced the z-score to measure the likelihood of bankruptcy in 1968, and the financial crisis has brought about significant changes in risk management for commercial banks and central banks alike.
But as we now emerge from the crisis, we would do well to reflect on what should persist from that period, and what the “new normal” for risk management should be. So today I would like to discuss what risk management has meant for the ECB in recent times, and what changes we can expect as we approach a phase of monetary policy normalisation.
But before I elaborate on this, I should note that we have entered the quiet period before the next monetary policy meeting of the ECB Governing Council, and therefore my remarks should be understood as high-level reflections and not be interpreted as containing any commitments or comments on upcoming monetary policy decisions.
Parallels between risk management at commercial and central banks
Risk management has gained in importance in recent years for both commercial and central banks.
The regulations that emerged following the crisis have led commercial banks to bolster their risk functions in a number of ways. They are now subject to more detailed and demanding capital requirements, higher standards for risk reporting and, in particular, more detailed rules for the building of internal models. More recently, the Targeted Review of Internal Models (TRIM) was launched to assess banks’ compliance with these requirements and thereby reduce inconsistencies and unwarranted variability in the outputs of their internal models.
The ECB’s involvement in risk management is perhaps most familiar in this context: as a banking supervisor. But the management and measurement of risks has also been of great significance for the monetary policy side of our operations. The ECB, like other central banks, has expanded its balance sheet substantially in recent years, resulting in several changes in our risk management framework. For example, we have expanded the range of eligible collateral for our lending operations and begun purchasing financial assets outright, including a wide array of private sector assets.
While there are many parallels in the way that we and the commercial banks have managed risks, there are also important differences due to our public mission as laid down in our mandate.
First, we conduct a single monetary policy for the euro area as a whole. Though our credit operations and risk mitigation measures are in some ways similar to collateralised lending operations by commercial banks,1 the financial assets we take as collateral and the lending rate we set are the same for all borrowers. That, in turn, requires a risk control framework which aims to achieve risk equivalence across all assets accepted as collateral.
Second, our primary goal is to maintain price stability. So, unlike commercial banks whose fiduciary responsibility is to maximise their financial income, central banks have to consider the wider macroeconomic picture when they set their risk management frameworks.
This is why central banks’ exposure to financial risks can – and may indeed have to – increase in order to honour their mandates, while commercial banks typically aim to reduce risks during crises. In exceptional times, central banks may need to take more risk on their own balance sheets so as to reduce risks for the financial system as a whole. This contributes to financial stability and, ultimately, to price stability.
Still, this is not to say that managing financial risks is not important for implementing the Eurosystem’s monetary policy. Quite the opposite, in fact! Just as a commercial bank must comply with its regulations, a central bank must follow its mandate and the risk management principles therein.2 For us at the ECB, these principles, which were established long before we embarked on non-conventional policies, underpin all our policy measures.
Broadly speaking, the principles are protection, consistency, simplicity and transparency. They imply that – if there are several monetary policy options that we can take to fulfil our mandate – we should select the measures that minimise our own exposure to financial risks. This idea, which underpins all risk management (including in commercial banking), is known as risk efficiency. In addition, our principles require risk management to be an integral part of our decision-making. And we embody transparency and simplicity by being rules-based and as predictable as possible in our operations.
This commitment to risk efficiency is vital for several reasons. First, central bank revenues are public funds, meaning any losses by central banks are losses for the public purse in each euro area country. Second, losses can affect the financial independence of central banks and therefore, potentially, their operational independence. Third, losses can harm our credibility and reputation in the eyes of the public, and thus their confidence in the central bank to maintain price stability.
For these reasons, our principles will continue to guide our approach to risk management in all our policy decisions. But as we now move towards a new phase of monetary policy, it is worth reflecting on what these principles imply for the future risk management framework.
In my view, we should aim to return as closely as possible to the pre-crisis state. But we also need to consider carefully whether some of the temporary measures should remain part of our toolkit. And since we have taken on new risks that will be on our balance sheet for a long time, we may need to retain certain elements of our current risk management framework.
As monetary policy begins to normalise, there are three areas in particular where our risk management framework needs to be reviewed.
Risk management principles while returning to a more conventional monetary policy
The first relates to the changes we made to our collateral framework during the crisis to enable greater access to central bank liquidity.
When we launched the various vintages of our longer-term refinancing operations, we introduced in parallel a number of adjustments to our collateral eligibility criteria. These adjustments contributed to the sizeable take-up of our operations and their effectiveness in reinvigorating the bank lending channel. And maintaining risk equivalence in haircuts meant that broadening the set of eligible assets did not reduce the level of protection for the Eurosystem.
But some of the measures introduced fragmentation into our collateral framework.
Before the crisis, the Eurosystem operated on the concept of a single list. Its purpose was to enhance the level playing field across the euro area, to promote equal treatment for counterparties and issuers, and to increase the overall transparency of the collateral framework. This changed, however, with the introduction of the temporary additional credit claims (ACC) framework in 2012.3
The temporary ACC framework deviates from the single list principle by allowing individual national central banks to specify their own frameworks adapted to their local needs, albeit fulfilling certain agreed minimum risk management requirements. This was acceptable to combat the severe financial tensions and the uneven distribution of collateral in the euro area at the time ACCs were introduced. But clearly, once out of crisis mode, we would not want such a renationalisation of our collateral framework to persist.
So I do not see the case for maintaining national extensions to the common collateral framework in the form they are in today. At the same time, since ACCs represent a considerable source of collateral for our long-term lending operations, there might be a case for retaining them in a different form.
One option would be to return to the fully fledged single list of collateral that excludes ACCs. Another would be to introduce stronger harmonisation into any future ACC framework, which could either be part of the regular framework or part of a state-contingent framework. The key issue is that any future framework should remove the fragmentation we see today.
Other temporary measures introduced during the crisis have less bearing on fragmentation. For instance, we also widened eligibility requirements for collateral, such as for certain asset-backed securities, and accepted non-euro denominated collateral.4 We did all this to achieve a specific monetary policy goal; but once we reach that goal and liquidity demand declines, there should be less need for those exceptional measures to continue. Of course, they will remain “on the shelf” to be used again, as necessary, to fulfil our monetary policy aim.
A central bank should be flexible and may need to have many instruments at its disposal to achieve its mandate. But it should not take higher risks than necessary.
So as we head down the path of monetary policy normalisation, we will have to decide whether some temporary measures need to be jettisoned, included in a state-contingent framework, or transformed into harmonised, more permanent measures.
Since the last of our long-term lending operations will only mature in the first quarter of 2021, this discussion does not need to be concluded today – and many questions are still open. But in any case, changes in this area will involve careful consideration, since experience has repeatedly shown that each crisis needs a tailored response.
The second area where our risk management framework needs to be reviewed is the risk control framework for our asset purchase programme (APP).
We will retain this framework beyond the horizon of our net asset purchases since, for an extended period of time past that horizon, principal payments from maturing securities purchased under the APP will be reinvested. For as long as we keep outright portfolios on our balance sheet, the principles behind the risk control measures, including eligibility criteria, purchase limits, benchmarks ensuring diversification and the different risk-sharing agreements, will continue to apply.
Still, in the reinvestment phase, some criteria and risk control parameters may warrant recalibration. This is to ensure that – given changes in portfolio composition when bonds mature and proceeds are reinvested – overall risk exposure does not increase. Moreover, with significantly lower volumes of purchases and the related increase in operational flexibility, some parameter adjustments may be possible that would actually contribute to risk efficiency gains.
The third area for review is how our counterparty and collateral framework should adapt to a post-crisis financial system. Certainly, in the future we will rely more on our own judgement on the quality of assets and counterparties and consider further expanding the Eurosystem’s internal credit assessment capabilities. The crisis highlighted the importance of having more information on these aspects.
This implies, among other things, further enhancing our due diligence on external credit ratings, for which greater transparency on the judgements underlying these ratings is essential. And it implies making better use of supervisory information. The introduction of European banking supervision has brought about fundamental improvements in this regard, as it facilitates the assessment of relevant information within the legal limits of the separation principle.
Moreover, we will have to balance the aim of returning to the simplicity of our previous framework with adapting to the new realities of the financial system. For example, we will need to keep the flexibility to apply the collateral framework to financial innovations, especially complex new financial products. The new “simple, transparent and standardised” securitisation regulation is a case in point. It will allow us to better assess the collateral we accept.
At the same time, if financial innovations simply present new types of risks, we will not be so accommodating. This is also a key lesson of the crisis. We will forcefully deal with new types of securities whose risks may not yet have been fully appreciated.
Conclusion
Let me conclude.
Thanks to our stable principles, the Eurosystem’s risk management framework has successfully weathered the challenges of the financial crisis. The size and type of our operations changed, as did the assets we accepted as collateral. But our principles stayed the same.
Like risk management in the banking sector, central bank risk management has to evolve with the times. So we need to reflect on where our principles will lead us in a post-crisis landscape. Most importantly, we need to start thinking about a financial risk management framework that will be appropriate in an environment of more conventional monetary policy.
While the benchmark for this future framework should be the pre-crisis state, it is not clear whether we can return entirely to the previous status quo. Instead, we might have to apply what we have learnt from the crisis, retain what is useful for the future, and leave behind things whose time has passed.
This will enable us to rely on a framework that is transparent and robust, but also flexible enough to deal with the challenges of the future.
1 ECB (2015), “The financial risk management of the Eurosystem’s monetary policy operations”.
2 In 2010, the Governing Council approved a set of “Principles of the risk control framework for the Eurosystem’s credit operations”.
3 The ECB’s Governing Council also decided in December 2011 to enable national central banks to accept as collateral performing additional credit claims (ACCs) that satisfy specific eligibility criteria. Such assets do not belong to the single list and they are also subject to a separate ECB Guideline: Guideline of the ECB of 9 July 2014 on additional temporary measures relating to Eurosystem refinancing operations and eligibility of collateral and amending Guideline ECB/2007/9 (recast) (ECB/2014/31).
4 This expansion included certain short-term debt instruments and derogations for own-use of government-guaranteed bank bonds. Non-euro denominated eligible assets included those denominated in pound sterling, US dollars or yen.