Yves Mersch: Risk management in times of non-conventional monetary policy
Keynote speech by Mr Yves Mersch, Member of the Executive Board of the European Central Bank, at the Joint Bank of Portugal and European Central Bank Conference on "Risk Management for Central Banks", Lisbon, 25 September 2017.
The views expressed in this speech are those of the speaker and not the view of the BIS.
Let me start by thanking you for coming to this joint ECB-Banco de Portugal conference, and the Banco de Portugal for hosting us on this occasion. I am delighted to see central bank risk managers, practitioners and academics here in Lisbon to exchange views on risk management for central banks. This topic has gained such importance in recent years as central banks across the globe, including the ECB, have sharply expanded their balance sheets, even though the time for ever more of the same is behind us. Like the many great maritime expeditions that started out from this country, some of our endeavours in recent years have taken us into uncharted territory. Now is a good time to take stock of what we have learnt and map out what may come next.
Following stable principles in the Eurosystem's risk management framework
The Eurosystem has a clearly defined objective, enshrined in the Treaty, to preserve price stability in the euro area.1 The Treaty does not speak of eligible counterparties, collateral haircuts or issue share limits for bond purchases. However, you would be mistaken in assuming that managing financial risks has no role in the Eurosystem's monetary policy implementation - especially in the light of the significant balance sheet expansion over the past few years. On the contrary, we have applied the guiding principles of our financial risk management framework in a stable manner. We established these principles long before we embarked on non-conventional policies, not least because we have a statutory obligation to lend only against adequate collateral. We continue to apply them to standard as well as non-standard policy measures.
Broadly speaking, these principles are protection, consistency, simplicity and transparency. This means that - if multiple monetary policy options exist to fulfil our mandate - we should select the measures that minimise our own exposure to financial risks. This idea is sometimes referred to as risk efficiency. These principles also imply that we aim to avoid undue distortions in the markets in which we operate. We trade at market prices and we adopt a risk-equivalent approach in setting our risk control measures. In addition, the principles require risk management to be an integral part of decision-making in the institution. So we are compelled to measure risk in an objective and consistent way across different types of operation. Lastly, our frameworks promote transparency and simplicity, and we champion adherence to these principles by being rules-based ourselves and as predictable as possible in our operations.
We have come together here to discuss risk management for central banks and the lessons from the crisis. But, in addition to looking back, I would like to take this opportunity to map out some fundamental elements that should be part of the future risk management framework, once we have returned to a more conventional monetary policy environment. Of course, these are largely informed by some of the actions we took over the past decade and founded on the principles I have just described. So while we are clearly not in a position to move to this new steady state in one go, its broad features can already be seen today.
Navigating through gusts, storms and crosswinds
Risk managers, including central bank risk managers, are generally prudent. Such prudence is not only necessary to preserve market confidence in the central bank's independence and ability to comply with its mandate, it is also appropriate since we are entrusted with public money. In a crisis, however, the central bank's exposure to financial risks can - and may indeed have to - increase, because it is still expected to fulfil its policy mandate. The non-conventional measures that we implemented were necessary and proportionate to achieve our mandate; they have also contributed to financial stability by reducing financial risks in the euro area. There are, of course, limits to the actions of central banks, as monetary policy cannot address structural or fiscal sources of financial distress, nor can it step in to rectify national deficiencies as our mandate applies to the euro area as a whole. It is important for central banks to strike the right balance in this regard, and for financial risks to be correctly measured, properly understood and sufficiently contained.
Moreover, the European Court of Justice, in its judgment on outright monetary transactions, has confirmed the need to provide sufficient risk management measures that are "likely to reduce the risk of losses to which the ECB is exposed".
Speaking in a country with a proud history of maritime exploration let me recall the words of Seneca, the Younger:
When a man does not know what harbour he is making for, no wind is the right wind.
The "harbour" of a central bank is its mandate. The principles it seeks to adhere to are the "compass" that helps it fulfil this mandate. When a central bank is facing a crisis, it needs to consider whether its monetary policy operations are still proportionate. This need stems from the obligation under primary law (proportionality principle) to properly weigh up all the components of a given measure against one another. The aim is to determine whether the expected benefits of the measure outweigh its costs. As a general rule, a stable risk management approach and a through-the-cycle approach are advisable. However, extraordinary circumstances may require extensions of the framework, such as the acceptance of additional collateral. If this still proves insufficient, unprecedented types of monetary policy measures may call for entirely new operations, accompanied by new risk management measures.
The evolution of the Eurosystem's monetary policy operations in response to the crisis featured all these activities. Our risk management framework has evolved accordingly, guided by our stable principles.
First, the "gusty winds": during the early stages of the financial crisis, we increased the amount of liquidity we provided to the banking system without tightening the risk control framework, despite the onset of stress in the system. In this respect, our risk management framework followed a through-the-cycle approach - thereby avoiding further distress to the banking system. It also stood in stark contrast to the more pro-cyclical nature of the commercial banking sector's risk management approach - and was thus part of the central bank's duty to lean against such potentially self-reinforcing winds.
It was possible to avoid tightening the framework without exposing the Eurosystem to undue risks because the pre-crisis framework had been calibrated in a sufficiently prudent manner. In fact, the framework proved to be robust enough to protect the Eurosystem from losses resulting from counterparties defaulting, such as Lehman Brothers or the Icelandic banks - although the complexity of the collateral in some of these cases led to lengthy liquidation processes. This experience certainly merited a careful review of the collateral rules for complex securities and financial innovations. More broadly, it justified taking steps towards reducing the reliance on external credit ratings, in line with the G20 agenda.
Second, the "storm broke out": after the default of Lehman Brothers, it became evident that such standard measures were no longer sufficient to achieve price stability in the euro area. We therefore provided enhanced credit support and, later on, non-standard credit operations in the light of the increased liquidity demands from the banking system. To avoid potential collateral shortages when satisfying this increased demand, we expanded the range of assets against which banks can draw liquidity from the Eurosystem. But at the same time, we followed the principle of risk equivalence with the assets accepted under the pre-crisis rules.
Let me give you a concrete example: since 2012, several national central banks of the euro area, including the one hosting this conference, have accepted additional bank loans as collateral. To ensure risk equivalence, haircuts on these assets were higher than those used in regular operations.
This brings me to the third phase of the crisis, the "crosswinds": in recent years the continued decline in inflation rates in the euro area called for additional non-standard measures - I am referring, of course, to the asset purchase programmes. This type of liquidity provision was a novelty for the Eurosystem, although the instrument of outright purchases has been envisaged as one of the legal and operational tools of the Eurosystem since its inception. Accordingly, additional risk control frameworks needed to be developed to keep the financial risks in check given the significant rise in exposures.
Our Asset Purchase Program, especially the PSPP, its largest component, is a case in point: the Eurosystem applies strict eligibility requirements. Purchases are subject to issuer and issue limits. We closely monitor the risks of the programme in place and any possible deviations from the benchmarks, e.g. in the CBPP3. Our due diligence can, under very strict conditions, lead to purchases of covered bonds that credit rating agencies rate below our regular eligibility threshold. On the other hand, our due diligence can also lead to additional risk control measures for particularly risky covered bonds.
Given that we have not suffered any impairment since we started the expanded asset purchase programmes, I feel justified in saying that the financial risks have been successfully contained so far.
Maintaining the risk management principles while returning to a more conventional monetary policy
It is clear that, once we have seen a sufficiently sustained adjustment in the path of inflation, we will continue to prudently adjust our tool-box of monetary policy instruments, as we have been doing since December last year. This will also be the moment when we have to carefully review how our operational and risk management framework has deviated from its pre-crisis state. Where necessary, some temporary measures will be rolled back. But we will also consider whether there are areas where we have taken new risks onto our balance sheet - and consequently where certain elements should be kept.
The developments I have just described offer some insights into how the future steady state risk management framework might look. At the moment, I see four key features:
First, while the temporary collateral framework has successfully eased potential collateral shortages, it could be argued that it should not become part of the regular framework. As liquidity demand declines, there will also be less need for an expanded set of eligible collateral. Whether some temporary measures might be transformed into a state-contingent framework, ready to be activated again in the future if circumstances so require, is an open question. As experience shows, each crisis needs a tailored response.
Second, we will need to keep the flexibility to adjust the collateral framework to financial innovations, especially when it comes to 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. In a similar way, we will need to forcefully deal with new types of securities, such as conditional pass-through covered bonds, whose risks may not yet have been fully appreciated.
Third, we will have to retain the risk control frameworks for the asset purchase programmes beyond the horizon of our net asset purchases. As long as we keep outright portfolios on our balance sheet, the reasons behind the risk control measures, including eligibility criteria, purchase limits, benchmarks ensuring diversification and the different risk-sharing agreements, will continue to apply.
Fourth, the Governing Council will have to decide whether we ought to rely more on our own judgement about the quality of assets and counterparties and to further expand Eurosystem-internal credit assessment systems, steps which would require additional resources. It means further enhancing the due diligence on external credit ratings, for which sufficient transparency on the judgement underlying these ratings is essential. And it means using supervisory information to ensure an efficient implementation of monetary policy operations - the introduction of the European single supervisory mechanism has brought about fundamental improvements here, as it allows for the sharing of relevant information within the legal limits of the separation principle.
To conclude, I believe the Eurosystem's risk management framework has successfully weathered the financial storms of the past few years. The reason is that we followed stable principles when the size of our operations, the assets accepted as collateral or the types of operations changed. Looking ahead, we can already start thinking about a financial risk management framework that will be appropriate for a central bank that has returned to a more conventional environment for the conduct of monetary policy, while still following these principles. I now look forward to the panel discussions, which will pick up on some of those issues in more detail, and to hearing your views.
1 Treaty on the Functioning of the European Union.