Yves Mersch: Banks, SMEs and securitisation
The views expressed in this speech are those of the speaker and not the view of the BIS.
Overcoming the current challenges faced by small firms
Ladies and gentlemen,
It is a pleasure to speak at this conference today. In my remarks I would like to address what I see as the key challenge in the euro area today: how to fix the flow of credit to the real economy.
During the crisis the credit transmission channel has become impaired in three mutually-reinforcing ways. First, as regards quantity, aggregate loan growth continues its steady deceleration and remains very heterogeneous across the euro area. Second, as regards price, low Eurosystem policy rates are simply not being passed on in certain countries. Third, as regards distribution, smaller firms appear to be suffering more from price and quantity effects, especially in weaker economies. Certainly, the reasons for these blockages relate in part to demand-side factors, but the supply-side - banks' capital levels and risk assessments - remains an important part of the story.
A malfunctioning credit channel has several unwelcome consequences for the real economy. It impairs the transmission of monetary policy throughout the euro area. For borrowers in weaker economies that are undergoing a relative price adjustment, tight credit means that real interest rates are rising while relative prices are falling, thus pushing up the real cost of servicing debt. And it also impairs the allocative efficiency of the real economy, as credit is allocated away from the smaller and younger firms that are by nature more risky yet create the most net jobs. As 99.7% of firms in the EU are SMEs, this is a crucial issue to be solved. The fortunes of the euro area economy are intertwined with the financial health of SMEs.
One solution to this problem could be for firms to bypass the bank lending channel and replace bank finance with capital market finance. This trend has already been observed for larger corporates, with European corporate bond issuance expanding rapidly in the last few years. However, for SMEs accessing non-bank finance is often simply not an option. Indeed, non-specialised investors and lenders are often wary of firms facing high degrees of competition and limited growth prospects, particularly if those firms have only existed for a short while. That is why SME lending is currently highly concentrated among a handful of large banks that have the scale and capacity to diversify idiosyncratic risks by investing or lending to a broad enough range of SMEs.
In other words, SMEs will remain largely dependent on a subset of banks for the foreseeable future. The key question then becomes, in the context of an already-accommodative monetary policy stance, how the role of banks in providing SME finance be structurally improved?
I believe there are two channels. First, through the Asset Quality Review transparency will be increased and investors' confidence can be restored. Removing the blockages created by non-performing legacy assets can unclog banks' balance sheets. Once banks' balance sheets are cleansed and strengthened and the market is reassured about the stability of European banks, they should be able to begin lending again. Second, connecting SME financing needs with the funds of bank and non-bank investors via securitisation of SME loans can assist banks' ability to fund and distribute risk. Both channels have the advantage of not interfering with the allocation mechanism of a market economy.
As the Asset Quality Review is already receiving much attention, thus I will focus the remainder of my talk on the second channel.
The benefits of SME loan securitisation
Let's pull together what we know right now. First, we know that SME lending in weak economies is risky. Second, we know that many banks are reluctant to take on additional risk on their balance sheets and indeed are trying to deleverage and reduce their balance sheet risks. Third, we know that non-bank investors in many parts of Europe and abroad are seeking investment assets with maturities and returns that match their liability profiles. I am thinking both of insurers and pension funds, but also alternative investors such as hedge funds. There are also many banks willing to hold ABSs1, but are reluctant to do so, for reasons that I will address later.
So we have a situation where certain banks own assets with risks and capital charges that they do not wish to bear. On the other side, there are investors who are willing to bear those risks. In my view, this is a match that should happen in an efficient financial system, and securitisation of SME loans is the way to achieve it. What is more, the funds received by banks from transferring bundled SME loans to outside investors can help support new lending, also to SMEs, and in turn support further securitisation.
One might rightly ask whether SME ABSs are too risky to deserve a comeback. After all, memories of the US sub-prime mortgage crisis are still fresh in the minds of many market participants. As a result, the conventional wisdom is that securitisations pose financial stability risks and should be largely discouraged by regulators. Thus we have seen a general drive to tighten the framework based on prudential grounds, both at the EU and international level. Let us break down the possible sources of securitisation risk and see what the facts are and whether the conventional wisdom is justified.
First, there is the risk that underlying assets are of poor credit quality. This indeed is a very real possibility, as demonstrated by the actions of many mortgage lenders in the US during the mid-2000s. Even in the EU, where loan underwriting standards are generally higher, loan delinquencies in certain weak economies have reached double-digit levels. Subject to certain legal conditions, securitisation allows originators to remove securitised assets from their balance sheets, which may give an incentive to move off the riskiest loans and package these for unsuspecting investors.
Yet we must also remember that important and sensible counter-measures have been put in place since the 2007 US sub-prime crisis. For example, since 2011 EU lenders are required to retain a share of the resulting ABS, which motivates them to securitise better quality assets. Furthermore, the Capital Requirement Regulation requires that originators apply the same underwriting criteria to loans which are securitised as those applied to loans that remain on the balance sheet. In addition, measures such as the recently-adopted Mortgage Credit Directive will improve underwriting standards across the EU.
But crucially as well, unprecedented levels of transparency have been brought to the EU ABS market thanks to the establishment of loan-by-loan reporting requirements in the Eurosystem and elsewhere. Transparency not only helps expose risks, but also acts as a disciplining device. It reduces information asymmetries between originators and investors. It is now possible for investors to independently analyse the credit quality of the loans and to assess correlation risk across the portfolio - for instance, the concentration of loans to, say, construction firms that are collateralised by real estate. This scrutiny discourages banks to package many of these loans in an ABS. And let me add that transparency requirements are doubly-effective because it is investors who are tasked to perform the due diligence on ABSs and verify that lenders comply with risk retention requirements.
So we see that the worry about underlying asset quality is likely to be confined to a number of increasingly-verifiable cases.
The second source of ABS risk is at the level of the structure itself. It is true that ABSs are complex instruments, involving many parties to a transaction and a range of possible constructions. Many in this room will be well aware of terms such as re-securitisations, synthetic transactions, and revolving structures. These arrangements indeed add complexity, but I must emphasise that the majority of current EU ABSs are static pool, true sale structures, especially the SME ABSs we are interested in. Let me also add that senior European SME ABSs generally have consistently greater credit enhancements than nearly all other senior EU ABSs, certainly so in relation to the delinquencies in their asset pools.
In addition to credit enhancement, I cannot avoid mentioning the role played by the Eurosystem collateral eligibility criteria in influencing the ABS structures we see today. The criteria impose strict conditions, for example requiring a proper legal true sale, allowing only publicly listed senior bonds with a minimum rating, and collateralised by granular homogeneous asset pools. Obtaining Eurosystem eligibility, and thus the prospect of access to funding in an emergency, acts as a powerful incentive for ABS issuers to structure relatively high-quality deals.
Given these asset pool and structural safeguards, it is not surprising that there have been very few European ABSs' defaults. Depending on the study and starting date, since the start of the 2007/08 financial crisis defaults range between 0.6-1.5% on average, against 9.3-18.4% on average for US securitisations. What is more, European SME ABSs are far below both broader EU and US securitisation default rates, with defaults occurring on about 0.1% of instruments. And, even more convincingly, our internal analysis suggests that ABSs eligible for Eurosystem operations have historically had impairment rates even lower than non-eligible ABSs. This suggests again that sensible asset and structural safeguards can go a long way to mitigating the risks of investing in ABSs.
Strengthening the SME securitisation market - looking forward
I have already discussed the potential role that SME ABSs could play in unlocking additional affordable credit for European SMEs. Many SME ABSs, in particular those eligible for Eurosystem operations, are on the lower end of the risk spectrum. The final piece however is to discuss how we can revive EU securitisation markets, including SME ABS markets. As has been widely discussed, both publicly placed issuance and secondary market activity has lagged in recent years.
What can be done to stimulate these markets? First and foremost, I believe it is important to recognise that ABSs come in many shapes and colours, and therefore ABSs can be structurally different in terms of risk profile. From the various statements made by myself and also the ECB-President in recent months, I think it is clear to everyone that the ECB feels that EU ABSs are being treated inappropriately by present regulations and proposals.
What is the issue here? The securitisation capital framework is currently being overhauled at the international level, with proposals being calibrated largely on a single pool of data that does not reflect differences in standards across the world, does not reflect the structural differences across ABS deals, and does not reflect the vastly-different default performance of SME ABSs at the European and global level. Nor can the proposed calibration fully reflect the future benefits of recently-adopted regulatory enhancements. For example, how does one capture the benefit of having loan-level data and more standardised investor reports in a numerical calibration? The resulting strict risk weights in the current discussions may, I fear, not only curtail any growth in EU SME securitisation markets, but further damage the market from its current state.
The revised securitisation framework should reflect the risk-mitigating features of high quality securitisation. It is therefore important that the EU moves ahead swiftly in addressing inconsistencies in the treatment of high quality securitisation.
The EU has already innovated in certain areas, for example introducing preferential risk weights for certain covered bonds. I do not wish to claim that ABSs present as low risks as covered bonds, but it is odd that certain covered bonds can earn preferential risk weights even though they are not subject to the same tight standards regarding transparency and harmonised regulatory frameworks at the EU level, such as ABS risk retention requirements. It is also odd that proposals for insurer securitisation capital requirements would require far more capital than if an insurer held the underlying asset pool directly. These are level-playing field issues that are specific to the EU and, in my view, should be kept in mind when reflecting on the EU-specific securitisation framework. In any event it reasonable to (re-) establish the consistency in the regulatory treatment across asset classes without prejudice to prudential principles.
What would I propose then? This of course needs to be carefully studied but I can already tell you that I favour distinguishing securitisation instruments into several categories, and assigning preferential treatment only to instruments meeting the strictest requirements. The key question of course is what criteria to use, and here I believe that, in order to avoid too much regulatory complexity, a principles-based approach makes sense. Yes, one can also go down the route of detailed calibrations, however from our own experiences with the Eurosystem collateral framework, we have found it is simpler to set out broad eligibility criteria, and then conduct detailed numeric analysis on those counterparties and assets that 'get through the gates'. In fact, I would propose that central bank ABS eligibility criteria could form a useful starting point for identifying 'qualified' ABSs. This is because these criteria are determined using a common risk-tolerance threshold, are widely-accepted by market participants, and are set without conflicts of interest. Of course, we should not entirely rely on central bank eligibility criteria. In this regard I believe the approach recently developed by EIOPA, and partly inspired by the Eurosystem eligibility framework, has many merits, not least being relatively simple while managing to exclude many particularly-risky ABSs.
Having defined criteria for so-called "high-quality securitisation", the next step is deciding what treatment to grant. There is more room to manoeuvre than is generally recognised. For example, supervisory parameters can be adjusted in the current securitisation proposals. Or, there is always the option of adopting straightforward preferential risk weights in a similar manner to covered bonds in the Capital Requirements Regulation. These arrangements could then be reviewed should the Basel Committee reach an agreement on a similar framework. What is key however is that action is taken soon: let us not forget that the investors I mentioned above will not simply grin and bear higher risk weights - they will simply put their money elsewhere, and fund growth in other asset classes that are not being treated so strictly but where investment may not be, from a macroeconomic perspective, as desirable.
There are of course other barriers to the securitisation market not addressed in my remarks, for example the somewhat-arbitrary imposition of caps on ABS ratings by certain credit rating agencies, which are becoming increasing untenable, not only view of asset performance but also because of structural reforms in the euro area institutional structure, not least banking union and the single resolution mechanism. In addition, the Liquidity Coverage Ratio also leaves out some ABS types that could have been included in my view, however taking a step back it seems that an appropriate capital treatment is more important to achieve right now, because with too-high capital requirements European ABSs are unlikely to be created in the first place, much less traded.
Conclusion
Let me conclude.
We have seen that SME ABSs can play a key role in bridging the gap between deleveraging banks and investors seeking to diversify their portfolio. There is growing recognition that, first, European securitisation instruments had a far better performance than their US counterparts and, second, significant policy improvements have been made which deem the regulatory calibration based on the worst performing global transactions and assets as inappropriate. Yet the potential for SME ABSs to provide some funding relief for SMEs is limited in part by the current proposals to overhaul the securitisation credit risk framework, and is also limited by level playing field issues within EU regulations and proposals.
It is not too late to change course: not all EU securitisations deserve the stigma attached to them for the past few years. There is a need to restore coherence across financial sectors in particular amid the unfavourable regulatory treatment of (high quality) securitisation instruments without violating prudential principles.
By doing so we must act fast and in a manner that is sensitive to our own European reality. If the Bank of England and the ECB were to put forward a joint statement on this issue at the forthcoming IMF Spring Meetings, it would underline the European determination to decisively move forward.
1 ABSs in this speech denote both ABSs and RMBSs.