Per Callesen: A central banker's perspective on European regulatory reform
Speech by Mr Per Callesen, Governor of the National Bank of Denmark, at the International Capital Market Association's (ICMA) and Nordic Capital Markets Forum's (NCMF) Annual Conference on Regulatory Reform and Nordic Capital Markets, Copenhagen, 18 March 2014.
Thank you for the invitation!
For this topic of European Regulatory reform I have chosen to highlight, first a number of the most prominent financial regulatory issues in Denmark over the past year that are also related to the broader European agenda, and second, the implications of the European Banking Union going forward.
Capital and liquidity regulation
At the heart of Basel III and the implementing legislation in the European Union (CRR/CRDIV) has been stricter regulatory requirements of capital and liquidity.
We have come a long way in many countries, including in this country, to ensure better capitalisation of banks. Authorities have raised capital requirements by implementing the EU legislation, and in several instances going somewhat beyond this by taking advantage of the room for national discretion included in the directives. Perhaps even more importantly, bank themselves have taken steps to become substantially better capitalised, often going well beyond the regulatory minimum requirement.
In recent surveys, Danish banks stand out as top performers in a European context when it comes to early compliance with fully phased-in stricter minimum rules for capital. This is wise. The largest Danish bank did very recently raise new additional tier 1 capital at a fairly low price, a price which was lower than recent, similar issuances by other EU banks. This confirms the old textbook lesson that solid capitalization is a cornerstone for lower funding costs.
Had banks across our continent had similar levels of capitalisation before the financial crisis struck as our Danish banks do today, I am confident that the crisis would have been a rather different one and also the international debate and actions for reregulating banking would have been quite different.
Should banks, well advised and at the broader international level, decide to move further in the direction of increasingly adding distance to the difference between actual capitalisation and regulatory requirement, interesting new aspects may come up. First, the financial system would be safer. Second, funding costs are likely to come down. Third, other regulatory and supervisory practices may possibly become less intrusive over time.
Liquidity regulation, especially the liquidity coverage ratio (LCR), which measures banks' ability to withstand short-term severe stress in funding markets, has been subject to extensive discussions over several years. The LCR will become binding for financial institutions across the EU from January 1, 2015. At the heart of the LCR is a requirement that banks should hold a buffer of high quality liquid assets (HQLA) to survive a 30-day period of market stress. The technical details are very complicated, but the basic idea is fairly simple and reasonable.
What can become a problem, however, is the definition of what constitutes high quality liquid assets. If that definition is too centralised, too rigid and based on an implicit assumption that all assets under the same formal classification perform equally well in all markets - then we will have new financial problems.
What we need is portfolios of liquid assets which are well diversified, and consist of assets with proven market liquidity and credit quality. We need liquidity standards which will facilitate a self-sustaining banking sector and not adding further to its reliance on central bank facilities, or adding further to the interdependence between banks and the sovereign. We need assets with a broad and diversified investor base.
In that respect it is surprising to see how strong the push has been towards a definition of liquidity which is heavily reliant on government bonds and central bank facilities, for example the Basel Committee proposal that at least 60 per cent of bank liquidity should include only such assets. If government bonds were defined as the only high quality liquid assets, the link between banks and sovereigns would be reinforced, especially in the absence of risk weights for such bonds and if there are no diversification requirements.
As central bankers, we are obviously in favour of prudent fiscal policies and a low level of public debt. But the combination of the current low Danish public debt and liquidity regulation insisting on government bonds constituting a dominant large share of banks' liquidity buffers would hurt liquidity in the Danish government bond market. The answer should obviously not be to increase public debt.
The solution is not difficult. Denmark has one of the largest and most liquid covered bond markets with a stable institutional framework and a very strong track-record of market liquidity and credit quality, including throughout the entire financial crisis. This track-record has been confirmed in a recent comprehensive study by the European Banking Authority (EBA). Danish covered bonds are currently used as the main instrument of bank liquidity management and there is no reason why they should not continue to serve this role. They are not so-called inside money (claims on other financial institutions) but claims on non-financial companies and, mainly, private households.
We are confident that the European Commission will take these issues into account when deciding upon liquidity definitions within the next few months.
The European Commission has recently forwarded another proposal, namely on the potential structural separation of large banks. As you may know, that proposal is seeking a "mandatory ring-fence", that is a separation of bank lending facilities from proprietary trading and certain other trading activities. Separation may be useful in some countries, whereas in other countries, such as Denmark, there is not much of a convincing case for such separation. Problems in our banking sector came from traditional lending activities, not proprietary trading. The broader debate over the merits of structural separation will continue. My point at this stage is only that due notice should be taken of the potential dilemma between on the one hand asking banks to hold larger amounts of highly liquid assets, and on the other insisting that such assets should be managed in separate corporate structures.
Maturity extension for mortgage covered bonds
As regards the Danish mortgage bonds, you may have noticed a public debate over the past few months, as well as a law passed last week, on the introduction of mandatory extension of mortgage credit bonds under certain conditions.
The underlying issue is the following. Over the last decade, Danish mortgages became increasingly reliant on long-term loans with a maturity up to 30 years, financed by the issuing of short term bonds, down to a maturity of 12 months. Borrowing costs are consistently passed through directly to borrowers - so called "match-funding". And demand for these bonds have been consistently very robust and high - for example, market rates of the 12 month bonds have on average been below 30 basis-points during the past year. But the system has been short of a comprehensive answer on what would happen in the extremely unlikely, but also rather scary, event of a failure of the now frequent auctions of new short-maturity mortgage bonds.
The solution has been to introduce a mandatory extension of mortgage bonds with shorter maturities than the underlying loan in case an auction would fail, or if the interest rate has increased by more than 5 percentage points in the year preceding the previous refinancing. In such a case, the bonds will be extended for a year with an interest rate equal to the interest rate at the latest refinancing plus an add-on of 5 percentage points. A comparable extension could happen in case a mortgage institution would become insolvent. To get the proportions of the debate right, it is key to remember that no Danish mortgage institution has failed for more than 200 years, and instances of a 5 percentage point increase in interest rates over a 12 month period have only been observed at very few instances during the last 150 years. But the substantive issue of refinancing risk has been taken care of.
One discussion has been whether the 5 percent trigger was needed or whether one should just refer more generally to the very unlikely event of a failed auction. The point is however that identifying a failed auction in practice is not very easy. A discretionary decision by authorities would probably be needed. The interest rate trigger is then a practical criterion, also allowing investors to price any such risk.
Another discussion has been if, and to what extent, the new law would add to the cost of the bonds. Will there be an additional risk premium for investors? As a first approximation, the answer is yes, namely the risk of having a maturity extension at a rate of an additional 5 percentage points in cases where the market rate increases by, say, 6 or 7 points. Our estimate is that the effect is less than 10 basis-points at the point of issuance, and probably smaller. Note however, that the new law provides a high degree of certainty for investors. This is a material change compared to the previous situation, where investors in the case of extreme events were bound to have open ended speculation on the possible actions of public authorities. This new certainty runs counter to a higher risk premium stemming from the new law.
A third discussion has been on whether a possible self-reinforcing negative market valuation effect could come into play in cases where markets rates would be approaching the five per cent trigger point, say by increasing by 4 percentage points. One answer is that such an effect would be rather limited, since the maturity of the extended bonds is so short. Another answer is that an increase in interest rates of 5 percentage points is so extreme that many other reactions would be triggered, rendering the final outcome extremely hard to estimate. For example, much larger uncertainty in markets could actually reallocate demand for bonds from the longer maturities to the shorter ones. Another possible effect would be a flight of borrowers out of the shorter maturities, in particular if they can thus repurchase the bonds at a market discount, implying that not only demand, but also supply of the short maturity bonds will be reduced, and possibly sharply so.
At some stage the public discussion of these issues ran out of proportion with the limited magnitude of these effects. Now, increasingly, investors and observers understand and acknowledge that the new law is a credit-positive event - complementing the existing safeguards and track-record of the mortgage system.
Finally, to the issue of Banking Union. Preparation by the ECB for the Single Supervisory Mechanism is well advanced. Regarding the Single Resolution Mechanism (SRM) and the setup of a sector-financed joint resolution fund, the window for reaching a final compromise between the co-legislators, the Council and the European Parliament, is closing fast due to the coming EP elections. Still, it should be in the interest of all parties to reach a compromise solution. I believe and hope that it will be done.
The Banking Union is a major change in the European financial regulatory framework. And the impact will be felt widely in EU countries, whether they participate in the union or choose not to. Danmarks Nationalbank finds that it would be to the advantage of Denmark to join.
Yes, the trigger event for leaders deciding to launch negotiations on a banking union construction was the sovereign debt crisis in a number of euro area countries combined with the weak banks and the interaction between banks and the sovereign.
But the broader reasons for banking union were around even before the financial crisis. Key are the need for a level playing field in the single market for financial services, the need for a comprehensive framework for crisis-management of cross-border banks, and the need for better and more harmonized supervision.
One may argue, and I do, that banking union is more of a single market issue than a single currency issue, while acknowledging that it is both. We need more competition in financial services and therefore more activities across borders. The reversal of cross-border financing activities since the financial crisis is not helpful.
Supervision by the ECB can better ensure a level playing field in terms of supervisory practices. It can enhance the quality and credibility of supervision by adding external eyes to work done by national supervisors - excellent as it may often be. The practice set by the ECB is in any case likely to establish an EU wide benchmark for supervisory standards.
A level playing field needs a strong single resolution mechanism which can effectively apply and implement bail-in. Note that the requirement of an 8 per cent bail-in before using resolution funds and public credit lines is very strong. 8 per cent of all assets often compares to 25-30 per cent of risk-weighted assets. The resolution mechanism will also work as an insurance scheme in case of even graver systemic events. This is helpful, provided it is based on sound principles.
The ECB is currently performing an Asset Quality Review of the largest banks of the euro area, and is preparing an ensuing stress test. The Danish supervisor will perform a similar review of the largest Danish credit institutions.
Obviously, any legacy problems unearthed by the AQR and stress tests have to be dealt with before any joint insurance element of banking union will come into play. The ECB has strong incentives to safeguard the success of the AQR and be a strong and hands-on supervisor. Legacy issues are not a particular dividing line between countries in or out of the euro zone, but something being a matter for all possible participants in the banking union.
There are still issues and details to be decided and dealt with, but we should work towards the success of the banking union, and as mentioned, we believe that Denmark should continue to engage strongly with a view to participating when appropriate.
Thank you for your attention. I am happy to take any questions.