top of page
  • Chris Thorne

Square Peg Found in Round European Hole.

Earlier this month the European Commission was advised by one of its supervisory authorities to change a regulation that would otherwise have the effect of imposing a uniform method of calculating "mortgage lending value" across the EU. Why has there been this apparent pushback against the tide of pan European integration of financial services regulation?

The question of the percentage of value that can be prudently lent as a mortgage loan has been a perennial problem for both lenders and those who regulate them. Many see a problem in the fact that the value of the property held as security can fluctuate significantly over the term of the loan. A loan that is well secured one year may not be the next if market conditions change for the worse. Over one hundred years ago, some German issuers of bonds secured on the income from mortgages (covered bonds) invented the concept that has become known as “mortgage lending value” (MLV), which is a method of calculating a figure that its proponents claim provides a value that will be valid throughout the life of a loan.

The idea of a value measurement that can always be relied upon regardless of what the market may be doing is clearly seductive to those concerned about lending risk. Because of this, in recent years MLV has migrated from its origins in the German covered bond, or Pfandbrief, market. For example it now features in regulations and directives issued by the European Union on the measurement of the capital that must be held by financial institutions. In the calculation of the capital a bank is required to hold to maintain its solvency, mortgage loans secured against the MLV of property carry a more favourable risk weighting than those secured against the more common measure of market value.

The European Union had included a provision in the Capital Requirements Regulation (CRR) that required the European Banking Authority (EBA) to produce Regulatory Technical Standards (RTS) for the calculation of MLV, with the objective of ensuring this would be done consistently across the EU. However, in the course of preparing the RTS the EBA has recognised the extent to which MLV is determined by national regulation in the covered bond markets in Germany and some other European states. It has advised the European Commission that to oblige those covered bond markets currently using MLV to change to a pan European method of calculation would create a substantial risk of introducing a disproportionate cost on both the institutions involved and the relevant supervisory authorities. The EBA has accordingly issued an opinion in which it advises the European Commission that the CRR should be amended to limit the scope of the proposed RTS so that they do not apply to covered bonds held by financial institutions.*

The method for calculating MLV to be introduced by the EBA would still apply where it is a permitted measurement option for secured loans under the Standardised Approach, Credit Risk Mitigation Framework and Large Exposure provisions of the CRR. However, there is a certain irony that, having studied the technical implications of drafting a pan EU method for calculating MLV, the EBA is recommending that the market from which the concept emerged should be taken out of the scope of its project.

I would suggest that the recommendation has come from the realisation that, unlike market value, MLV is not based on observable behaviour and prices in the relevant market but is an artificial construct. The methods of calculating MLV currently in place in national regulation use certain market inputs but forbid the use of others in favour of inputs that are determined by regulation. For example, under the German BelWertV regulation, while MLV is supposed to have regard to normal regional market conditions, there are stipulated minimum yields that can be applied to investment property and a minimum percentage of income that must be deducted to reflect operating expenses. Furthermore, temporary, economically induced fluctuations in value have to be disregarded along with speculative elements. But notwithstanding the direction to disregard economically induced fluctuations, the MLV cannot exceed the market value at the date of its assessment, so if this is the case a further adjustment has to be made.

MLV undeniably is a regulatory construct, unlike market value, which is an observable and measureable economic concept common to all markets. As such MLV is what the relevant regulator wants it to be. The EBA seems to have recognised that since there is no "right" scientific answer involved there is little to be gained in creating a set of RTSs that will conflict with the well-established regulations in the covered bond markets, such as those in Germany. The alternatives of either extending the limits and conditions deemed appropriate for the German Pfandbrief market across the whole of the EU, or of requiring the Germans to replace their long established method of subjecting a market price to various haircuts with something that was workable across the whole of Europe would both be unpalatable. Either would cause problems out of all proportion to their relevance to the central objective of the CRR, which is to ensure that banks hold adequate capital reserves.

The reaction of the European Commission to this recommendation is awaited, but hopefully it will recognise the pragmatism of the EBA’s position and put in hand the suggested amendment. Of course the elephant in the room is whether MLV does actually offer what it claims, or whether there is other information that would be more useful to a lender in assessing the contribution that the security offered makes to the overall risk profile of a loan. However, that is a debate for another day.

* EBA/Op/2015/17 5 October 2015


Featured Posts
Recent Posts
bottom of page