ERU, Alm.del - 2015-16 - Bilag 318: Regeringens høringssvar til Basel-komiteen

Erhvervs-, Vækst- og Eksportudvalget 2015-16
ERU Alm.del Bilag 318
Offentligt

ANNEX

10. June 2016

Danish response to the consultative document on reducing vari-

ation in credit risk-weighted assets – constraints on the use of

internal model approaches

General Danish remarks

Denmark would generally like to encourage the Committee to consider

the following:

•

To reassess the need for output floors in light of other similar

safeguards in the regulatory framework (e.g. the leverage ratio).

To carefully consider the calibration of parameter floors and to

abandon the use of floors on exposure level.

To explicitly allow the advanced IRB approach for Income Pro-

ducing Real Estate.

It is our view that these three changes would address the main Danish

concerns while preserving the essence in the Committee’s proposal.

Hence, addressing these three challenges would be the Danish preferred

way forward.

Alternatively, we encourage the Committee to consider a more nuanced

approach. Following this, we would recommend a solution where the

Committee’s current proposal would more or less be implemented as is

but where this would happen in combination with special treatment, e.g.

less restrictive calibration, for demonstrably low risk markets and busi-

ness models.

Particularly, we find that the differences between national housing mar-

kets are too substantial to warrant a one-size-fits-all approach. In our view

there should be room for a more differentiated approach in this area, par-

ticularly considering the high importance of local housing markets to the

different jurisdictions’ economies.

In this respect, we find it prudent and appropriate to base such a more

nuanced approach on quantitative criteria regarding the availability of

long time series of low losses as well as more qualitative criteria regard-

ing the functioning of the market. For example, we find that the exist-

ERU, Alm.del - 2015-16 - Bilag 318: Regeringens høringssvar til Basel-komiteen

2/6

ence/lack of full recourse is an important distinguishing feature between

markets with different degrees of risk. Another important feature, in our

view, is the degree to which an efficient liquidation process can be ham-

pered by extensive consumer protection.

Specific Danish remarks

As already mentioned Denmark finds that the need for output floors

should be reassessed in light of other safeguards in the regulatory frame-

work.

However, should the Committee decide to introduce an output floor based

on the revised standardised approach, we would find it prudent to consid-

er a more nuanced approach. Specifically, we wish to encourage the

Committee to consider lowering the risk weights for exposures secured by

real estate in demonstrably low risk markets under the revised standard-

ised approach.

Furthermore, the Committee proposes to set exposure-level floors on

IRB-model parameters (PD, LGD, and EAD). We are not convinced of

the merits of using floors on exposure level and we question this approach

as it is solely targeting low risk exposures. Furthermore, it also distorts

credit risk models’ ability to correctly rank customers and exposures and

consequently interferes directly in credit institutions’ daily risk manage-

ment. Portfolio level floors, as currently known for retail residential

mortgages, would in our view be far less interfering although the calibra-

tion would still need to be carefully considered.

As mentioned in our main letter, it is our understanding that the proposal

regarding Specialised Lending (SL) is based on concerns about the

modellability of such exposures, for example due to lack of data or corre-

lation between creditworthiness of the borrower and the value of the asset

being financed.

There are five sub-categories in the SL segment, one of which is Income-

Producing Real Estate (IPRE). While we acknowledge that concerns

about modellability may be valid for some sub-categories of SL, we seri-

ously question this concern when it comes to substantial parts of IPRE

where Danish credit institutions are able to develop well-functioning

models.

Data is widely available in the Danish mortgage credit institutions for

some segments such as property rental exposures and exposures are rela-

tively homogenous (possibly more than other corporate exposures).

Likewise, it is our experience that concerns about correlation between

borrower and the asset can be handled directly in the models.

ERU, Alm.del - 2015-16 - Bilag 318: Regeringens høringssvar til Basel-komiteen

3/6

Furthermore, in a Danish context large parts of the IPRE segment are

generally low risk. Exposures are typically highly over-collateralised and

in some cases there is also partial governmental support. Following this,

we are overall concerned that the proposal will lead to an unwarranted

increase in capital requirements, a loss of risk sensitivity and as a conse-

quence incentivise credit institutions to shift their portfolios towards

higher risk.

We therefore encourage the Committee to explicitly allow the AIRB ap-

proach for IPRE if the institutions can document the availability of suffi-

cient historical data and fulfil all requirements for IRB modelling.

Alternatively, given the inherent low risk of many of the exposures in

question we call for a less restrictive treatment under the standardised

approach or the supervisory slotting approach. Again, such a change

could be targeted at markets with demonstrably low risk.

The Committee’s proposal contains a requirement that assignment to rat-

ing categories should remain stable throughout business cycles such that

migrations are due to company-specific or industry-specific changes and

not business cycles fluctuation.

We assume that the objective of this proposal is to achieve relatively sta-

ble capital requirements. However, we doubt that credit institutions will

be able to comply with the requirement in practice as all the rating sys-

tems we know exhibit some migration over the cycle (e.g. when financial

ratios change). On this background, we therefore encourage the Commit-

tee to consider a less prescriptive wording of the requirement.

The Committee proposes that exposures to corporates belonging to con-

solidated groups with total assets equal to or exceeding EUR 50bn should

be subject to the standardised approach. We acknowledge the challenges

to obtain reliable estimates of PD and LGD for corporates belonging to

consolidated groups with total assets equal to or exceeding EUR 50bn due

to the low-default nature of these corporations. However, using the stand-

ardised approach would mean increases in capital requirements which we

do not find justified due to the same low-default nature of these expo-

sures. We therefore encourage the Committee to reassess the need for

using the standardised approach, and as an alternative consider the F-IRB

approach for these exposures. In our view, using the F-IRB approach

would be a way to preserve a risk sensitive approach but at the same time

take into account that it can be difficult to obtain reliable estimates.

The Danish mortgage credit sector

ERU, Alm.del - 2015-16 - Bilag 318: Regeringens høringssvar til Basel-komiteen

4/6

As the proposal will significantly affect low risk market segments, includ-

ing the Danish mortgage credit sector, we find it useful to briefly outline

some of the special characteristics regarding this market.

The Danish mortgage credit sector is inherently low risk. This can be

demonstrated by the empirical development as well as by the specific

features of this sector.

Impairment charges for Danish mortgage credit institutions have averaged

around 0.2 percent over the last 30 years. During the most recent financial

crisis, which did affect the Danish economy severely, annual impairment

charges did not exceed 0.2 percent. Please see figures A1 and A2 below

for the development over an extensive time period.

Particularly worth highlighting, the mortgage credit institutions were able

to increase their loan volume during the most recent financial crisis. Thus,

the mortgage banks acted as an important stabilizing factor at a critical

point where credit granting was otherwise contracting.

The historic performance of the mortgage credit institutions is not a result

of chance, but due to the regulatory framework within which the institu-

tions operate. Naturally, changes have been introduced to the framework

over time – including introduction of new product types and removal of

obligors’ joint and several liability – but all in all Danish mortgage credit

institutions have operated in a markedly low risk environment throughout.

Thus, while mortgage credit institutions must comply with the same regu-

lation as all other credit institutions, they are subject to additional national

requirements which set narrow limits for their business activities. In order

to limit credit risk, mortgage credit institutions are only allowed to grant

loans secured by real estate and only within specified loan-to-value limits.

For residential real estate, loans can only be granted if they are within 80

% of the property value and for most other segments the equivalent limit

is 60 %.

It is also important to highlight that Danish mortgage credit institutions

have full recourse to all assets of a defaulting obligor and Danish obligors

are thus personally liable for their debt. This provides mortgage credit

institutions protection from losses beyond the risk mitigation originating

from the collateral provided. Perhaps even more important, it additionally

serves as a strong incentive for obligors to service their debt as nothing is

gained from defaulting, even if the value of their property should fall be-

low the value of the outstanding debt.

ERU, Alm.del - 2015-16 - Bilag 318: Regeringens høringssvar til Basel-komiteen

5/6

An additional feature of the regulatory framework surrounding the mort-

gage credit institutions is their very efficient access to liquidate the collat-

eral in the event of default. Again, this provides a direct protection against

losses, while incentivising the obligors to service their debt as it is not

possible to prolong the liquidation process.

While not directly related to the proposal at hand, we would also like to

further stress that the regulation of Danish mortgage credit institutions

further ensures the low risk nature of the sector, e.g. effectively prohibit-

ing mortgage credit institutions from taking significant market risk just as

liquidity risks are very limited.

In addition, the legislation has in recent years been reinforced, limiting

refinancing risk for the mortgage credit institutions’ adjustable-rate mort-

gages. The legislation prescribes that bonds issued with a shorter maturity

than the loans which they fund are automatically rolled over in the low-

probability event where a refinancing auction of covered bonds related to

the mortgage-credit institutions adjustable rate loans is unsuccessful, or if

there’s a steep, sudden rise in the bond interest rate.

As mentioned above, impairment charges have historically been very low

for Danish mortgage credit institutions. Figure A1 below shows the low

impairment charges of Danish mortgage credit institutions compared to

those of commercial and savings banks. Over the last 30 years the average

loan impairment charge for mortgage credit institutions has been 0.2 per

cent while the corresponding average for commercial and savings banks

has been 1.0 per cent. This period even includes two periods of financial

distress as illustrated in figure A2 where real GDP growth and relative

changes in house prices are added to figure A1.

ERU, Alm.del - 2015-16 - Bilag 318: Regeringens høringssvar til Basel-komiteen

6/6

Figure A1: Loan impairment charges for commercial and savings banks

and credit institutions

Figure A2: Loan impairment charges for commercial and savings banks

and mortgage banks, relative changes in house prices, and real GDP

growth.