Response to consultation on RTS on conditions for capital requirements for mortgage exposures
Go back
However, it remains unclear how the final expected loss rate would be obtained. Making adjustments on the basis of the forward looking immovable property market developments is problematic, and there is no reliable methodology for forecasting asset prices, while adjustments made based on such forecasts would be subject to high error margins.
We would therefore recommend adjustments are based on historical experience are preferred.
The ranges of loss expectation relative to risk weights are broad and, at the lower end, a small increase in loss expectations may lead to a large increase in risk weight.
We note that in the UK the largest single year loss is 0.26% in 1993.
Without further consultation it’s difficult to comment fully on the proposed benchmarks. At the upper end of the 35% RW benchmark, 1.5%, it’s not clear to us why a standard deviation would apply. Whereas at the lower end why recovered assets would not be recognised as in the case of default BTL regulations (CRR Articles 125/126 paragraph 2(b)).
We suggest regulators also take into account discretionary controls should not lead to duplicative or multiplicative capital requirements. As explained above, increasing RW% and LGD% Floors will overlap with Pillar 2 Capital measures, the Capital Planning Buffer is already in place to cover stressed losses, and the PRA Counter-Cyclicality buffer is in place to promote financial stability.
We would also expect regulators to provide comprehensive explanations of any decisions they make.
Thought should be given to the principle behind why a minimum LGD is set i.e. because some exposures with very high levels of security and where in the event of default it is very unlikely that any loss will be experienced might carry little or no capital which regulators are reluctant to sanction. We are wary that any change in the minimum LGD demands a good understanding of the status of the portfolios caught by the minimum LGD and the scope for limited losses even in a severe downturn, which in turn requires specific national expertise.
We notice that no mechanical link is proposed between the expected and the minimum LGD and agree with this approach as changes to the minimum LGD should only be considered in exceptional circumstances, and judgement is required.
Making adjustments on the basis of the forward looking immovable property market developments is problematic. There is no reliable methodology for forecasting asset prices, and adjustments made based on such forecasts would be subject to high error margins. We recommend adjustments based on historical experience are preferred.
Increasing the LGD portfolio floor decreases IRB model sensitivity and disproportionately affects low risk organisations, thus favouring higher risk strategies. It not clear under what economic conditions this tool would best bring about financial stability. Model changes focussing on collateral values would maintain sensitivity, for example ‘Peak-to-trough’ house price assumptions or ‘ Collateral Haircuts’.
There needs to be a better explanation of how average LGD forecasts can be converted into practical LGD floors.
Adjustment should be made to the Mortgages market as a whole. If segmentation is preferred this should be predefined with sufficient lead time to enable institutions to adapt their model infrastructure. The current infrastructure makes it difficult to act on individual portfolios, e.g. where short market history, as LGD floors may overlap.
As highlighted above, any change in the minimum LGD should be at the discretion of the Competent Authorities, but decisions should be transparent and supported by a comprehensive explanation.
Discretionary controls should not lead to duplicative or multiplicative capital requirements. Increasing RW% and LGD% Floors will overlap with Pillar 2 Capital measures, the Capital Planning Buffer is already in place to cover stressed losses, and the PRA Counter-Cyclicality buffer is in place to promote financial stability.
The question should be asked as to whether the LGD Model infrastructure enables the desired action to be implemented across all institutions on an equal basis.
Regulators should demonstrate why this is the best discretionary control tool to use at this moment in time.
Question 2: Do you agree with the conditions for specification of the loss experience and the loss expectations? Do you agree with the adjustments allowed to be made to the loss experience on the basis of the forward-looking immovable property market developments?
In principle we do agree with the CP, but it should be noted that developing a view as to future expected loss is inevitably judgemental. The suggested areas of investigation are reasonable, and we agree it is right to emphasise the impact on risk weights of a switch in exposure to unsecured retail as property prices decline.However, it remains unclear how the final expected loss rate would be obtained. Making adjustments on the basis of the forward looking immovable property market developments is problematic, and there is no reliable methodology for forecasting asset prices, while adjustments made based on such forecasts would be subject to high error margins.
We would therefore recommend adjustments are based on historical experience are preferred.
Question 3: Do you agree with the indicative benchmarks for the assessment of the appropriateness of the risk weights and to guide the setting of higher risk weights across immovable property markets in different member states as specified in Article 4(3) and 4(4)? What levels of these indicative benchmarks would be most appropriate and why?
We are unconvinced by the use of possible triggers and/or benchmarks. The CP suggests two possible approaches with different outcomes underlining the judgemental aspect of the approach. We believe the burden should be on the Competent Authorities to justify any change in minimum risk weight with reference to the previously identified forward indicators.The ranges of loss expectation relative to risk weights are broad and, at the lower end, a small increase in loss expectations may lead to a large increase in risk weight.
We note that in the UK the largest single year loss is 0.26% in 1993.
Without further consultation it’s difficult to comment fully on the proposed benchmarks. At the upper end of the 35% RW benchmark, 1.5%, it’s not clear to us why a standard deviation would apply. Whereas at the lower end why recovered assets would not be recognised as in the case of default BTL regulations (CRR Articles 125/126 paragraph 2(b)).
Question 4: Do you agree with the specification of the term of “financial stability considerations”?
We agree with specification, and the definition of financial stability is sufficiently wide ranging to give Competent Authorities the necessary discretion to make prudent decisions. Despite this, we would however refer to the key message above that LGD is not an appropriate tool for controlling overall financial stability.Question 5: Do you agree with the other conditions for the setting of higher risk weights? (Please provide your feedback related to the indicative benchmarks (in Article 3(3) and 3(4)) in your response to Question 3 above.)
We would again direct the EBA to the key message, and question why changing risk weights is deemed the most appropriate approach.We suggest regulators also take into account discretionary controls should not lead to duplicative or multiplicative capital requirements. As explained above, increasing RW% and LGD% Floors will overlap with Pillar 2 Capital measures, the Capital Planning Buffer is already in place to cover stressed losses, and the PRA Counter-Cyclicality buffer is in place to promote financial stability.
We would also expect regulators to provide comprehensive explanations of any decisions they make.
Question 6: Do you agree with the conditions for specification of the exposure weighted average LGD and the LGD expectation? Do you agree with the adjustments allowed to be made to the average exposure weighted LGD on the basis of the forward-looking immovable property market developments? Do you agree that it is not appropriate to set indicative benchmarks for the setting of higher minimum LGD values because of the specificities of national immovable property markets and because of the relationship of the LGD parameter with the other internal model parameters?
We ask for the following to be considered: Thought should be given to the principle behind why a minimum LGD is set i.e. because some exposures with very high levels of security and where in the event of default it is very unlikely that any loss will be experienced might carry little or no capital which regulators are reluctant to sanction. We are wary that any change in the minimum LGD demands a good understanding of the status of the portfolios caught by the minimum LGD and the scope for limited losses even in a severe downturn, which in turn requires specific national expertise.
We notice that no mechanical link is proposed between the expected and the minimum LGD and agree with this approach as changes to the minimum LGD should only be considered in exceptional circumstances, and judgement is required.
Making adjustments on the basis of the forward looking immovable property market developments is problematic. There is no reliable methodology for forecasting asset prices, and adjustments made based on such forecasts would be subject to high error margins. We recommend adjustments based on historical experience are preferred.
Increasing the LGD portfolio floor decreases IRB model sensitivity and disproportionately affects low risk organisations, thus favouring higher risk strategies. It not clear under what economic conditions this tool would best bring about financial stability. Model changes focussing on collateral values would maintain sensitivity, for example ‘Peak-to-trough’ house price assumptions or ‘ Collateral Haircuts’.
There needs to be a better explanation of how average LGD forecasts can be converted into practical LGD floors.
Adjustment should be made to the Mortgages market as a whole. If segmentation is preferred this should be predefined with sufficient lead time to enable institutions to adapt their model infrastructure. The current infrastructure makes it difficult to act on individual portfolios, e.g. where short market history, as LGD floors may overlap.
As highlighted above, any change in the minimum LGD should be at the discretion of the Competent Authorities, but decisions should be transparent and supported by a comprehensive explanation.
Question 7: Do you agree with the other conditions for the setting of higher minimum LGD values?
The subjective nature of the requirements, combined with the difficulty of answering all questions with a high level of confidence in a timely fashion is causing concern. The following additional factors should be taken into account: Discretionary controls should not lead to duplicative or multiplicative capital requirements. Increasing RW% and LGD% Floors will overlap with Pillar 2 Capital measures, the Capital Planning Buffer is already in place to cover stressed losses, and the PRA Counter-Cyclicality buffer is in place to promote financial stability.
The question should be asked as to whether the LGD Model infrastructure enables the desired action to be implemented across all institutions on an equal basis.
Regulators should demonstrate why this is the best discretionary control tool to use at this moment in time.