Variable mortgage risk weighting – Procyclical or anticyclical timing?

Increasing mortgage loan risk weights in a depressed property market is likely to be procyclical, as would reducing risk weights in booming property market. Strangely, this procyclicality appears to be acceptable under the contemplated EBA standards on adjusting risk weights due to financial stability considerations that are currently out for consultation. The draft binding rules do not specify when they should best be adjusted up, and when down, nor how to take into account such potential procyclical effects. Nothing in the proposed binding rules clarifies at which part of the cycle this lever should be used, which is a bit odd for nominally technical rules that have as their key ingredient that a specific lever can be used for financial stability considerations.

The EBA proposals do give a clue as to what information is relevant, but mainly leave the type of response to the supervisor itself. A supervisor eager to apply the law in a conservative manner is left scratching his head as to the optimum course and timing. A supervisor eager or under political or monetary policy pressure to boost a growing economy, or to stop a sliding property market, is free to do whatever it wants even if the longer term effects might be a less safe banking system. For example, the economies of many of the member states currently need a stimulus. Increasing house prices and office prices based on cheaper lending – if banks do not need to hold so much capital – could help provide such a stimulus. Even though bad lending practices and too low risk premiums and risk buffers for mortgage loans in the USA subprime sector actually kicked off the latest worldwide crisis, the solution to help growth in the short term could be to keep risk weights low, and to keep all options open for national legislators and supervisors. As a result of such pressures it is difficult to blame EBA and its voting members for building in this leeway. However, it does mean that the new binding rules are not very useful if a supervisor or financial stability regulator would like to be able to take measures to ensure the stability of the banking sector and/or the property market. The standards instead excel in less than clear guidance such as ‘Take into account housing market developments’, which kicks in a wide-open door, and says nothing on whether rising values or buyers interest should lead to an increase in risk weighting (and thus higher capital requirements), or to a decrease in risk weighting (and thus lower capital requirements).

This leaves aside that a discussion could be had on whether a higher risk weight would be best from a technical point of view in the upslope of a boom (to stop irrational exuberance, and build up capital buffers for the eventual decline in property values a few years hence and thus in an anticyclical manner), or on the downslope towards a trough (to increase the potential for bank capital being sufficient to deal with future losses in a value-declining property market, thus limiting the scope for banks to lend to potential new purchasers and forcing them to double down capital for existing and new downward developing mortgage loans, even though for the wider economy this would be procyclical). In this light, an analysis performed by supervisors on the basis of the lengthy data sets available over the boom period and the bust in immovable property markets in almost every member state could have been used to base these standards on an analysis of the costs and benefits of heightening and reducing risk weights in each national or regional property market in the period from e.g. 2000 until now. Indicating when Dutch, Spanish, Irish or any other national supervisor in hindsight would have wished that they used the existing risk weight-adjustment instrument either in a pro- or anticyclical manner during that period might lead to useful indicators as to when it should be used in the future with the best impact on wider financial stability as well as on the resilience provided by larger bank financial buffers.

A compromise solution could be to try to aim for the upper slopes of the boom for an increase, and reduce it when property prices have gone below reasonable long term values. At the bottom of the trough this would stimulate the housing market, especially if the expected losses on the housing portfolio have already been written down in full under a possibly wider definition of default and/or lower valuation of the collateral. Higher risk weights on the remaining fully covered mortgage loans would then no longer be necessary, if – and only if – the risk weight setter is able to correctly call when a boom is under way, or when a property market recession is entering irrationally depressed territory.

It would thus be helpful if the standards clarify whether their primary target is to stabilise the immovable property market in a certain market segment, or to stabilise the banks that lend in that area even if that means restricting loans to a already plummeting property market, or both. That would also help indicate whether there is a need to coordinate across financial sectors and across banks on the standardised and IRB approach (to ensure that banks, insurers, pension funds and other non-bank mortgage loan providers increase or decrease their exposure to the market segment involved in the same manner) which I would favour, or not (to ensure that the banks are safe by being able – to put it bluntly – to offload the risky and more costly exposure to the overheated property segment, even if that is to unsuspecting insurers or securitisation-investors such as pension funds).

This overall lack of clear indicators and purposes means that I am a bit reluctant to criticise the only clear benchmark that EBA does provide, which has been referenced in the draft standards and made more concrete in the impact assessment. According to it, loss expectations should be a key factor to determine how high the risk weights should be. It is a welcome clarification of intent, and something supervisors might be benchmarked to. However, though I applaud its inclusion, this specific benchmark does clarify two things that in view of pro and anti-cyclical thinking are a bit unwelcome. The first is that higher loss expectations are expected to be the trigger for an increase in risk weighting. As soon as market based loss expectations are made the determining factor, any irrationality in the market suddenly becomes less easy to deal with. This irrationality is part of the accepted market wisdom at that time, so if for ten years prices have gone up, no one ‘expects’ losses any more. Only once the bust period actually arrives, loss expectations suddenly swing up (sometimes to irrational heights in a panic). Increasing risk weights at that point in time will only strengthen the slide into the abyss. If risk weights instead are already up when loss expectations are still close to nil, then the lever could helpfully be used to lighten the load on the way down, helping to dampen the cycle. That does, however, require supervisors actually to take a stand against ‘the sky is the limit’ politicians and realtors, which as indicated above may not be their favoured role.

Second, the table appears to indicate that the lowest risk weights are appropriate in ‘normal’ times. If so, the lever of risk weights is unavailable during the entire trough of the cycle, meaning it has no dampening effect to get the market (and the banks’ capital requirements) into a mood that indicates light at the end of the tunnel. From a macroprudential point of view, that seems unhelpful. The lowest risk weights should be only in force at the ‘apex’ of the bust, so that the lever can be used both in the downswing and the upswing. No doubt this is more the role of the ESRB to point out, but strangely their role as providers of warnings and advisors on the cyclicality of draft-rules is not visibly reflected in the EBA draft standards.

In conclusion, it may be good to re-assess and clarify some of the key concepts, main goals and direction of adjustments in the draft binding rules before they enter into force. Building upon the experience in the past crisis with a ‘in hindsight’ analysis as to when and how this tool would have been most effective and efficient would be helpful. Both changes would help shelter banking supervisors from being put under pressure to sacrifice long term bank stability against short term political pressure for economic growth.

Also see:

The separate comment on adjusting the mortgage risk weights
EU Banking Supervision, chapter 6.2, 6.5, 8, 18.3, 21.2-21.4, and 22.5
124-126 CRR
128.2 sub d CRR
164-166 CRR
EBA consultation paper EBA/CP/2015/12 of 6 July 2015 on determining higher risk-weights

This drug needs to be utilized only on line levitra once every day. Vacuum Devices: Such devices levitra side effects are help to improve semen load. By all means, he online pharmacy for levitra prides himself in keeping her happy. Although women use about 15 buying viagra without prescription percent of the testosterone be it administered or natural, remains in a state that is bound which makes it very useless for muscle building.