Central Bank publishes Financial Stability Note, “Risk Weights on Irish Mortgages”

22 February 2022 Press Release

Central Bank of Ireland

  • The Central Bank is currently conducting a review of its macroprudential framework for bank capital, including the interactions of macroprudential buffers with other elements of the capital regime, such as risk weights. Risk weights are calculated under global rules established to ensure banks with riskier assets have more capital to be able to absorb higher potential losses on these assets.
  • In this context, research published today analysed risk weights on Irish residential mortgage exposures, which are high in a European context.
  • Higher risk weights in Ireland reflect a number of factors, including:
    • the higher default risk of the stock of Irish mortgage exposures, driven by loans issued before the financial crisis that remain on banks’ balance sheets;
    • the length of time it has taken for banks to resolve distressed loans arising from the financial crisis, including through repossession, and;
    • the implications of the internationally agreed regulatory framework requirement for loss estimates to reflect an economic downturn, which is particularly severe given the loss experience from the financial crisis in Ireland.

The Central Bank of Ireland has today (22 February 2022) published a Financial Stability Note, “Risk Weights on Irish Mortgages”. Authored by Paul Lyons and Jonathan Rice, the Note examines the main contributing factors to risk weight density (RWD) on mortgage loans in Ireland. In this context, RWD is defined as risk-weighted assets expressed as a percentage of mortgage loan exposure.

Risk-weighted assets link the minimum amount of capital that a bank must have to the risk profile of the bank’s asset book. Under rules established by the Basel Committee on Banking Supervision (BCBS), translated into European regulation via the Capital Requirements Regulation (CRR), banks with riskier assets are required to have more capital to be able to absorb higher potential losses on these assets.

Understanding the drivers of risk weights is an important element of the Central Bank’s broader work on the evolution of the macroprudential capital framework. Over the past decade, taking on board lessons from the financial crisis, there have been widespread reforms to the overall bank capital framework, including the introduction of an extensive macroprudential toolkit around bank capital. The Central Bank is currently conducting a review of its strategy around macroprudential capital buffers. This will also consider the interactions between macroprudential capital buffers and other elements of the capital framework, such as minimum requirements and risk weighted assets, the focus of this Note in the context of mortgages.

The Note published today specifically examines residential mortgages covered by Internal Ratings-based (IRB) models. Average Irish retail mortgage risk weights are around twice the reported European averages. The authors illustrate how care is needed when performing relative comparisons to other European countries as the published risk weight contributions in some countries do not consistently capture ‘add-ons’ imposed by regulatory authorities to address ‘low’ risk weights in these countries.

To shed light on the drivers of higher RWDs on Irish mortgages, the Note examines the two key parameters in risk weight modelling: probability of default (PD) and loss-given default (LGD). Both are found to be higher than European averages. Specifically:

  • On Probability of Default, the authors find that Irish mortgage default rates have been consistently above EU averages, both in recent years and over a longer period of time. The relatively higher modelled PD reflects the underlying higher riskiness of the mortgage stock in Ireland, relative to other countries. Loans issued prior to the Irish financial crisis (2009-2013), originated under weaker lending standards, continue to drive the higher default risk. Around 85 per cent of defaults in recent years stem from loans that were originated prior to the financial crisis, which continue to contribute significantly to the riskiness of Irish banks’ loan books. The authors also find that average modelled PD feeding into risk weights on performing mortgages in Ireland has been gradually declining since 2015, reflecting the improved economic environment and improved lending standards. All else equal, modelled PD is expected to continue to decline over time, as loans originating before the financial crisis are gradually replaced by newer loans issued under stronger lending standards.
  • On Loss Given Default, the authors find that the modelled LGD for Irish mortgages is also higher relative to other countries. Under internationally-agreed regulatory requirements, LGD estimates should be based on banks’ loss and recovery experience during an economic downturn. For Irish banks, this has been particularly severe. A key contributor to higher LGDs in Ireland is the long workout periods of mortgage loans that defaulted during the financial crisis. This, in turn, reflects two factors. First, Ireland’s repossession system, which results in a longer timeframe to realise collateral for defaulted assets. Second, the unusually large shock observed in the Irish mortgage and housing markets during the financial crisis and, relatedly, the prolonged period to work through the volume of defaulted loans during the financial crisis.

The Note’s authors conclude that the differential in risk weighting for Irish mortgage loans as compared with European loans will narrow over time. That is partly because loans originating before the financial crisis will gradually be replaced by newer loans issued under stronger lending standards. But it is also due to the introduction of Basel III reforms, which are expected to lead to higher risk weights for those banks in other countries where modelled risk weights are considered too low. In the medium-term, Irish mortgage risk weights would still be expected to remain higher than average in Europe, even if the differential narrows, given the factors outlined in this Note.

Notes to Editor

Risk weight density is defined in this Note as a firm’s risk-weighted assets expressed as a percentage of the firm’s total loan exposures. Banks can choose between two broad methodologies for calculating risk weights -

  • Standardised approach: assigns standardised risk weights based on the exposure type (for example, mortgages, corporate lending, etc.). This approach is therefore less sensitive to the risk of each specific loan.
  • Internal Ratings-Based approach: banks use their own internal models to estimate the inputs used to calculate risk weights. The key parameters used in this calculation are probability of default (PD), loss-given default (LGD), and exposure at default (EAD) and maturity (M). This approach can be further divided as follows:
    • The Foundation Internal Ratings-based approach (F-IRB) requires banks to determine PD only. The other parameters, for example, LGD, are prescribed by supervisors.
    • The Advanced Internal Ratings-based approach (A-IRB) requires banks to estimate all key parameters – PD, LGD, and EAD. To use this approach, banks’ models are subject to explicit approval from supervisors.