We suggest that the scope of the report should be increased to consider the risk exposures that Defined Benefit Pension schemes represent to Institutions and ensuring that CET1 capital impact reflects the risk exposure.
The funding of Defined Benefit Pension schemes is a long term risk to Institutions, and employees cannot accelerate receipt of benefits. However under CRD IV significant short term volatility to CET1 capital is likely to be experienced by an Institution – this may require short-term capital raising actions in response to adverse temporary movements in Defined Benefit Pension deficits, which are not reflective of the actual levels of risk to which the Institution is exposed.
We believe that that the CRD IV treatment of Defined Benefit Pension schemes can lead to undue volatility in the level of CET1 capital to Institutions, causing adverse impacts to Institution’s capital planning and introducing another source of pro-cyclicality to Institutions’ capital positions. We provide more information on these issues in our answers to Question 7 and Question 8 below.
We understand the internationally agreed Basel 3 rules have been reflected in CRD IV, but we would recommend that the EBA and EU Commission consider mechanisms to reduce the short term volatility to Institutions CET1 capital that the CRD IV rules in their current form introduce. These mechanisms could include:
• an approach where CET1 impacts are based on the level of Defined Benefit Pension scheme contributions to be paid by an institution (an approach used in the UK prior to CRD IV)
• through the ‘smoothing’ of movements in the high quality corporate bond discount rates used under IAS 19, with a threshold level for impact on CET1 capital. It could be appropriate for a regulatory body to determine (adjusted) interest and inflation rates to be applied for the valuation of Defined Benefit Pension schemes for prudential purposes. Accumulated gains and losses would only impact CET1 levels when they exceed the greater of 10% of Commitments or 10% of Assets.
We also believe that it is important that consideration is given to the equality of treatment of Institutions, other Financial Sector companies and non-Financial companies that are exposed to Defined Benefit Pension scheme risk. Non-financial companies are also exposed to Defined Benefit Pension scheme risk, but are not required to make explicit capital provision through capital deductions or non-recognition of Defined Benefit Pension scheme assets. It is important that the rules for Defined Benefit Scheme risk covering Institutions do not excessively penalise Institutions compared to other Financial and non-Financial companies.
We believe that the proposed methodology is a reasonable basis to address the objectives. We however believe there are significant shortcomings with the quantitative second part of the analysis and that significant additional items should be considered in the additional qualitative assessment – we comment on this in our answers to Question 7 and Question 8 below.
We agree with the conclusion that for Defined Benefit Pension Schemes which remain in IAS 19 surplus (and are an asset) the CRD IV rules would mean that there is no direct impact to an Institutions CET1 capital position (however there may be second-order taxation-related effects).
We do agree with the EBA that actuarial gains and losses are expected to be the most significant driver of the CET1 capital position impacts.
We do agree with the analysis performed. The immediate recognition of actuarial gains and losses is the most relevant change for the the scope of this report.
We generally agree with this analysis.
We believe that there are significant shortcomings with the methodology and scope adopted for the quantitative analysis performed:
• Data Used and Time Period: The analysis is restricted to 3 year-end positions (2010, 2011 and 2012) - a significantly longer dataset is required to provide robust analysis on the expected levels of volatility to Institutions CET1 capital levels under the CRD IV rules. Moreover, some Member States and International Jurisdictions require Defined Benefit Pension Scheme surplus / deficit assessments to be performed more frequently than once a year. It is therefore also important that analysis should be performed to consider the CET1 volatility over a shorter time horizon than one year which some Institutions could be subject to.
• Group Level Analysis: The analysis focuses on the impact to EU banking groups. Under CRD IV impact also can occur at a solo level (individual legal entity) where Competent Authorities require such capital ratios to be assessed - for such smaller entities where the levels of CET1 capital are lower and Defined Benefit Pension liabilities are much larger relative to the level of CET1 capital, the impact of the CRD IV Defined Benefit Pension Scheme treatment is likely to be much more severe than at EU banking group level.
• Netting: It is not clear if the EBA has applied full netting of all Defined Benefit Pension Scheme assets and liabilities for each institution when performing the analysis. In reality, Institutions may have a series of individual Defined Benefit Pension Schemes and may not be able to apply full netting across the individual assets and liabilities of each Scheme. This will increase the impact on CET1 capital; as such netting restrictions can increase the Defined Benefit Pension assets which have to be deducted under CRD IV.
Regarding the results of the EBA analysis, we consider that the outcome of average volatility on an annual basis around 20 b.p. is significant in magnitude.
For some of our associated Banks, annual movements of their capital ratios have moved around 20 and 50 b.p. historically. Therefore an additional 20 bp volatility introduced by pension obligations valuation would be a significant part of the capital fluctuations if not filtered or reduced via the adoption of an alternative CET1 capital treatment for Defined Benefit Pension schemes.
We also would highlight that there is a significant range across Institutions in the results achieved by the EBA for CET1 volatility impact and this needs to be considered when assessing significance to Institutions. For institutions with pension schemes which are partially funded or unfunded, the actuarial gain/loss is not compensated by a parallel move in the value of pension assets which would generate significant CET1 volatility under CRD IV.
Taking into account these considerations a cautious approach would be advisable. This approach could consist on the introduction of transitional mechanisms/filters to mitigate the potential volatility until a new assessment is carried on after an observation period long enough to ensure meaningful results (e.g. from 2015-2019).
As recognised in the EBA’s qualitative volatility analysis, the IAS 19 valuation of Defined Benefit Pension scheme obligations is very sensitive to both discount rates and inflation. As the duration of commitments is generally very long, a 50bps variation can lead to wide fluctuations on the total value of commitments (between 5% and 10%). This can lead to significant CET1 volatility of an institution under CRD IV.
Moreover, it should be noted that external Defined Benefit pension schemes invest in a diversified portfolio (equity, fixed income, sovereigns, real estate etc.) while the discount rate is defined on the corporate interest rate curve. These assets and the discount rate are therefore not correlated, introducing IAS 19 valuation volatility. Also, for Defined Benefit Pension schemes which are partially funded, the changes in the IAS 19 present value of gains or losses can differ significantly from the increase or decrease in the value of the pension assets simply due to differences in the magnitude of assets and liabilities.
As noted above, one of the key drivers of Defined Benefit Pension scheme deficits is the discount rate applied to projected defined benefit commitments. In times of economic stress, recent event have shown that the expected monetary policy response is the reduction of interest rates – such actions are likely to adversely impact IAS 19-based surpluses / deficits in pension schemes through reductions to discount rates. In addition, in such periods asset values are also likely to be adversely affected. Under CRD IV / Basel 3 this dynamic has the potential to be a significant source of pro-cyclicality to Institutions levels of CET1 capital, and from a prudential standpoint mechanisms to reduce such pro-cyclicality should be considered.