Since the PRA published its approach to setting Pillar 2A risk capital back in 2015, most firms have based their Pillar 2A assessment on the Herfindahl-Hirschman indicator (HHI) approach as recommended by the PRA. The requirements are prescriptive and involve calculating the add-on for certain exposures based on pre-defined geographic, single name and sector categories.

What are the issues with following the PRA's approach?

The main criticism of HHI is its simplicity, which gives rise to add-ons that are less relevant to firms? business models. For example, the HHI uses a geographic breakdown which includes the UK and eight other regions and specifically excludes residential mortgages. This may be a sensible split for a large international bank or a well-diversified UK building society, but it does oversimplify the actual concentration risk by ignoring regional concentration and concentrated mortgage portfolios in one particular area.

The example below illustrates this point where a firm has an £80 million portfolio of corporate exposures - in Western Europe (£20 million), Scotland (£20 million), London (£20 million) and the rest of the UK (£20 million).

Current HHI approach                                                           

 HHI approach with additional UK regions

The impact of adding additional regions is significant and it greatly reduces the HHI factor, resulting in a reduction in the Pillar 2A capital midpoint add-on from 1.03 per cent of RWAs to 0.35 per cent although there is no change in the underlying risk. This issue is amplified for small UK focused banks where the UK concentration risk HHI requirement is the most significant add-on (normally 1.33 per cent of the relevant credit RWAs).

How do firms manage concentration risk?

Approaches to concentration risk vary from firm to firm and there is still inconsistency in the way firms apply HHI concentration risk in their ICAAPs. Some of the approaches firms have taken include:

• Calculating HHI for all exposures on their balance sheet (all products/exposures) rather than following the product-driven approach set out by the PRA.
• Partly applying HHI, for example, not calculating the geographic add-on for a UK-only business.
• Not calculating HHI and stating that it is inappropriate for the firm's business model. For example, firms that are only exposed to one sector and one geographic area argue that the HHI method is not appropriate as it would result in 100 per cent HHI and therefore they use an internal method to demonstrate more granular diversification. 

A sample of SREP reviews have shown that a number of firms have not developed adequate risk appetite and limit structures within their risk management framework to support the management of concentration risk.

Not following the PRA's HHI approach has resulted in firms being penalised by SREP add-ons for concentration risk, coupled with scalars to address management deficiencies regarding concentration risk appetite.

How can firms improve the management of Pillar 2A concentration risk?

Some common themes:

• Firms should, in line with the PRA guidance, ensure that the HHI method is included as a ?floor? in their Pillar 2A assessment and that the outputs are reconciled to the FSA078 return and COREP capital adequacy and large exposures templates.
• Implement risk appetite and concentration risk policies that fully incorporate the HHI model as a way of optimising Pillar 2A and pricing products correctly. This is mainly done through aligning limits and key risk indicators with the HHI requirements. Some firms calculate the HHI impact on a deal by deal basis and add/allocate capital accordingly.
• Having internal concentration risk limits aligned to the firm's risk appetite in addition to the HHI limits/metrics. This would, for example, include additional regions and sectors in line with the institution's overall risk profile.