17 October 2012

IPE: Brussels bows to pressure over IORP II confidence levels

The European Commission has recognised the importance of allowing pension schemes participating in the first quantitative impact study for the revised IORP Directive to submit results based on a Value-at-Risk below 99.5 per cent confidence levels, after the measures raised concerns in the industry.

Writing to the European Insurance and Occupational Pensions Authority (EIOPA) last week to give its green light for the launch of the QIS for the implementation of a holistic balance sheet (HBS) within the IORP II Directive, Jonathan Faull, director general of internal markets and services at the Commission, said it was "important" QIS participants submit results based on a VAR at the 99.5 per cent, 97.5 per cent and 95 per cent confidence levels.

The move came after the pensions industry recommended EIOPA conduct stress tests at two other confidence levels rather than conducting stress tests at a 99.5 per cent level only.

In its response to the consultation paper on the QIS exercise conducted by EIOPA this summer, the Dutch Federation of Pension Funds recommended deriving results at 97.5 per cent and 95 per cent security levels.

The EIOPA's Occupational Pensions Stakeholder Group (EIOPA OPSG) echoed those remarks, arguing that it was unclear how EIOPA would infer other security levels from the calculations on a 99.5 per cent level. "Due to the (option) valuations of the adjustment and steering mechanisms, other security levels cannot be derived from the 99.5 per cent level since there is no longer a normal distribution", the group said.

EIOPA OPSG went even further in its response, saying the events of recent years had shaken confidence in the underlying assumptions of the HBS. "Indeed, these model risks are not new", the group added. "Unfortunately, they did not receive adequate consideration when VaR became a prudential tool for the banking and insurance sectors."

Additionally, Faull stressed in the Commission's letter to EIOPA that the matching adjustment of the discount rate was not part of the benchmark scenario, but would be tested as an option.

Full article (IPE subscription required)

© IPE International Publishers Ltd.