Lack of credit for dynamic hedging in QIS 5 driven by ideology, not economics

Old Mutual questions the lack of capital credit for dynamic hedging in QIS 5

The decision not to give capital credit for dynamic hedging programmes, in contrast to rolling hedging programmes, under the fifth Quantitative Impact Study (QIS 5) is driven by ideological considerations, not economic ones, says Andrew Birrell, chief risk officer of South African insurer Old Mutual.

The Committee of European Insurance and Occupational Pensions Supervisors (Ceiops) launched QIS 5 at the end of August, and it will run until November, with results published in April 2011. And while Ceiops has responded to industry concerns by taking a less conservative approach to several issues, dynamic hedging still attaches no capital credit.

Critics of dynamic hedging – in which a hedge is adjusted as the underlying asset moves – argue that it is either prohibitively expensive or near impossible to find counterparties in stressed markets, unlike rolling hedges which are inherently of longer duration. However, Birrell says the complexity of both rolling and dynamic hedges meant delineating between the two was fraught with difficulty and Ceiops has failed to adequately define this difference.

“The first point is that without a firm definition of dynamic hedging it is difficult to comment definitively – let’s wait and see before they do. But we need to be careful not to be wedded to ideology – and saying dynamic hedging is bad and rolling hedging is good, is sailing close to taking an ideological position.

“Instead Ceiops should be more interested in the economics of the hedge and its longevity.
If you have a hedge that goes beyond the 12-month view of Solvency II, it seems inconsistent not to take credit for it just because it is dynamic,” says Birrell.

Dynamic hedging is mainly the provenance of variable annuity providers rather than the traditional life products that dominate the European market, and a lack of capital credit under QIS 5 appears to act as a barrier to further market penetration. Lukas Ziewer, Dublin-based partner at consultancy Oliver Wyman, says Ceiops’ stance on the issue of dynamic hedging is part of a broader reluctance to move away from a simple downward shock to assets in the QIS 5 specification. 

“Ceiops didn’t want to go into the discussion of how companies will make dynamic changes to their asset portfolio in times of crisis. It’s a simplification to just say, ‘these are the assets you have and just apply a downward shock’. Though in its defence it would be an overly critical approach to say Solvency II is not sophisticated enough for hedging products.”

Ziewer says instead that  “clearly” Solvency II must have conditions in place where an insurer can take capital credit for dynamic hedging. However, he cautions that the very real risk associated with this risk-management approach must be included in any measure. Pointing to the significant losses suffered by European insurers such as ING and Axa on their hedging programmes in 2008 – and even larger ones experienced by their North American counterparts – he says any capital relief must be “real”.

“You must stop companies from saying they will sell assets when markets go down and then not doing so. There must be controls in place to see the company’s actual plans and measure the resulting capital. A dynamic hedge still includes market risk – you cannot hedge everything away.” 

Instead, Ziewer suggests an approach where companies determine what their capital relief is from hedging and they should apply a factor to that, less than one, to allow for uncertainty. That factor should be driven by a number of factors, including the chosen hedging strategy, “so a tighter hedging programme would be closer to one, whereas an approach that relies heavily on illiquid markets would take a lower factor”.

Ziewer emphasises that any solution to the dynamic hedging issue “must not be mechanistic”, and requires from the regulator both an understanding of the hedging approach and the previous experience of the company. “It is an area where judgement is required,” he says.

Another area of industry concern over QIS 5 relates to the issue of equity volatility stress. Previous versions of the specification included a test for this but it is believed that lobbying from smaller insurers saw these requirements dropped from the final version.

Market sources have told Life & Pension Risk that the CRO Forum was so concerned about this issue that it asked all members to run an equity stress alongside QIS 5 so the results could be used in future discussions between the industry group and the European Union.

A spokesman for the CRO Forum confirmed that it was seeking additional information from QIS 5, but declined to give specifics and says that none of this will be made public.