It is difficult to argue against the notion that funded pension benefits should be well capitalised. Those who argue that the benefits should be more secure would say this is so precisely because they are such an important part of a person's lifetime earnings. But there are plenty of arguments to say there needs to be an element of risk in the assets precisely because the ambitions of the pension promise are so great.

But the EU single market commissioner, Michel Barnier, speaking at the European Commission's public hearing on the IORP Directive, spoke of the "hyperbole" of claims for the economic cost of Solvency II-type rules if applied to occupational pensions. He was undoubtedly referring to the estimates of JP Morgan and PwC relating to the cost of Solvency II rules if applied to pensions. Or those of the Dutch minister for social affairs, Henk Kamp, who said that a move to 99.5% certainty under Solvency II would cost his country's employers and employees €11bn in extra capital buffers.

PwC estimates the potential cost to the UK could be as much as £500bn (€600bn), "depending on how much leeway there is for healthier businesses". The CBI estimates that schemes that comply with Solvency II would need to offload more than £800bn of equity, while JP Morgan says UK companies will be forced to pour another £600bn into their defined benefit (DB) schemes.

High solvency margins imposed on risky assets would have a detrimental effect on the sponsor and the employer, which would have to pay more for a spurious notion of certainty. The value-at-risk (VaR) models behind the notion of 99.5% certainty are themselves questionable; the 97.5% certainty rate of the Dutch FTK has not prevented benefit cuts and there is no proof that a higher certainty rate protects benefits from exogenous shocks. In any case, VaR was not conceived to determine the security of occupational pensions. The actuarial and investment assumptions on which occupational pension promises rest are furthermore mutable, not fixed and measurable.

Those few countries - the Netherlands, the UK and Ireland - with sizeable occupational pension assets have already addressed this issue, and the markedly different ways they have done so only highlights the difficulties in imposing harmonised solvency rules. This is not to mention the scope issue: so few countries would be subject to IORP II and so many pension systems exempt, such as Germany's book reserves, that the legislation is barely justifiable. Indeed, only 11 of the 27 EU countries responded to EIOPA's call for advice.

Throughout the 1990s, the occupational pensions industry pushed for a directive that would create the ideal pan-European pension fund, as promoted by the likes of Unilever, which was a strong supporter. In reality, there are fewer than 100 cross-border pension schemes under the IORP directive, and the world has moved on from this concept. Denmark's ATP now operates in the UK as an occupational pension provider, but it did not need IORP to create a sterling-hedged segregated account in Denmark to run British pension assets or to create a trust vehicle to oversee the scheme. Nor does UK retailer Tesco see the need for a pan-European IORP.

The 2003 IORP directive is, indeed, flawed in that it references Solvency I. But that is not sufficient grounds to impose an IORP II with expensive copper-bottomed promises that impose significant costs on employers and employees.

Countries with an occupational pension tradition are moving away from old-fashioned DB schemes towards greater conditionality. The premise is that investment risk is needed to achieve these goals. The capital cost implied by Solvency-II is too great for a European economy that desperately needs to promote funded pensions. Applying too high a certainty rate would make occupation pensions the preserve of the wealthy few.