PensionReforms' summary and comments
In a Defined Benefit scheme, the people who run the sponsoring employer also tend to have considerable say in the appointment of the trustees who run the normally separate scheme.  We might therefore expect that the scheme's investment strategy could reflect the employer's way of doing business.

So it turns out; at least it seemed to be the case in the ten years to 2005 in the Netherlands, based on this December 2007 report.  Given that the employer is underwriting the balance of the cost of providing the benefits, this could be seen as a reasonable arrangement.  The fact that, in the Netherlands, half of the pension scheme's board "represent" employees makes this a more interesting finding. 

"This study presents empirical evidence on the influence of sponsoring companies on the funding and portfolio allocation decisions of their defined benefit pension funds.  Several hypotheses taken from the theoretical literature are tested using a microdataset of around 550 Dutch company pension funds over the ten year period 1996-2005, combined with a microdataset on 100 of their sponsoring firms.  The wide variation in funding levels over this period provides a natural experiment in the determinants of underfunding and portfolio composition.

The report describes four key findings:
". Pension funds have lower cover ratios when their sponsoring companies have high leverage...
". Pension funds have lower cover ratios when their return on assets is relatively low..
". Pension funds have lower cover ratios and receive lower sponsor contributions when their sponsoring firm is small..
". Defined benefit pension funds invest more in shares when their sponsoring companies have high leverage.."

The report notes that these findings accord generally with the literature on these topics.  The facts do seem to conform to the theory.

The report suggests that regulatory authorities should pay attention to the corporate balance sheet as well as to the Defined Benefit scheme's own balance sheet and actuarial reviews.

"It is notable that such links apply consistently in the Netherlands despite the absence of pension benefit insurance that gives rise to moral hazard on the part of the sponsor vis-à-vis the insurer. Such patterns are likely to be even more marked when such insurance is present, as historically in the US and now in the UK also."

PensionReforms suggests that the report's findings are much as might be expected.  In the end, the strength of a Defined Benefit scheme's promises depend not only on the economic strength of the sponsoring employer but also on the employer's willingness to take that risk on.  The report's findings suggest to PensionReforms another very good reason for governments to avoid offering 'insurance' arrangements for Defined Benefit schemes, like the UK's Pension Protection Fund or the US Pension Benefit Guarantee Corporation.  Given the employer's apparent, indirect capacity to influence the risk profile of the scheme's assets, this sounds like something governments should stay away from.

It would be interesting to re-visit the findings after the 2008 global economic crisis.   (File size 842 KB; 36 pp) 356