A peculiarity of the Dutch pension system is that the bulk of pensions are occupational: key choices for participants are made collectively in collective labour agreements (CLAs). This limits the flexibility for individuals, but similarly employers are legally bound by agreements negotiated with trade unions on their behalf.

Since participants and employers are effectively captive clients, there is no market discipline for pension funds. Hence the need to provide discipline from a regulator, so there is some escape hatch to exit funds that fail to meet the needs of their stakeholders. Funds that significantly underperform an investment benchmark can, theoretically, be disqualified from the CLA mandate

This had become an entirely theoretical exercise, and a prime example of regulatory capture. Funds were effectively setting their own benchmarks and excluding a large part of their performance—allocation choices, hedging policies, as well as private market investments—from the calculation. I am not aware of any fund that ever failed this test, though many will have had anxious moments.

New law (WTP, the Future Pensions Act) justified a review of the performance test, and the ministry just held a consultation (NL). My comment is that this test should be a meaningful benchmark for the entirety of the fund’s activities—not only a benchmark of a fund’s investment management as currently, but evaluating all its policy choices as well. CFA has a list of characteristics of good benchmarks1, that says

In essence, a high quality benchmark or index should be

1. free of conflicts of interest;
2. provide independent review/pricing; and
3. have transparent methodology.

Judging by the proposal and the submissions, this is surprisingly hard to do well.

I also have a proposal2 that I’ve discussed before: buried deep in the law’s description of risk appetite is a prescription for how to describe a fund’s risk-free reference.