How do you ensure accountability without removing the judgement that good decision-making requires? There is no cost-free answer; metric-based systems cost less to administer but can distort behaviour, while qualitative supervision is more flexible but can place greater demands on regulators and schemes.
The VFM framework was introduced to address the problem of high charges for poor investment returns, with no consistent way to identify underperformance. The framework’s traffic-light rating is designed to make comparison possible, and to push underperforming schemes to consolidate or exit the market.
Louise Farrand of DCIF warns in Professional Pensions that standardised metrics discourage trustees from deviating from the peer group average, penalising long-term risk management and deterring investment in private markets. Louise warns that forward-looking performance metrics may pressure schemes into presenting optimistic projections to avoid negative ratings, which would increase rather than reduce herding.
Australia’s pension scheme performance test highlights the tensions; it identifies underperforming schemes and encourages consolidation, but Australia’s Treasury review found it produced short-termism, benchmark hugging, and metric gaming. Industry Super Australia reported that poorly performing funds shifted assets to categories with lower-return benchmarks, and exploited fee loopholes to improve test scores, without delivering better returns to members. Attempts to address this by extending the assessment period and recalibrating benchmarks have not yet resolved it.
The main alternative is the Dutch approach. Rather than applying the same metrics to every fund, the regulator focuses on higher risk funds, using a combination of standardised assessments and direct supervisory dialogue with boards.
Unlike in master trusts, Dutch multi-employer schemes tend to include employee, employer, and pensioner representatives on their boards, and often have separate advisory bodies. Accountability is built into the structure of the fund itself, rather than imposed through external metrics, and results in diversity of investment between schemes.
The DNB model is expensive; boards of small pension funds spend over 30% of their time on regulatory requirements, and indirect supervision makes up 75% of small schemes’ costs. The Netherlands is also in the process of moving from CDC to DC, which limits how directly its experience transfers to the UK DC market.
However, it does show that the choice is not between measurement or no measurement: it is between a system that costs less to run but risks rewarding conformity, and one that is more demanding but better placed to support investment strategies tailored to members’ needs.
Pensions Goth
Leave a Reply