In the Netherlands, pension reforms are slowly picking up steam. Long overdue and in many respects a sensible direction, pension boards can choose between variants that redistribute investment risk between individual participants and the collective pension scheme. For boards trying to shape the transition, it is helpful to share experiences to create sustainable investment approaches that incorporate the latest thinking in behavioural research and decision making. This article describes how SPH, the pension fund for general practitioners, approached this.

Positioning the fund

With €12bn in pension assets, and over 22,000 participants, the scheme could be considered a hybrid between a collective defined contribution (CDC) and defined benefit (DB) scheme. Compared with other funds in the Netherlands, the nominal payouts were relatively low. However, this freed up matching assets to be allocated to the return portfolio, to generate relatively high yearly conditional indexations.

This approach resulted in a combination of structurally high coverage ratios as well as high indexations. An additional benefit is that the fund in recent years maintained its regulatory coverage ratios and therefore did not have to announce that pension benefits might be reduced.

The financial crisis in 2008-09 was a pivotal moment for asset allocation strategies. The high allocation to return assets resulted in a strong drop in cover ratio. Although nowhere near regulatory limits, it was clear that work had to be done on two fronts.

First, there was the focus on efficiency and increasing the grip on so-called ‘principal-agent’ issues. From 2013, we cleaned up the investments by rationalising mandates, lowering costs and increasing control measures, like most funds in the Netherlands.

While relevant, perhaps the more important issue was to tackle balance sheet management. Most Dutch pension funds continued and today still have the same approach to asset allocation – a static strategic asset allocation mix, complemented with a policy that linked the level of interest rate hedges to interest rates. In other words, business as usual.

What set SPH apart is that the board did not assume business as usual. Like a general practitioner where making choices based on evidence is key, the board undertook a soul-searching approach after the financial crisis, facilitated by risk advisory firm Cardano. Not simply hoping that things would revert to ‘normal’ (in terms of equities and interest, for example), the board reviewed academic literature, took a hard look at the drivers of the crises, mistakes we and other funds made, and analysed what was needed to redesign the asset allocation approach. This led to four design choices.

Design choices

• Define your circle of influence, focus on the impact. It is great to have a discussion on central bankers and currency, but as an investor this is outside your influence sphere, and as a board member you can only focus on the impact this has on your own balance sheet. Will you withstand these shocks? At SPH we only discuss and manage what we can influence. So we have eliminated this part of the discussion, freeing up valuable time for board and investment committee to manage the balance sheet.

• Pre-commit yourself to difficult decisions. A sensible conclusion from behavioural literature would be that boards should basically be training 364 days for that one difficult day during a crisis when markets are volatile, outcomes are uncertain and risk aversion among boards and advisers is running high. We have applied this insight consistently in balance sheet management. We define difficult decisions beforehand, and act on them, ranging from how to rebalance during normal periods and stress to how much risk we should allocate given the level of cover ratio. The amount of risk capacity and tolerance is not static over time, neither should the asset allocation.

• Make risk-savvy decisions. A lesson learned from the financial crises as well as behavioural finance, is that boards and investors are unable to grasp and steer on percentages and probabilities, which become useless anyway when stress increases. So we should not and would not. They are important, but for a long-term perspective. The board bases the risk budget on a deterministic, plausible short-term financial shock scenario – how much can we absorb before our fund goals get into trouble (the impact part). The investment staff then re-engineers this to a workable asset allocation, and tests in an asset liability management context whether this leads to long-term desirable pension outcomes. In other words, the risk budget is key for asset allocation, not the other way around.

• Nurture a healthy risk culture. Complacency, confirmation bias and overconfidence are all avoidable but need constant attention. Major decisions are only taken after conducting a pre-mortem, investment cases are challenged on their existing evidence, not desirability, and the investment committee’s meeting are structured to avoid known pitfalls.

These ideas are not new. They are widely accepted and researched extensively and confirmed over the years. What sets the fund apart is that we take them seriously and apply them rigorously to the balance sheet and investments. An open and inquisitive board culture is crucial. Most of our board members have a medical background; with all policies they want to know what the evidence is, and the more consistent and stronger the evidence is, the more compelling the argument to apply it.

This policy has been in place for almost five years. All the changes were set in motion to generate more stable returns – the same returns with less volatility. Also, the board is fully in control, focusing on choices that matter, and the fund optimally uses the governance budget. Neither Trump, Brexit, COVID and the volatility in interest rates were predicted but the policy accommodated them. Although not set as goals, SPH was one of the few funds in the Netherlands that has steered far away from any discussion on decreasing pension rights due to low cover ratios.


The board, investment staff and advisers have grown to work effectively in this model. Yet it might not be suitable for most funds. Board members and advisers from other funds visit us regularly, are curious, agree fully with the assumptions, but are hesitant to follow up on the consequences. SPH departs from current practice in the pension sector, and while this is innovative, it might not always be to everyone’s taste. One of the reasons is probably that our underlying world ‘vision’ is that we operate in volatile, relatively unpredictable markets where some assumptions more or less hold (equity risk premium) but we should be extremely prudent with other ones (mean reverting elements), base our choices on existing research and results and thus focus on impact – prepare for the worst and hope for the best, rather than hope for the best and ignore the worst.

Looking ahead, as the Dutch pension sector is shifting towards DC, will this framework still hold? We think it can and it should, because it combines the best insights in portfolio construction, behavioural finance and decision-making has to offer. The board has implemented the DC scheme five years ahead of the schedule laid down by the pension reforms, and in discussions with stakeholders, we found that for the construction of the life-cycle, return and matching policies, the same principles apply, albeit in amended form. Based on our experiences, we hope that the pension reforms generate more insights from other funds on how to successfully merge behavioural insights, investments, and balance sheet management, allowing us to learn and improve decision-making. All with the single goal of improving the results for our participants.

Alfred Slager is board member of SPH and professor of pension fund management at TIAS School for Business and Society, Tilburg University. Pieter de Graaf is CIO at SPH