The UK has nearly 36,000 trust-based DC pension schemes, over 90 percent of which are tiny (with fewer than 12 members).
Only 120 have more than 5,000 members. In addition, there are some 5,900 DB schemes, 80 percent of which have fewer than 1,000 members. Conversely, only 392 schemes (fewer than 7 percent) have more than 5,000 members.
Meanwhile, the pursuit of scale is evident in many countries, notably Australia, where the contraction in the number of DC schemes has been particularly marked. 25 years ago, there were over 10,000 schemes; today, there are just 236 regulated schemes with more than four members (average membership is some 39,000), and a diminishing shoal with four members or fewer (accounting for perhaps 2 percent of pension assets).
Denmark and the Netherlands have also actively pursued scale. The number of Dutch pension funds has halved in the past ten years (to 365), and five funds now have roughly 7.5 million members between them, with over 72,000 affiliated employers and over 50 percent of total pension assets.
One reason why UK pension schemes have not scaled up is that it is not in the interests of the financial services industry to encourage it. Fewer schemes means fewer clients for professional service providers (actuaries, accountants, lawyers, trustees and a whole gamut of consultants), as well as fund managers, insurance companies, platform providers, and others.
In addition, large schemes can attract the talent necessary to make them expert clients of the market, and use their purchasing power to negotiate prices down. They are more inclined to build in-house teams to manage assets, to the detriment of external asset managers. Last year, for example, the California State Teachers scheme moved $13bn of its assets in-house (and plans to move $20bn more). Denmark’s ATP pension fund recently pulled £2.2bn from external asset managers, halving the amount of money with them and, similarly, Sweden’s AP2 fund has moved $6bn in-house, over the past three years. Meanwhile, the UK’s Railpen cut its external equity managers from 17 to two (saving £50m in annual costs).
In theory, trustees are uniquely well positioned to drive the consolidation of pension schemes. In practice, running a business is incompatible with acting in a fiduciary capacity, and many do not really behave as the principals they are supposed to be. Indeed, the FCA now recognises that trustees are conflicted: fewer schemes means fewer clients.
Consequently, regulators should consider requiring anyone performing a governance (or fiduciary) role to adopt mutual status and adhere to criteria that include meeting a benchmark cost per scheme member. If not met within the three years, say, then mandatory merger should ensue. Tougher governance is what drove many Dutch scheme mergers. Meanwhile, trustees should be required to publish, annually, the actions they have taken to evidence that they are really acting in members’ best interests (including confronting the principal-agent problem).
The UK needs fewer than 100 DC schemes to serve the UK’s workforce. Merging DB schemes is more challenging, but asset pooling surely beckons (note the on-going pooling of LGPS assets). In many consumer arenas this is the age of the customer. It is time that this applied to pensions too.