PDOs: is net value added good enough?

I recently came across a piece in which Keith Ambachtsheer argues that you must assess pension funds on value-for-money (linked here), and not the absolute level of fees or costs. The point of this piece hinges on a confession – namely that I, wrongly, read Keith as talking about ‘net value added’. This could be semantics, but I want to give Keith the benefit of the doubt as he has repeatedly extolled the virtues of integrated reporting (<ir>) which proposes assessing value creation through the lens of six capitals (and multiple time horizons) not just in terms of financial capital. Net value added, I would argue, looks very much like a financial-capital-only, single-time-period assessment.

What is the point? Well, I have been thinking recently about the size of pension funds, or ‘pension delivery organisations (PDOs)’ to use another Ambachtsheer term. For defined contribution assets, does any country need more than five (say) master trusts? Enough for viable competition, but sufficiently few to enable economies of scale to be harvested. I am beginning to settle on the belief that, as far as operational aspects are concerned, almost any single-employer DC arrangement is likely to be sub-scale and therefore inefficient. The arguments need finessing when we leave DC, but I believe the principles remain the same.

My beliefs regarding scale and investment performance are less settled – and particularly where the combination of operational economies and competitive investment diseconomies might fall. However it is the growth of internal investment teams within asset owners that I am finding interesting. I assume that the growth of ‘operational’ staff is relatively easy to judge and manage relative to the harvesting of economies of scale (cost per member should fall with scale). But how do we judge or manage the size of internal investment teams? With more staff, asset owners can pursue more complex investment strategies which offer, but do not guarantee, higher returns. But more staff also means more agents and more career risk. At what point do the management / employees capture the PDO and run it for their own purposes? If investment returns are always strong, then maybe this concern never becomes material. The financial capital lens suggests that we can safely ignore the high absolute costs, because the benefits are even higher. But if the PDO’s investment returns are weak for a period then not only will the financial capital lens show red ink, but we may also find that the social capital is in serious deficit too. At that point the governing board could find themselves with a serious headache.

So at the margin I do disagree with Keith, in that I think the absolute level of PDO costs do matter. In most cases PDOs are profit-for-member entities and so are not subject to the market discipline facing profit-for-shareholder entities. It is therefore relatively easy to add cost under the cover of enhanced net value, but I suspect much harder to reduce cost. My thinking up to this point has been about the number of employees, but I can’t resist a brief mention of compensation and incentives. How should a PDO compensate its staff? Let us assume the same base pay and a spectrum for variable pay ranging from 0% (pay for the job) to 200% (pay for performance). Beliefs (and values) can (and do) differ about the extent that the investment return streams will vary as a result of the incentive structure chosen. But when it comes to the difference in the risk of internal agency capture I think there is only one answer. High variable pay means a significantly higher risk that the employees run the PDO for their own benefit.