CDC pensions are better for the economy
At the TUC’s ABC of CDC conference, my good friend Bernard Casey of Warwick University asked Gregg McClymont, the Shadow Pensions Minister, the sources of the superior projected performance of Collective Defined Contribution pension schemes (CDC) compared to individual Defined Contribution (DC).
“Different speakers have emphasised the “main” advantage of CDC. One said it reduced volatility. Another said that it generated higher returns because it allowed investing long term, and in higher risk assets, and a third that it generated better returns because, via scale economies, costs were lower. Which is it?”
The answer, of course, is all and none. All are sources, but the dominant one in practice will vary with circumstances.
The various projections that have been undertaken, by the Government Actuary’s Department, by Aon Hewitt and by the Royal Society of Arts, have used models. Therefore the design of the model and its calibration will have determined the answer. In fact, as the order in which effects are considered in an attribution analysis determines their magnitude, the question of a unique most important source is not even well framed.
Bernard’s question and its target were a piece of pure devilment, of which I might have been proud myself. The various projections are all concerned with scale and scope in pension management.
However, there is a more important aspect to CDC as a form of organisation for pensions. We have known since the work of Teresa Ghilarducci that how pensions are organised can have effects upon the broad economy.
Empirical work on US data showed that individual DC is more procyclical than (collective) defined benefit. In other words it exacerbates economic downturns and heightens booms. The driver of this analytic result was rather more the collective risk sharing of DB than the pooling aspect – most individual DC assets are invested in collective mutual funds.
At the conference Steve Webb, the Pensions Minister, repeatedly emphasised the risk pooling aspect of CDC to the point that the audience might have thought that risk-sharing among members in a CDC scheme was entirely absent. It isn’t – well designed CDC schemes are both risk-sharing and risk-pooling.
CDC schemes allow greater commitment, in both amount and over time than individual DC, even after considering the collective nature of DC investment funds. The key insight here is that it is long-term commitment that allows industrialists and entrepreneurs to create further wealth. Simply put, greater commitment can be expected to deliver greater wealth, a larger pie from which pensions may be delivered.
There are a number of approaches to the analysis of commitment. We might have used Claude Shannon’s 1948 work on information theory in combination with Georgescu-Roegen’s 1971 “The entropy law and the economic process”, but that is perhaps too abstract for pensions practitioners. It would though deliver the insights that although all wealth is information, not all information is wealth, and that commitment is ineluctably related to irreversibility.
Colin Mayer recently wrote an excellent book “Firm Commitment: why the corporation is failing us and how to restore trust”, which offers a corporate finance perspective on these issues, but more directly relevant is Pankaj Ghemawat’s 1991 “Commitment: The dynamic of strategy”. From Colin Mayer, we may note the relation between trust and commitment; and from Ghemawat, we may distinguish between strategy and tactics, and by extension between investment and speculation.
The irreversible nature of commitment, or trust, makes it risky. Indeed, the analysis may be expanded to consider soft and hard commitments, the value of flexibility as an option, and decisions that are, rather than absolutely irreversible, merely costly to reverse, which is the problem more usually faced by investment managers.
Several analysts have purported to address empirically the question of whether CDC schemes do, in fact, invest more for the longer term or are more committed. They suggest that conservative asset allocations are observed. However, there are two problems with these studies: the data are derived from legacy Dutch DB schemes that many have wrongly rebranded as CDC, and secondly, and more importantly, these schemes are subject to a strict regime of risk-based regulation (FTK) as if their liabilities were hard promises.
Solvency ratios are the heart of this FTK regulation and simply to calculate a solvency ratio it is necessary to attribute an estimate to liabilities. Indeed, the Dutch regulator has required cuts in pensions in payment. Against this background, it is perfectly sensible to maintain conservative, low volatility, asset allocation strategies. As we are still to see the detail of the UK CDC regulation, we should bear in mind this caution when thinking about that.
Many commentators have raised the sustainability of CDC – ‘if I am a young employee, why should I join a scheme that is already in deficit?’ is one way that this is often put. It is not clear how potential members become aware of a possible shortfall or indeed if that concept has any meaning given the absence of hard pension promises.
The reality is that younger members face far more risk and uncertainty than older members – and the magnitude of that risk and uncertainty can be estimated from the high relative cost of buying deferred annuities for that group.
They also need to consider their alternatives – if the investment performance of their individual DC choices is as much lower than CDC, as the various projections suggest, the deficit in a CDC scheme would have to be substantial to offer worse potential outcomes and justify not joining. Finally, as this article has pointed out, if these employees do not participate, they are facing a future that is, by their own making, less rich and satisfying.
Note:  The cuts to pensions in payment seen in the Netherlands need never be seen elsewhere. If the notional accrual rate that defines a (soft or hard) pension promise, rather than the amount of pension in payment, is used as a control variable, the balance between assets and notional liabilities could always be maintained. Indeed, using it as a control variable offers the greater certainty of smoothing of pensions over the working lifetime of an employee. This method is a projected rather than accumulated benefit obligation forecast, and is appropriate when we are considering action today in mitigation of possible future developments.