10 January 2012
Richard Butcher writes for Engaged Investor
Richard Butcher considers one way to improve DC, in the first of a new series of blogs for Engaged Investor
Trustees have to look after the financial interests of their members. Recently, in the context of defined contribution (DC) schemes, the Regulator has been helping trustees to focus on this by setting an objective of ‘delivering good member outcomes’. The regulator’s intent is good (albeit not terribly well defined) but they also miss the point.
When you drive a car, you look at the road ahead and use the controls to achieve your objective. You do not look at the controls with an objective in mind and then drive the car. If you did you would crash. Probably before leaving your drive. Yet we ‘drive’ DC schemes in this way: we look at the controls of contributions and investment with our objective only vaguely defined.
To achieve good member outcomes this has to change. We have to focus on the objective not the controls. Professor Robert Merton spoke about this at the last NAPF investment conference, coining the phrase ‘managed DC’. A similar set of arguments were put forward in research carried out by the Cass Business School and by Lord Hutton.
How could managed DC work in practice?
In principle it’s not difficult. We ask the member to define what benefits they want (say, a pension of £100) and when they want it (say age 65). We ask them what margin of error they could tolerate (say +/- £10). Under the bonnet we stochastically model that objective. The result is the level of contributions required. If the contributions are too high (and the member can now make an objective decisions about contributions), they have to lower their target, lengthen their plan or increase the margin for error (i.e. take more investment risk).
The modeling wouldn’t be a one off. It could be web based and done, in real time, at any time. The model would imply a cone along a time line, narrow at the point of retirement and progressively wider before. Progress along the cone would be tracked. Near the bottom and more contributions may be needed. Near the top and you can relax. Out of the cone, time to reconsider the objectives.
Under the bonnet, this approach would lend itself to, for example, dynamic derisking, derivative usage to manage down side risk, target date return funds and many of the other innovations that are now common in DB investment.
It sounds complex but it isn’t. All the member would see upfront would be £100, +/- £10 and age 65 = £x a month. When they review it all they see is, ‘You are currently on target to hit £98’ (or whatever).
A sceptic may argue that we would be creating a DB scheme by the back door, but this wouldn’t be the case. It isn’t beyond our ability to make the model pretty good and to come up with a form of words that makes it clear there is no guarantee. In any event, the approach is far better than just driving in the dark (as we do now).
Nor is this approach intended to remove choice. This is aimed at the 95% of members who end up in the default fund. Those who want to play with their pension investments still can. In other words, this is entirely consistent with the objectives of providing robust and proactive defaults in an environment where members can make informed decisions.
DC doesn’t work well. This is one very effective way of fixing it. If we don’t fix it we will end up with dissatisfied and possibly even litigious members. If we don’t fix it, we are heading for a car crash.