Liability driven investment
It does seem strange that it has taken so long for so many people running pension schemes to realise the importance of creating a liability driven framework when establishing their fund’s investment strategy.
Until recently investment risk tended to be thought merely in relation to deviation from a stock index. Now pension funds realise the importance of developing their own liability benchmark and that, for them, risk is really the mismatch between the way assets perform and liabilities change.
Of course, there is no such thing as a perfect asset liability match. First you have a duration problem, then you have mortality issues and, in particular, longevity risk. In addition benefits can and do change and it is very difficult to immunise your fund against cost increases when benefits are related to unknown future salaries. However if you do not understand your liabilities it really must be difficult to set an appropriate asset strategy.
For most funds it would be entirely inappropriate to simply invest in bonds, there are simply not enough bonds around with the right duration. Yes, I know that banks tell us that we can swap our way out of trouble but this will only take us part of the way and does not solve the longevity and final salary liability issues and may even increase other risks. In any event few funds can actually afford to lock into bonds – their sponsoring companies cannot pay the increased costs that would result – but that takes us to a different question and the reason why many corporates are moving from DB to DC pension provision.
Of course we know that members want security and fund sponsors want affordability and trustees have to sit in the middle trying to marry these different objectives.
Paul Myners in his famous report for the UK Treasury published in March 2001 really helped trustees by recommending they concentrate on the important issues and in particular said that “Trustees should set out an overall investment objective for the fund, in terms which relate directly to the circumstances of the fund, and not to some other objective, such as the performance of other funds.”
Many funds thought they already did this but in reality it has taken a long time for trustees to stop looking over their shoulder at what other funds do and in particular how they have performed. Of course, funds gain comfort from the fact that they are doing what other schemes are already doing, so that if things go wrong they are not alone.
It is like the old argument that nobody in the seventies or eighties got fired for buying IBM stock – perhaps they should have been!
In any event now we realise, especially after the market crash, that assets should bear some relationship to liabilities, so the question is now – how close a relationship and where can we deviate? Now many actuaries start with the premise that bonds are the closest match for pension scheme liabilities and that any deviation is a risk. I don’t think this is quite true as any efficient frontier always seems to show that bringing in a small proportion of equities or indeed almost any other asset class will reduce risk.
Whatever we decide is the minimum risk portfolio, in order to deviate from it we must establish a risk budget for our scheme.
Neil Walton, a senior consultant at Mercer Investment Consulting believes that when setting a risk budget: “Objective setting is probably the most challenging part of the process”, although, first you must define what one is trying to achieve in managing the scheme’s investments. Walton be-lieves that as the objectives of trustees (who want to provide security) and fund sponsors (whose key need is affordability) conflict, so we need to consider ranking and trade-off. This he believes is an iterative process involving three elements for each objective: each players’ tolerance level, time frame and the probability of breaching the tolerance levels.
The aim of the exercise is for the fund to target a particular funding level to be achieved in so many years, assuming an acceptance level of contributions expressed as a percentage of pensionable payroll. To do this we must be able to manage the risk of the funding level falling below a set level at any point in the process or at least over a set number of years. For some schemes this process will be relatively painless. The trustees of other schemes will however realise the enormity of the task that lies ahead of them.
However, for me, the key lies in the assumptions trustees use in these calculations. I do feel that many trustees are still being hopelessly optimistic in the assumptions they are willing to accept for actively managed equities. Maybe too many managers are accepting mandates that require them to produce performance above their benchmarks that are beyond their capabilities. But that, as they say, is a different story.