Home > News > 2003 > October 2003 > Determining an asset allocation with cashflow modelling
Determining an asset allocation with cashflow modelling
Chris Watts, an actuary in London's pensions actuarial team looks at the critical asset allocation decision.
This article first appeared in September's edition of Pensions World.
Introduction
Pension scheme trustees need to establish an investment strategy for their scheme's assets. However, the needs of each scheme is different and the strategy adopted will depend on a range of factors including anticipated future liabilities, the size of the assets held and the acceptable level of investment risk.
Setting an appropriate strategy involves determining an appropriate asset allocation. This will be followed by adopting a statement of investment principles, reviewing investment manager arrangements and ongoing performance and risk measurement.
In many cases a suitable asset allocation can be determined by a cashflow study, rather than a more sophisticated economic analysis of the possible relationship between the assets and the liabilities. This can provide a more cost-effective approach to asset liability modeling, and this simpler approach may be a more appropriate alternative for some pension schemes.
Cashflows
To conduct a cashflow study you need details of estimated projected outgoings. Today's actuarial valuation software allows projected cashflows to be relatively easily and cheaply produced.
The cashflows then need to be analysed, with a view to producing an investment strategy which will aim to match expected expenditure, maximise long term investment returns (subject to certain risk considerations) and allow sufficient flexibility in unexpected situations.
Projected liabilities, not projected net cashflows, are the most important factor in determining the matching investment strategy for existing funds. It is, therefore, important to distinguish between those of the future liabilities that might be thought of as relating to the existing funds and those that might be thought of as being met by future contributions.
Chart 1 shows how an example pension scheme might split its anticipated future liabilities between those to be met by existing assets and future contributions. The chart assumes that the assets and liabilities of the scheme are in balance.
If there is a surplus or deficit thought should be given as to how this might affect the actions of the trustees or sponsoring employer, and how this might affect the projected cashflows. For example, if there were a surplus a contributions holiday may be taken or discretionary pension increases awarded.
Alternatively there may be a deficit. Chart 2 shows an example of the outgo of a scheme with a projected deficit. In this case I have assumed that the trustees would begin to cut back benefits in 3 or 4 years time. Working out such a precise course of action may well be impossible in practice, although thought should be given to how the existence of a deficit might affect future cashflows.
Having modeled future liabilities we then need to consider which assets should be used to match the various cashflows. For now we will consider only two broad asset classes, equities and bonds/cash. Other asset classes can be incorporated later in the analysis.
To help decide how the assets should be split between these two broad classes, we have conducted some analysis on the historical outperformance of equities over gilts. This is often referred to as the equity risk premium, because it represents the additional return investors get for accepting the higher risk associated with equities.
Equity Risk Premiums
Compared to bonds, equities have shown a much higher volatility of capital value over the short term, but higher total returns (the combination of both capital growth and income). An investor will typically invest part of his fund in bonds, to secure assets to meet short term liabilities, and part in equities to try and generate high long term returns.
Bonds represent the "low risk" asset because they share the most similar financial characteristics to future liabilities. The capital value of bonds varies as interest rates change, so they are more volatile in absolute terms than cash). However, if bonds are purchased with a view to providing income at the same time as future expenditure then the value of income and outgo will move in a similar way.
There are two key questions when deciding how much exposure one should have to equities and bonds. First, over what future periods do we expect equities to outperform bonds, and second, what long term outperformance can we expect?
Although history is not always a good guide to the future, it seems a sensible place to start. For example, over the last 50 years (1953 to 2002) the UK stock market outperformed UK fixed income gilts by an average of over 5% per cent a year.
The spread of results making up this average is also worth considering. Over the same period the UK stock market outperformed fixed income gilts in more than nine out of ten periods of seven years, and more than three out of four periods of three years. Chart 3 summarises the results of this historical analysis.
This chart is based on data from the last 50 years. It shows, for example, that equities have outperformed gilts by more than about 3 percent per annum in three out of four ten-year periods. Correspondingly, in one out of four periods they have outperformed gilts by less than 3 per cent pa. Meanwhile, looking at periods of 1 year, equities have underperformed gilts by more than 7 per cent pa just one time in four.
We can see that equities have shown a relatively high level of risk over short term periods. "Short term" for most investors could be defined as up to perhaps seven, ten or more years, depending on how conservative a view is taken, and it is the expenditure for this period that should be backed by cash or bonds. Payments due after that can be backed by assets expected to deliver higher long term returns such as equities. The longer time frame means that there is a relatively low risk of equities under-performing cash or bonds.
Fine Tuning
It is also unlikely that the trustees will want to continually review their asset allocation. However, it is quite possible that the appropriate investment strategy should change over time. It is worthwhile projecting a scheme's finances, including income levels, forward a few years (perhaps until the next investment review is planned). The analysis can be reworked to determine what the recommended asset allocation might be in a few years time if the projections were borne out.
This would give an indication as to how the recommended strategy may need to change over time. If a significant change is anticipated then a dynamic rather than a static investment strategy may be appropriate. Alternatively this might mean that more frequent reviews are needed.
The trustees might also want to consider how a particular investment strategy will affect the volatility of the scheme's funding level on a particular basis (or two!).
What we have considered so far will help to determine a recommended bond/equity split. There are a number of other potential issues that may require this strategy to be fine tuned. Larger funds may be better placed to cope with the extra expense and complication that some of these considerations will involve.
In particular a scheme may consider the precise nature of its bond investments. Depending on a scheme's liabilities, bonds of a particular duration to maturity might be more appropriate than others. Similarly, a scheme may like to invest in corporate bonds as well as gilts, or invest in a mixture of bonds that provide fixed and inflation-linked returns.
If a scheme has a large investment fund it may also consider investing in other asset classes, such as property, overseas assets or "alternative asset classes" such as hedge funds or private equity. Investing in different asset classes can help to diversify a scheme's investments and may reduce volatility of investment returns.
This article first appeared in September's edition of Pensions World.
Chris Watts, October 2003.