[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

The foundations of charity

uch discussion has occurred in the past as to the application of asset liability models to pension funds; this has proved a great success in improving the investment advice given to pension funds and in expanding the reach of the actuarial profession. Less has been said about the wider application of this approach and in this article we consider how such models can be extended to provide advice to foundations, endowments, and other charitable institutions on both their investment strategy and on a planned approach to expenditure.

As background, data from the European Foundation Centre show that the charitable sector in Europe consists of some 26,000 foundations with total annual expenditure of over e50bn and combined assets of over e170bn. These institutions finance their charitable projects in part with the proceeds of invested funds and in part using donations from governments, corporations, and private individuals. The relative importance of alternative funding sources varies considerably across different countries and across different institutions within countries.
As encouraged by recent Charity Commission guidelines, standards of professionalism within the charity sector have been increasing, with an emphasis on bringing the skills and techniques used in the management of similarly large sums of money by pension funds and insurance companies into the charitable sector. By strengthening the governance and financial management of charitable organisations the value of every pound donated is increased and charitable goals can be brought one step closer to realisation.
Watson Wyatt has surveyed some 250 European foundations and found a broad consensus that international diversification and diversification between asset classes is desirable (although a definite preference towards investing in domestic or ‘close to home’ assets was observed). When asked about future spending plans, the surveyed charities were less clear and we observed that the link between future investment returns and the ability to spend was not always uppermost in the minds of those managing foundations.
Looking at the overall picture, it is clear that charitable institutions have two conflicting goals as follows:
– maximising charitable expenditure, while avoiding sudden jumps in spending; and
– maintaining the real value of assets, while minimising the risk of financial failure.
Many charities wish to spend evenly over current and future generations. Spending too much now would disadvantage future generations, while spending too little will lead to current generations losing out. Maintaining this balance is the key to effective management of foundations and to achieving the ultimate goal of preserving equity between generations.

Asset expenditure modelling
To investigate this further, we have developed an asset expenditure model (AEM) that looks at both potential investment returns and charitable expenditure in a stochastic framework. An underlying assumption is that the charities we have investigated intend to exist for perpetuity, although we recognise that this may not apply to all charitable institutions, and indeed such a long-term view is not always practised when planning investment and expenditure, even where this is the stated intention.
The concept of stochastic investment returns will be familiar to most, and we will not dwell on it here. For the charity under review, the AEM creates stochastic investment returns, based on the asset allocation reported in the most recent report and accounts and a simplified investment model. These investment returns are combined with the expected income and expenditure of the charity (both in terms of the costs of running the charity and the expected level of charitable expenditure) to project forward a range of possible outcomes in terms of the charitable expenditure and the financial balance sheet of the foundation under review.
There are many approaches to modelling charitable expenditure, three of which are as follows:
– Maintaining current charitable expenditure, with increases slightly above inflation.
– Spending according to the Yale rule, as used by the Yale University endowment (described below).
– Spending according to the ab rule (described below).
The holy grail of spending rules is to find an approach that maintains the real value of assets while allowing a smooth level of expenditure at the highest level possible. Our research to date indicates that current expenditure, when mixed with current investment strategies, is not sustainable for a significant number of charities. Clearly, we would expect expenditure to adjust to a deteriorating financial position rather than be maintained, but figure 1 does illustrate that current expenditure is unlikely to continue.

Spending rules
In summary, a spending rule as described by a formula can be applied and the effects of this spending rule investigated. Two of the rules we have considered are the Yale rule and the ab rule, described below.

Yale rule
An example of the simplest form of a Yale-type rule would be to spend 4% of the average of the last three years’ assets. More typically, the rule may allow for a weighted average of last year’s charitable expenditure and the 4% rule, so smoothing out expenditure. As an example, the charts in figure 2 illustrate the projected funding and expenditure for our sample charity under the following rule:
Spend 40% of last year’s charitable expenditure plus 60% of the average of the last three years’ assets times 4%.
The charts illustrate that the adoption of this rule would reduce expenditure relative to current outgo, and continue to reduce expenditure over the projection period. However, the reduced level of expenditure is still too high to maintain the real value of assets at the median, and the funding level continues to fall.
Further areas to investigate here would be how the spending rule would behave at a lower level of target expenditure, eg at 3% of the average of the last three years’ assets.

ab rule
The ab approach is suggested by Perry Mehrling in his 2004 paper ‘A Robust Spending Rule: The ab Approach’. The key to this approach is to split the total endowment into two distinct funds, although both funds can be invested in the same way. The first fund, denoted E, is the original endowment (or the fund value at a particular point in time), and this is allowed to grow at the rate of inflation. The second fund, denoted F, is the difference between the current endowment and the first fund and is known as the stabilisation fund. By construction the original endowment fund remains constant in real terms over time and all fluctuations in the endowment value are absorbed by the stabilisation fund.
The spending rule can be described as follows:
Spending at time t, denoted by St is given by St= aEt-1+bFt-1
where a is the regular expenditure from the first fund and b is the expenditure from the stabilisation fund.
The level of a should be set below the expected return on assets, allowing the fund to maintain its real value, but, on average, leading to growth in the stabilisation fund. The level of b should be set higher than the expected return allowing the stabilisation fund to disperse the accumulated assets.
In scenarios where investment returns match or exceed expected returns, the stabilisation fund will grow and boost overall spending by the value of b times the stabilisation fund. In scenarios where investment returns fall below expectations, the stabilisation fund will reduce, and in the extreme become negative, leading to a reduction in overall expenditure.
For our purposes, we have run our model with a=4% and b=8% and, as per the Yale rule, applied a weighted average of 40% of last year’s expenditure and 60% of the expenditure according to the ab rule. The results are shown in figure 3 and illustrate that for this charity, the ab rule appears to provide a good fit with a narrow range of likely asset values and a quite stable range of expenditure outcomes.

The future
The scope for expanding traditional asset liability modelling is huge, as demonstrated by our work on asset expenditure modelling for European foundations. We look forward to extending our work and evaluating how best charities can invest while having a stable platform for future spending based on realistic estimates of the money available to meet charitable goals.
Perhaps of particular interest is our initial conclusion that for many European charities, their current levels of expenditure appear unsustainable and perhaps now is the time to review both investment arrangements and how future financial expenditure is planned and predicted.