Charities have the freedom to be the most interesting investors around, but are hesitant to break the bounds of tradition, says a new report from Sarasin & Partners, which manages money for 175 of them. The report, Sarasin & Partners Compendium of Investment for Charities , looks at current trends in endowment investing - globalisation and alternative asset classes are key terms here - and describes the ideal process for setting investment policy.

"While trustees and investors are perhaps more able than ever to create portfolios that match their particular circumstances, agreeing and implementing investment policy still represents a major challenge," says the foreword to the report, written by Richard Maitland, Edward Campbell-Johnston and Henry Boucher.

Guy Monson, Sarasin's chief investment officer, points out that the UK's Trustee Act 2000 says: "A trustee may make any kind of investment that he could make if he were absolutely entitled to the assets of the trust", with the exception of direct purchase of land. Once the statutory duty of care is fulfilled, which Mr Monson interprets to mean diversification, and a risk process should be in place, "you can do anything you like".

Charities have traditionally been the poor relation of investing in the UK, says Mr Monson. They have not been adventurous in their investment policies, even though in the US endowment funds such as those of Harvard and Yale Universities have been exemplars of innovative and successful investment. "But things are changing here very, very fast," says Mr Monson. The changes highlighted in the report include the increased use of pooled vehicles, a move away from measuring performance against peers to using bespoke benchmarks, a tendency to aim for returns higher than cash or higher than inflation instead of beating indices, and increasing willingness to allow derivatives to be used in investment management.

While these techniques have become mainstream in pension fund investing, they still count as new for charities. Charity investing has different requirements and objectives than pension funds, however. While each charity has different objectives, as a group they are likely to have an investment horizon at perpetuity and need to generate a regular income stream, which is not dependent on variables such as longevity.

Endowments therefore can approach the task of setting their investment policy on different terms than other investors. For a start, their investment horizon is likely to be much longer than most investors. On its own, this can make a difference to the basic assumptions used in setting investment policy, especially when it comes to asset allocation, deciding what proportions of the fund to put into equities, bonds and other asset classes.

The classic analysis looks at various combinations of two asset classes, usually equities and bonds, to find the proportions that offer the highest returns for an acceptable level of risk. This produces a curved line called the efficient frontier, the furthest right on the graph below. It shows, for the UK market, that although holding the whole portfolio in government bonds is slightly riskier than including a small allocation to equities, beyond that, more equities means more risk.

The limitation of this useful exercise is that it only applies over a 12-month period. If the same calculation is done over a longer period, it turns out that bonds in the long run are more volatile than equities.

If UK charities take on board this kind of investment thinking and combine it with increasing the range of asset classes they use, as is already happening, they could find themselves at the cutting edge of investment innovation, like their US counterparts.