Picking the right benchmark for charity investment performance
As part of their duty of care, charity trustees need to demonstrate effective monitoring of their investments and benchmarks are frequently used as part of this task. They face an array of different benchmarks to choose from. But what exactly is a benchmark and how are they best used?
Benchmarks are used for three main types of purpose. First, they can be used to set constraints on the types of asset and their proportions held in a portfolio. Second, they can be used to set targets to beat and third, they are often used to provide a comparator of similar strategies or client types.
There are three main types of benchmark: composite indices, target return (also known as absolute return) and peer group benchmarks.
A composite benchmark typically combines a series of market indices in similar proportions to the portfolio’s long term strategic asset allocation – assets held can be mapped/matched against long term liabilities to help determine the long term risk/return objectives.
The portfolio might include a mix of fixed interest, equity, alternative and real estate assets. For example, the UK government fixed interest exposure might be measured against the Markit iBoxx All Gilt index, the UK equity exposure against the MSCI UK Equity index, and so on. In turn these can be used to set constraints on the types of assets held - often minimum and maximum ranges are added within which a manager can act with discretion.
Comparison with subcomponents
This type of benchmark is also useful because it allows performance between the portfolio and the benchmark to be directly compared with its subcomponents. This in turn allows analysis of the differences at each level to be broken into asset allocation, stock picking and timing decisions. This can give important information to both the investment manager and client about the drivers of performance.
A target return benchmark measures portfolio progress against a fixed yardstick. For many charities the yardstick is preservation of the portfolio capital in real terms over the long term. A target return benchmark might therefore be an inflation measure (consumer price inflation - CPI, or retail price inflation - RPI) plus a margin to provide for income generation, costs and a buffer – in all this could be 3%-4% or more depending on risk profile and the inflation measure used.
It is important to remember that this is a longer term benchmark, most often used as a target to beat, and is less appropriate over short periods or in providing comparative data. Although not recommended due to distortions in index construction, composite benchmarks can also be used as targets to beat.
Peer group benchmarks were the first type of benchmark to be popularly used, and arose from the desire to compare portfolio performance with those portfolios invested by other managers. This usually involves looking at the performance of a group of charities or other funds which may or may not share similar objectives and risk profiles to your own. This can be a useful sense check to see how your portfolio is performing compared to others.
However, significantly, peer group benchmarks often do not consider the bespoke objectives of a charity and often group together portfolios with very different asset allocations and investment approaches. Thus focusing too much on relative peer group performance can distract from the ultimate objectives of a charity’s investments. As noted, their main use is as a comparative, but composite benchmarks can also be used for this purpose.
Evolution of benchmarks
It is helpful to note how benchmarks have evolved over the past 30 years, and how they can help meet the needs of the latest generations of investment managers and trustees. As already mentioned, peer group benchmarks were the first type of benchmark to become popular. The problem with the early peer group benchmarks was that only return was considered, irrespective of the risk taken to achieve those returns.
Thus trustees of portfolios underperforming the peer group may have been tempted to shift to riskier, better performing managers in order to achieve these higher returns. This realisation often encouraged fund managers to take on more risk in a race to the top of the table. The result of this was that in the inevitable downturn portfolios proved to be riskier than trustees were comfortable with.
Best practice led to the greater use of composite index benchmarks which looked like an ideal solution. However, trustees began to monitor achievement of their longterm objectives on increasingly short time frames, not over the full investment cycles which most active investment styles need.
Unsophisticated and impatient clients tended to buy managers at the peak of their outperformance and then sell out at their trough, thus damaging investment returns. In turn, investment managers learnt that it was safer in terms of mandate retention to tightly (“closet”) track the composite benchmark. Benchmark tracking error became a substitute for real risk.
With less real active management came less outperformance, opening the door for the rise in passive management. More importantly for this discussion, with the bursting of the dot.com bubble in 2000 and the financial crisis of 2008, trustees tended not to be that impressed when told how good performance had been – with, say, portfolios down by “only” 28%, when the benchmark fell by 30%.
This in turn has led to the greater use of the target return benchmark. However, as we go through the tenth year of the recovery after the financial crisis, there are a lot of investors wondering whether they made the right decision to track RPI + 3% during a period when markets have averaged RPI + 9%. In addition, these types of absolute returns can typically only be achieved in a downturn through the use of derivatives to protect from losses.
The cost of maintaining this type of insurance is usually prohibitively expensive, and so there are likely to be many portfolios that struggle to achieve these targets through the end of the cycle, whenever that may be.
Which benchmark is best?
The main problem with choosing benchmarks is that the moment anything is specifically measured, it tends to distort manager behaviour. This is as true of benchmarks to monitor hospital waiting times as it is of portfolio monitoring. In addition, investment managers and their clients often confuse the uses of different benchmarks and the time horizons over which they are most effective. Accordingly it is best to establish the types of benchmark, their use and the effective time horizon at the start of each mandate.
Each of the three types of investment benchmark discussed capture desirable characteristics that trustees would want to encourage and monitor in a portfolio. However, the focus on just one benchmark invariably leads to other desirable portfolio characteristics being neglected. As there is no single solution charities increasingly use a combination of all three types to reflect the uses to which each is best suited.
This approach helps provide a balance of desirable characteristics being monitored and encourages both the trustee and investment manager to pay close attention to all the characteristics. While this approach does not provide a hard and fast test of success and failure in the short term, it does help ensure that everyone is more focused on what really matters through the whole investment cycle.
How to use benchmarks
Charity trustees need to monitor the performance of their fund manager to make sure performance and risk are in line with expectations. However, expecting performance to match the benchmark at all times and in all situations is unrealistic. In order to beat the benchmark investment managers have to take positions away from the benchmark. Indeed, the question arises as to whether paying too much attention to the benchmark risks putting the cart before the horse.
Performance measurement is relatively easy to do but it is not a panacea. Moreover it can distract from the more detailed monitoring that investors frequently carry out when awarding mandates and then often ignore thereafter.
There is also a need to consider appropriate time horizons. For example, one might ask if measuring something over only three years is a misalignment of time horizons with a charity’s long term objectives. Given potential volatility, it is generally preferable that the longer the period a benchmark covers, the better. With any period of less than 10 years, it is hard to statistically distinguish skill from luck.
Some kinds of investments or strategies will work better than others in a bull or a bear market which means trustees should be looking at the whole investment cycle – peak to peak or trough to trough when assessing an investment manager’s performance. It is surprising how often there are requests to supply historic data for only five years by charity trustees or consultants looking to review investment managers.
A measured approach
The optimal solution is likely to depend on what a charity is looking to measure, as well as its sources of funding, its need for returns and its risk appetite, as well as its financial condition, its size and its scope of operations. Nevertheless, most charities with significant invested assets could probably benefit from a variety of benchmarks to ensure they are getting the most comprehensive view of the performance being delivered.
Ultimately, trustees have an obligation both to the charity’s beneficiaries and to the regulator to ensure that due consideration has been given to the ways in which the charity funds itself, including its investments. Benchmarks represent a useful tool in terms of providing some much needed transparency, but these should be used alongside monitoring of the ongoing investment process, portfolio construction, risk mitigation measures and the investment firm’s culture.
