Accurate and regular portfolio analysis is fast becoming a necessity for both retail and pension funds. But discrimination is needed as an overload of information can be as bad as the wrong information
Performance measurement has come a long way since the days when fund managers hid this function somewhere in the depths of the organisation. To some, it would be 'nice to do if time permits ' once all the accounting reports have been sent out.' In other firms it may have been in operations, middle or front office, or even in the marketing department.
Pension funds were trying to outperform their benchmarks, which were almost exclusively a universe benchmark. Indeed, in many cases, the fund's liability profile meant its asset allocation was inherently different from the average. Despite this, people still benchmarked themselves against 'the herd.' Performance was calculated quarterly, occasionally monthly, with varying degrees of automation, ranging from none at all to little.
Performance analysts did not really exist as we know them today. Their function was mainly one of data processing and the value added was amending accounting data that would otherwise have been sent out incorrectly. Little time could be spared for actually analysing portfolio out-performance ' risk was confined to looking at standard deviation, which not everyone understood and those that did had nothing to compare it against.
UK performance measurement was dominated by CAPS and WM, which suited most parties at the time. Regulation took the form of the National Association of Pension Funds (NAPF) voluntary codes.
The right information
Are you getting enough information? Many people say they are getting too much, but it is the wrong information. We have moved on to more frequent and accurate data. Monthly calculation is the norm with weekly and sometimes daily calculations being provided.
Many argue this is overkill ' and for your average pension fund, this may be true. However, if you are an index tracker, who may have penalty clauses if you stray from the benchmark, or for funds very active in asset allocation decisions, you will want to receive daily performance. Interestingly enough, daily performance is not only for the fund management community, as some of the larger global pension funds will testify.
One of the dangers the performance measurement industry faces is supplying reams of information to clients who do not need or find it relevant.
For example, does your average pension fund want to receive a 150-page performance report every month with the relevant pieces of information buried in pages 140-143? Surely it is far better to sit down with clients and develop reports that suit their individual needs, rather than drown them in paper every month? Why does it even have to be paper? Why not electronic delivery? It gets there quicker, will not give the postman a bad back and is ultimately more useful to the client.
We are now seeing far more people providing performance measurement solutions and there is definitely healthy competition. Far more emphasis is being placed on taking mundane administrative tasks away from the owner/user. If you are taking a dozen different reports from fund managers every month, consolidating them manually into Excel spreadsheets to consolidate into management/board report information, ask yourself: 'Should I be doing this? Is this a good use of my time?'
The stock market performance of the past few years has been a fantastic advertisement for risk analysis and risk management tools. Historical risk analysis measures such as volatility as defined by standard deviation, tracking errors, Sharpe ratios, information ratios and betas have always been popular in the US. Some of these measures are now standard tools for European fund owners, with tracking errors and information ratios becoming popular with UK pension funds. It is also important to understand that one risk measure in isolation will not be very informative. For example, 'the tracking error of my portfolio was 1%.' Is this good or bad? This will depends on:
• Is your fund an index tracker? If yes, 1% is too high.
• If your mandate is very aggressive, with a performance target of +2% a year, then 1% is probably too low
• What is your risk appetite anyway?
• Is this tracking error consistent over time?
• How does this compare with similar funds?
People always assume a high tracking error is a bad thing, but this is not necessarily the case. Active managers have been criticised in the past, quite rightly in my opinion, for being 'closet' index managers. They have hugged the benchmark ' whether the universe or an index ' and taken very few big positions. For an active manager to justify his active fee, they must deviate from the benchmark to add value, thereby increasing volatility and tracking error. If they are not being 'active', why not select an index tracker and save on the fees? Despite this comment, I am a believer in active fund management.
So which one risk measure is best? The answer is none of them. Everyone has their favourites for various reasons, but no one risk measure should be taken in isolation. Rather, use a combination of risk measures that complement your own preferences and your fund's style. Ideally you are attempting to monitor and get comfortable with the risk in your portfolio, while maximising return. People often assume that anything to do with risk will mean reducing risk within their portfolio into safer asset classes (bonds and cash). However, it often highlights the opportunity costs they could be missing by investing too conservatively.
Looking backwards is still an important part of performance measurement. This also really depends if the quants are kept separate from the performance measurement team. Risk measures such as Value at Risk (VaR) have really taken off over the last five years, especially in continental Europe. Although VaR can look backwards, it can also estimate the value at risk in portfolios going forward, either at an individual fund or total plan level. Stress testing is also a powerful tool for seeing how actual or simulated portfolios react under extreme market conditions.
Forward looking (ex-ante) tracking error estimates are also now available. Another area that is gathering interest is risk budgeting, which systematically apportions risk, for instance, tracking error, between asset classes or managers to obtain an overall risk budget for the plan. Portfolio optimisers can be used to assist in this process, to determine the optimum allocation between asset classes.
We need to ensure advances in technology and systems do not mean we are overloaded with irrelevant information.
Daily performance measurement for the majority of funds could become the norm eventually and the fund management community will most likely demand daily calculations. For the majority of ordinary pension funds, this would be information overload ' they are unlikely to be willing to pay for the slight increase in calculation accuracy.
Therefore, this change would have to be driven by increased efficiency of processing accounting and performance information. Clients will receive all the information they need, either online or electronically and not on paper. On this issue, great progress has been made.
Finally, Transaction Cost Analysis (TCA) is a new concept that will increase in popularity, especially after the UK Myners report recommendation that trustees pay closer attention to transaction costs. TCA evaluates the contribution of the trading process, in order to enhance future performance.
One of the dangers the performance measurement industry faces is supplying huge amounts of information to clients who do not need or find it relevant.
A combination of risk measures that complement the manager's own preferences and fund style would be the best solution.
'Right thing to do'
£69m spent on upgrades
European fintech market 'underserved'