Performance attribution and cost control are two areas which can help investors to decide how indexing best fits into their portfolio, explains Olaf John at Barclays Global Investors
Since the 1950s, when Professor Harry Markowitz first conceived the idea of investing in whole markets, indexation has engendered a vast amount of research and debate . By the early 1970s indexation had moved from theory to practice with the launch of the first index fund by Wells Fargo Nikko Asset Management, now part of Barclays Global Investors. Index mandates have grown enormously as the advantages of indexation have become more widely understood, yet there are still entire markets where the concept of indexation is hardly used. Why should this be?
It is quite common to do comparisons among different countries, but it is even more important to understand why things are different . It is interesting to observe that there is a correlation between the degree of influence held by investment consultants in a particular market and the extent to which indexation is used in that market. Fig 1. gives an overview of the degree to which indexation has penetrated some of the world’s markets. If we combine this with what we know from a variety of studies and interviews with investors, it appears probable that the influence of consultants has led to greater transparency of fees, performance and manager skill, prompting institutional investors to index portions of their assets.
In the UK and the US, greater transparency of fees and skill, not to mention discipline in the process of appointing asset managers, has come through the use of performance measurement. In part, this is because the benchmark concept tends to split investment decisions into smaller parcels relating to, for example, equities, bonds, countries, currencies, industries and individual stocks. The result is that skill in each area and the fees paid for those skills can no longer be masked by overall returns. By contrast, in countries such as Germany and Austria the main focus is still on absolute return objectives, with performance relative to a benchmark and scrutiny of skills in individual asset classes taking a back seat.
In the example in Fig 2, Manager A has an absolute return of 7.67%. It has outperformed Manager B by 0.5% and the benchmark by 1%. Performance attribution shows that it achieved this mainly through its choice of asset allocation between equities and bonds. Note, however, that it failed to beat the index in the underlying asset classes.
While some people might look at absolute return and not care about the source of performance, others might prefer to employ Manager A only for asset allocation decisions and then index the underlying asset classes, a combination that could bring higher performance and lower costs.
However, performance is affected by many other decisions, whether implicit or explicit, that asset managers make concerning regions, currencies, industries and other factors. Performance analysis can be used to examine these decisions in a variety of ways and over different time periods. For active asset managers this analysis is invaluable in understanding which decisions are compensated and which are not. This enables them to allocate returns and understanding risk. Risk is often measured as tracking error, ie the deviation from a benchmark. 1
More importantly, performance attribution is a tool for ascertaining whether the level of return achieved was justifiable in terms of the level of risk taken, or whether there might have been a more efficient, less risky way of arriving at the same conclusion.
We do not know the risks Manager A took to achieve its results, but we do know that these risks were mainly associated with equities. From this we can infer that Manager A took a higher level of total risk than either the benchmark or Manager B.
The aim of good performance analysis is therefore twofold - to distinguish skilled from unskilled investment managers and to bring greater control to investment decisions. As the example in Fig 2 shows, cross-sectional comparisons of returns can distinguish good from bad investment decisions. Time series analysis of the returns can help to separate skill from luck by measuring return and risk. This checks that the manager has indeed delivered the performance and risk control that he promised.
The transparency of investment management fees varies considerably between markets. The US and the UK are probably the most transparent, while Germany and Austria lie towards the other side of the spectrum . When one considers overall fees - investment management fees, transaction fees, portfolio turnover impact and custodial charges - the allure of indexation grows. Lower turnover rates, the crossing of transactions internally rather than trading in stock markets and the absence of expensive research visits all ensure that passive management is far less expensive than traditional active management.
By taking into ac-count both fee considerations and the risk/return trade-off, one can get a much better picture of where to use active and where to use passive management . For example, a prime candidate for indexation is the Euro-zone bond market. A Euro-zone index such as the Salomon Euro BIG Index now poses a severe challenge to active asset managers attempting to beat it, and the levels of fees typical of traditional active management will erode the client’s returns still further. In the Euro-zone the old mainstay strategies of active Euro-zone bond management, based on betting on currencies and interest rates, have been rendered near useless by the advent of the euro. Current low interest rates in the Euro-zone also make it much harder for active managers to outperform, leading to greater competition for added returns through stock-specific plays on a small corporate bond market and use of duration bets. With this limited opportunity set indexation provides a less risky more cost-effective approach to the investors.
The evidence is increasing that the growth of indexation has been spurred by the role of investment consultants. This certainly appears to be the case if we compare Austria and Germany, where investment consultants still play a minor role, and markets such as the US, the UK, The Netherlands and Switzerland where they have gained significant influence. Consultants have supported analytical disciplines, which has lead to higher fee transparency. They have introduced performance-related fee structures , raised the level of the active versus passive debate and increased the level of competition in the industry. In this growing culture of accountability in Germany and Austria, indexation will come to the fore as an efficient investment option for many investors.
1. Tracking error is typically defined as the standard deviation of the last 60 observations (eg monthly or weekly) of a, where a is the difference of portfolio return rP and benchmark return rB, annualised, ie with n=60 and m=12 for monthly observations
Olaf John is director of European institutional business at Barclays Global Investors in London