Benchmarks are often used to define and control investment strategies. However, there are so- called ‘absolute return’ investors aiming to achieve constant positive returns unconnected with market movements. Their ‘benchmark’ is seen as a fixed return (7%, for example). Many of these investors consider benchmarks that consist of bond or equity indices as useless for defining their asset strategy. This is because they do not want their assets to fall with the benchmark – that is, to fall with bond or equity markets. The question is, is there still a role for a benchmark with these absolute return investors?
In many continental European countries, such as Italy, Spain, Germany and Austria, institutional investors focus on absolute return targets. For example, insurance companies and Pensionskassen in Germany have achieved almost constant returns. In the past 10 years, they have distributed yearly returns (Nettoverzinsung) of between 6.8% and 7.6% on average. This has been possible, because they were able to level out volatility by using hidden book reserves (Stille Reserven) as a fluctuation cushion. Thanks to a regulated market, herd behaviour and local accounting rules, which focused on book values, it was possible to build these hidden book reserves in times of above-target returns and use them to bolster below-target returns in other years (Figure 1).
Since the deregulation of the insurance market in the European Union in 1994, greater competition has emerged and new entrants to local markets are more likely to distribute profits to attract new business, rather than build additional reserves for rainy days. In Germany last year, some local insurance companies were forced to reduce their distributed returns from their long-term target of above 7%. This was due to a lack of hidden book reserves that could otherwise have subsidised the lower market return.
This year, more insurance companies in Germany will have to reduce their distributed returns below the psychological level of 7%. In addition, lower levels of inflation have led to lower nominal yields and a fixed number, such as seven, is no longer as meaningful as it once was. All this suggests that in future the level of distributed returns from insurance companies may not be as high or as stable as we have seen in the past. As a result of legislation, 90% of these reserves belong to the beneficiaries, yet it is not clear how they should be distributed among the beneficiaries. Currently, reserves built from long-standing customers are used to subsidise new business.
A risk-free or guaranteed return can be determined by the current cash or zero bond yield, depending on time horizon and the definition of ‘risk’. Additional returns above this level can only be achieved by taking risks. This could be duration risk, credit risk, currency risk, equity market risk and other active risk (due to other active decisions such as stock selection). The returns of all investors can therefore be seen as a combination of a risk-free rate and a risk premium (as in modern portfolio theory). For example, the return of an equity portfolio r(p) can be split into a risk-free rate r(f) and a risk premium {r(p) – r(f)}:
r(p) = r(f) + {r(p) – r(f)}
The size, volatility and relationship between the risk premiums available from different asset types should be the driving force in the investment decision to achieve the target return. The ideal combination is a risk premium with low volatility, high expected return (alpha) and low correlation to the other constituents of the overall portfolio.
The twin objectives of achieving absolute returns and withstanding falling markets can be better illustrated by splitting a traditional equity portfolio (for example, European equities) into two parts, comprising its indexed and active positions, as in Figure 2.
This shows us two things. First, it is now obvious that the traditional actively managed portfolio contains index positions, which provide market returns, and active positions, which provide an absolute return. Secondly, it shows how reducing or removing the part with index positions can lessen the market exposure or market risk. The active positions – the deviations from the market positions – can be replicated in a long-short portfolio that reflects the view of the active portfolio manager relative to the market or relative to the consensus view.
“Market neutral: the only way to invest, if you believe in active management” – This famous quote from William Sharpe is based on the theory that one should hire active managers for their active views, not for the consensus view (ie, the index positions). The consensus view is reflected in the index positions, because all investors in aggregate hold the market and thus on average achieve the index return.
Besides buying the companies that they like and excluding or underweighting the stocks they dislike, managers can also short sell the companies they dislike. This can be achieved in a long-short portfolio by replicating deviations from the consensus. The return r(l/s) of such a long-short strategy consists of the returns of the long positions r(l) and short positions –r(s):
r(l/s) = r(f) + {r(l) – r(s)}
To construct a long-short portfolio neutral to market movements (ie, market risk), it is necessary to have rigorous risk controls that check projected and actual returns.
Such a portfolio is often referred to as a market neutral portfolio because its returns are uncorrelated with market movements and therefore do not succumb to falling markets. The risk premium of a market neutral fund can be about 4%, somewhat higher than the expected risk premium of a traditional long-only equity fund, which can not deliver any alpha from short selling. Furthermore, the volatility of a long-short portfolio may only be one third of the volatility (risk) of a traditional equity fund.
In practice, we do not see many long-short strategies. There could be several reasons for this:
q implementation may be difficult;
q short selling may not be permitted;
q the strategy may be difficult to understand; or
q asset managers may want to reduce their business risk by holding index positions (closet indexing).
For absolute return investors, however, market neutral strategies do have their appeal – low risk, low correlation, additional expected returns and strategic diversity. To achieve absolute returns these strategies can be combined with cash or bond investments. This would also appeal to investors who have used up their maximum equity exposure under local legislation; German Pensionskassen and other insurance companies, for example, are allowed a maximum equity exposure of 30%.
The example in Figure 2 has illustrated how a traditional portfolio can be represented as a blend of two poles, pure passive and pure active asset management. All strategies can be seen between these two poles, which span a spectrum between passive and active management.
A systematic way to actively achieve an absolute return, without succumbing to the risk of falling markets, is to combine a pure active market neutral strategy (long-short) with a bond, cash or other risk-free strategy. Thus, the absolute return can be transported to a separate portfolio, which is also known as portable alpha. A pure market neutral strategy aiming for a 4% portable alpha, combined with a bond or cash portfolio delivering 3.5% return, can achieve a target absolute return of about 7.5% without equity market exposure. The short positions in the long-short equity portfolio hedge against falling equity markets.
The example in Figure 2 above, in which the benchmark defines the market to split a portfolio into two components, illustrates nicely the difference between pure active risk and market risk. In this respect, relative to benchmark performance measurement is risk control that describes how much the portfolio is connected to the market (for example, by correlation). In falling markets, it helps to determine how much market risk the asset manager could avoid. For those who do not like the concept of tracking an index, this demonstrates clearly the advantage of looking at a benchmark.
This means that for the absolute return investor, there certainly is a role for a benchmark, and that role is twofold. First, it provides an element of risk control, and secondly it helps to establish market neutral strategies, which target absolute returns that are uncorrelated to equity markets. Market neutral funds offer investment managers the potential to reduce market exposure and to add value via uncorrelated non-consensus decisions.
Olaf W John is a director, European institutional business at Barclays Global Investors in London .
No comments yet