There is an apocryphal story of a bursar responsible for the large investment portfolio of one of the oldest Oxbridge colleges who was told by the hot-shot salesman from one of the big multi-asset fund managers: “Your investment portfolio is crazy, it’s all in land and if you had been in equities over the last 40 years you would have done much better.” The bursar indignantly replied: “Listen young man, we have been in property for the last 800 years and we have done fine.”
Of course, the bursar should have added that some of the college’s wealth was also tied up in its medieval silverware, which was brought out for formal dinners at high table, while the dons favoured the liquidity drawn from the college’s wine cellars over any available through bank deposits.
More important now?
An asset allocation consisting exclusively of property and silver with a little art and wine thrown in might be seen as too ‘alternative’ by most institutions in today’s environment but, as the story shows, it can all depend on what your time horizon has been.
Richard Horlick, CIO of Schroders, points out that during the past 20 years, asset allocation has been largely irrelevant. “In 1982 then Fed chairman_Paul Volcker decided to crush inflation and there has been falling inflation ever since,” he says. “Whatever asset class you were in had a tremendous boost from a disinflationary environment. Everything you bought went up except for commodities. Asset allocation was just nibbling around a benchmark when all assets gave similar returns over the period of time.”
This was a once-in-a-lifetime scenario, a one-off boost that will not be repeated. What is clear now is that there has been a secular change that makes the issue of asset allocation and the diversification of risks through incorporating a much wider range of asset classes a central plank of any investment strategy. “In the new environment, you will not get supernormal returns,” Horlick adds. “What assets will give you extra return and diversification in an environment of reasonable economic growth and modest inflation? If equities give 4% real plus inflation, that is not very exciting unless you move away from the index.”
Most pension schemes face the issues of either being underfunded or having sponsoring companies that are anxious to reduce the burden of contributions. The net result is the same – everyone looking for extra returns but reluctant to accept extra risk. This has of course been exacerbated by the herding instinct of pension schemes which made them reluctant to deviate too far from peer group asset allocation weightings in the past, heavily weighted towards equities with little interest in other assets apart from bonds and cash.
Institutions that were prepared to have very different approaches to investment tended to be endowments and family foundations. Yale, for example, has had an asset allocation policy for a number of years heavily weighted towards alternatives, driven by David Swensen, CIO of the Yale endowment fund and a board member of Schroders.
Horlick says of Swensen: “At Yale he got into asset allocation by asking for real returns and diversification that drove him to other real assets and as it became more successful, he asked the question – how do I keep diversifying and keep equity returns?”
The key to the systematic and structured incorporation of alternative asset classes in a well-designed investment solution to a liability profile and return objective is to have an appreciation of four critical areas. First, how sensible are current valuations for the asset class? Second, how stable are the correlations with other asset classes and how could they change during an economic or market crisis? Third, how liquid are the investments should the scheme have to change its investment stance quickly, for example, because of a corporate restructuring. Finally, and most important when it comes to manager selection and fee negotiations, are the returns generated from purely beta, market risks with no manager skill involved, alpha by management skill or some combination of the two.
Valuations of asset classes can be notoriously difficult, as the TMT bubble of the 1990s, the Japanese bubble of the 1980s and any number of historical bubbles has shown. Although there is strong evidence that asset prices over long periods are mean-reverting, the problem is that over the time scales that fund managers and trustees are judged, deviations from long-term averages can be large and persistent. Diversification through investing in as many ‘cheap’ asset classes as possible reduces the exposure to any one asset class.
Hedge funds, private equity and property are seen as the three largest alternative asset classes, but other possibilities include physical commodities, currencies, specialist sub-sectors of equity and bond markets such as emerging market debt, and natural resources such as forests. Correlations between these can vary dramatically. Private equity valuations are highly correlated with stock market levels and as the TMT crash showed, can be heavily geared towards systematic market revaluations. Extreme market movements in the past produced flights to perceived havens such as gold and US treasury bonds. In more normal times, correlations can be very different to those during extreme crisis and can even change sign.
Liquidity is an issue for some classes of assets, particularly private equity investments where capital is committed for up to 10 years. This has been alleviated somewhat through the development of the secondary market for private equity with specialised firms such as Coller Capital focusing purely on this sector.
The distinction between beta and alpha returns is critically important when it comes to deciding what level of fees is justifiable. Exposure to returns from owning physical commodities can be made, for example, through rolling over futures in indices such as the GCSI representing a broad basket of commodities. This would represent predominantly a beta risk, with manager skill coming in at the margins in the timing of rollovers to take advantage of any backwardations in the marketplace, when futures prices are lower than current spot prices.
In contrast, hedge funds conceptually at least, are meant to represent pure alpha risk. Derek Doupe of Schroders argues: “Hedge funds succeed for structural reasons. Most market participants are not profit maximisers, eg in the UK equity market a large part is held by pension funds and insurance companies, the former maximise profit relative to a tracking error target and insurance companies are constrained by their liabilities. Hedge funds have complete freedom of choice. It’s Game Theory, everyone can win because they have different objectives.”
While this is a pleasant thought, the reality is that many hedge funds will fail for operational reasons as well as the fact that they cannot maintain the ability to generate alpha. There may be over 7,000 hedge funds and 1,500 fund of hedge funds but as Ian Martin, managing director of GAIM Advisors, complains: “The dispersion of returns is going down and the beta of multi-strategy funds is going up. Now that institutional players are looking at the market, there is much more forensic analysis done by their advisors The challenge is to reward people for alpha and not beta. Lots of mediocre managers are getting rewarded for beta, for market returns. The amount of skill in hedge funds over the last few years has been disappointingly low.”
Advice and fees
For the trustee, there are two questions to consider: who is best fitted to give the advice on asset allocation incorporating a wide range of alternative asset classes and what are reasonable fee levels for management of them? As Horlick notes asset allocation: “At the end of the day, it is the trustees’ responsibility. However, it only makes sense when looked at in the context of the liabilities and the assumptions behind those, which are determined by the actuaries. You would have assumed that investment managers are best positioned but for so many their involvement has been in relative returns in a narrow benchmark area. What most firms are trying to do is to rebuild the skill sets of the 1960s and 1970s that went into abeyance in the 1980s and 1990s. We are going full circle.”
As for fees, as Ian Martin asks: “Why pay 1.5% to generate beta returns when you can get an index fund to do that for a few basis points? Perhaps performance statistics should be broken down between alpha and beta components and fees should be based just on alpha returns.”
A view that may attract more
support from schemes than fund managers.