It’s been about two years since “hedge fund replication” hit the mainstream business press. Over that time, many products have been launched and there has been much navel-gazing about the future of the hedge fund industry. But today “real” hedge funds remain. With such an elegent value proposition, and a stellar academic pedigree, why have the hedge fund clones not yet taken over the world?
This is one of the questions being addressed at a Terrapinn conference next week in New York. To provide a bit of perspective on the challenges faced by hedge fund replication, we invite one of the speakers at that event, UK-based Nicola Ralston of Liability Solutions, to share her thoughts. Ralston is also a former Chair of the CFA Society of the UK and a Governor of the CFA Institute.
Special to AllAboutAlpha.com by: Nicola Ralston, PiRho Consulting
Asset owners are an understandably sceptical group of people. They often feel it’s a jungle out there; investment banks and asset managers seem to make a very good living at their expense, yet everyone knows that real skill is a scarce commodity. Where to turn? In the equity markets, passive investing cuts the cost of gaining exposure to markets. It reduces the risk of being fleeced by managers who are trying, but failing, to add value and taking a big fee along the way. Why should hedge funds be any different?
Working out what to replicate
The concept of hedge fund replication assumes that we have an understanding of what we seek to replicate. This may seem fairly straightforward – even basic – but there are serious issues to consider. Even indices that broadly measure the same thing can differ significantly.
Over the five years 2002-2007, for example, dollar returns on underlying hedge fund indices vary between 7.2% pa and 12.6% pa, though Fund of Hedge Fund indices vary less, from 7.1% pa to 8.2% pa. Investable indices have performed consistently poorly, ranging between 4.4% pa and 6.0% pa over the same five year period. Criteria for inclusion, however, are extremely limiting, and hence such indices represent a very small sample of the universe.
It is, therefore, possible to argue that “hedge fund returns” have varied between 4.4% pa and 12.6% pa over the same period. When it suits presenters to show their own performance in a favourable light, they often show comparisons with investable indices on the basis that it is the performance “real” funds have achieved. But funds of hedge funds would be given short shrift for using such an argument. Similarly, when presenters want to make the rationale for investing in hedge funds as an asset class, they generally show a fund-based index such as the HFR Composite, which is consistently several percentage points higher. It is hardly surprising that investors are confused.
What do replication products aim to achieve?
Deciding which index to track is only the start of the decision-making process. The stated purpose of hedge fund replication products tends to vary widely. Some, such as Morgan Stanley’s Altera platform, have targets related to the HFRI Hedge Fund Indices. Others, such as Goldman Sachs’ Absolute Return Tracker, do not publicly target any specific index. In Goldman’s word’s, its product “reflects the total return of a dynamic basket of investable market factors determined through an algorithm to approximate patterns of returns of hedge funds as a broad asset class”.
The use of an unstated proprietary benchmark is also common among the smaller asset management hedge fund replication type products as well.
Is replication really all about alpha?
Proponents of traditional passive investing frequently make the point that the average fund cannot outperform the index, since by definition the sum of all equity alpha is zero (before costs). They point out that on average, passive equity products perform better than active equity products due to their lower costs. There is an (often unstated) assumption that the same logic will hold true for hedge funds.
But is this a helpful way of thinking about hedge fund investment? The stated objective of most hedge funds is not to “create alpha”, but to generate positive (absolute) returns. The creation of alpha is merely a means to the end of generating positive uncorrelated return. The return on the HFR Fund of Hedge Funds Index was 7.5% pa over the period January 1994 – May 2008, nearly 2% pa above the 10 year bond return, but with a volatility of 5.8% pa – well below the bond volatility of 7.0% pa. That index has also outperformed equity indices handily. So it can be argued that hedge funds as a whole have done a valuable job for investors.
So how concerned should we be that only a modest component of hedge fund returns is ‘pure’ alpha? As a result of extensive academic work on hedge funds, we now understand much more about the complexity and diversity of their sources of return. For example, we know that they contain a significant element of mainstream equity beta. Yet hedge funds have also been successful at reducing exposure to equity markets on the downside. Is this skill in limiting exposure to equity beta at the right time ‘alpha’?
Furthermore, we now know that hedge funds contain many other ‘alternative’ or ‘exotic’ beta exposures. They contain not just straightforward betas like credit spreads or commodities, but also others such as volatility and liquidity – and even more esoteric ones such as lending without being subject to banking regulations.
It is both difficult and expensive to generate alpha in the complete absence of beta. This is another way of saying that any almost every investment has some embedded exposure to one or more market factors along with security-specific risks. Low exposure to equity beta, which is generally considered a desirable trait in hedge funds, often entails substantial exposure to non-equity beta factors such as asset backed lending.
To some extent that debate around the generation of alpha revolves around a tautology. Alpha can be defined as everything which is left once systematic return sources have been removed (Ed: the Victorian “God of the Gap” in last week’s Economist – see related post). On this basis alpha can only be known retrospectively, and almost certainly not by those who are trying to generate it minute by minute.
Fees, half truths and statistics
It might be argued that debates about alpha, and benchmarks are interesting up to a point, but most hedge fund replication investors only need to be concerned that they do a good job of delivering a substantial element of the attributes of hedge fund returns at low cost.
Fees for hedge fund replication products vary enormously but are generally at least ten times as high as the fees for passive equities. This reflects, in part, the vastly greater complexity of the hedge fund products vs. their traditional counterparts. Some also charge performance fees.
Perhaps more surprising, however, is that determining the actual fees is not always a straightforward matter. Many replication performance statistics are quoted gross of fees, while the hedge fund indices they are compared with are universally net of fees. This practice should always be challenged by investors. The case for replication products should not depend on unfair or irrelevant performance comparisons – particularly when greater transparency is often described as one of the advantages of this approach.
The jury is very much out on the performance generated by hedge fund replication products so far. The divergence of returns between different replication products is extremely large. In the seven months to July 2008, for example, performance on the replication products we monitor varied between -15.7% and +3.8%. It is also interesting to compare the spread of returns on hedge fund replication products with equity trackers, where a range of the order of 0.6% might be seen. On this basis, the range of returns on hedge fund replication products is more than thirty times wider than on passive equity products.
In fairness, hedge fund replication products bear little relation to the straightforward world of passive equity products. Equity index products are almost universally based on widely used, capitalisation-based indices. In that business, investors decide which index they wish to match, and have a realistic expectation of ending up with a performance within a few basis points of that index, for a fee in single digit basis points.
It’s no surprise that investors in hedge fund replication products would like to believe that they are getting something similar. In reality, however, they are not. The point is not whether individual replication products can do a good job relative to their chosen objectives (some clearly have). Rather, it is that investors should not invest blindly in a replication product and assume that they will broadly match the return of the hedge fund industry. By extension, they also can’t assume that any fee savings will necessarily translate into superior after-fee returns.
– N. Ralston, September 2008
The opinions expressed in this guest posting are those of the author and not necessarily those of AllAboutAlpha.com.