Complexity adding value?
In constructing a more efficient portfolio investors are typically looking to add assets (or strategies) that are uncorrelated with the current portfolio and have a positive expected return.
One of the assets/strategies that has been considered by Australian superannuation funds is that of hedge funds. The purpose of this paper is to examine to what extent the addition of hedge funds to a typical superannuation portfolio has been useful in improving the overall risk and return characteristics of the portfolio.
The proponents of hedge funds in portfolios have typically argued that the addition of hedge funds offers one or a combination of benefits that I would categorise as “diversification”, “alpha” or “flexibility”. Amongst others, these benefits would include:
- they represent a different “asset class” or strategy and access returns typical strategies cannot;
- they have a low correlation with equities/balanced portfolios;
- objective based or outcome orientation rather than benchmark orientation; and
- the ability to change strategies readily as market conditions dictate.
Investors in hedge funds (or indeed any asset class/strategy) are looking for positive expected returns (presumably at least above cash) and diversification (low to zero correlation with traditional assets). In the long term, investors are looking to make money and earn a positive return even if the stock or bond markets do not do well. In the short term, particularly in the case of short-lived cataclysmic market events, investors are looking to preserve capital.
It is our view that while it may be possible to identify individual hedge fund strategies or fund of funds which “add value”, the complexity of such structures, significant fee levels involved and the thousands of strategies available means the odds of getting it right are firmly stacked against regular institutional investors. It is our view that most superannuation funds would be better served by keeping things simple and staying away from hedge funds.
In the remainder of this paper we review the extent to which hedge funds have been able to deliver (or not) their purported benefits.
What are they?
Hedge funds have often been characterised as an additional “asset class”. In more recent years this has been as a subset of either the defensive component of the portfolio or a broader “alternatives” allocation. Part of this is due to the fact that most hedge funds offer little transparency and as such there is the mystique about exactly what they are doing. Several studies have investigated the underlying “beta” of hedge funds to determine to what extent their returns are due to accessing “non-traditional beta” including trend following (momentum) and other derivative based factors. Ultimately, we would suggest that in the context of accessing capital markets there is equity and debt, which can be accessed via listed markets or directly (unlisted), most of the analysed “non-traditional” betas really represent factor exposures to an active strategy.
We would conclude that hedge funds in aggregate are not an asset class. Certain hedge funds may invest in a sub-asset category (e.g. distressed debt) and for these strategies they should rightly be considered as a sub-grouping within the respective asset class. However, in aggregate hedge funds represent a myriad of different strategies built around one or a number of asset classes, not an asset class in their own right. The real source of expected value investors are accessing is in our view manager skill, not some new beta.
While this may sound at first a relatively subtle point, the implications are somewhat greater:
- Strategies (as a general rule) cannot be “valued” in the traditional sense. A strategy is an approach to investing, it is not an investment in itself – important concepts in deriving future returns such as “acquisition cost” and “cashflow” are not meaningful for a strategy.
- Most hedge fund strategies are relatively dynamic – meaning that the historical performance track record has dubious predictive power (to the extent that any historical track record has meaningful predictive power).
- The underlying process, strategy or the people who implement that strategy are subject to change (sometimes significantly).
- The strategy is subject to being arbitraged away as the volume of money chasing that strategy increases – a major issue for hedge funds which have seen phenomenal growth in assets (and the number of offerings).
- The costs (and method) by which the strategy is accessed are a large part of the total return.
In summary, the expected return of a collection of hedge funds strategies (much like any active strategy) should realistically be zero, or negative to the extent of the fees (which in this case are not insignificant). In our view the above are not reasons to disregard hedge fund investment outright – as in fact much of the above can be said in relation to any active management strategy – however they are reasons to a) be cautious and b) help define how we should look at hedge funds from an investment analysis perspective.
In regard to the analysis of hedge fund returns, given that they represent an amalgam of strategies not an asset class, the objective of our analysis is to review:
- the total returns from these strategies versus broader portfolios;
- to what extent these returns correlate with the broader portfolio; and
- is accessing the above actually worth it, post fees, in the context of the impact on the overall portfolio.
While it is not the objective of this paper to conduct an exhaustive study into 1 and 2 - there are and will continue to be plenty of academic studies that attempt this - we do review some of that academic evidence and conduct our own analysis from an Australian dollar perspective. Using this analysis and that conducted by others we can then draw some conclusions around point 3.
One critical point to be aware of in any analysis is that when we are comparing hedge fund returns with other asset classes (or groups of asset classes) we are really comparing an active portfolio with a passive portfolio. To the extent that these portfolios have generated positive returns this could bias results to the inclusion of hedge funds which might be less obvious if active strategies had been used as the basis for comparison.
Given that hedge funds are not an “asset class” per se but rather a strategy, the sourcing of return information is somewhat more difficult – clearly there is no aggregated list of assets with defined prices and market values around which to construct an index of returns in the normal way. Additionally, there is no uniformity to the definition of hedge funds nor a specific set of assets which hedge funds invest in and as such, defining the performance of hedge funds as a universe is quite difficult.
By various estimation there are circa 10,000 hedge funds globally, many of which would be either very small and some which are huge and/or closed to new investors. For the purposes of this analysis we have used the Dow Jones Hedge Fund Index to represent the performance in aggregate of hedge fund strategies. To be included in this index funds must have a minimum of one year track record, audited financial statements and assets of at least US$50mn.
We would note as shown in the chart below that the construct of strategies within this index has changed significantly over the last 15 years largely representing the change in structure of the industry. This will no doubt have had an impact on the historical and future performance characteristics of the index.
Historical Sector Weights – Dow Jones Hedge Fund Index
While other indices exist, including strategy specific indices, hedge fund of fund indices and “investible” indices, for the purpose of this paper we have mostly considered the index which represents the broadest set of available strategies.
Performance of hedge funds
In deciding whether to add any asset or strategy into a portfolio, the obvious first question is “does it add incremental return”? In the case of a traditional asset, this is will be primarily a function of the cost of acquisition and the future cash-flow generation of the asset. However, in the case of hedge funds the underlying assets are typically (albeit not always) listed securities. As such, it does not make sense to think about hedge funds as an “asset” or collection of assets, but rather as a strategy for investment in a group of assets.
A recent study by Ibbotson, Chen and Zhu1 decomposed hedge fund returns from 1995-2009.That study highlighted a number of important findings:
- Hedge fund indices are subject to a large element of survivorship bias (the removal of failed funds from an index – either just before they fail or the entire historical series after they fail) and backfill bias (the adding of return data for a successful fund after it has been successful). These biases are not at all insignificant – and across a number of studies including the Ibbotson one have been estimated at circa 3% p.a. for survivorship bias and up to 4% p.a. for backfill bias – clearly quite significant.
- Allowing for some of these biases (but not all), the total pre-fee return of hedge funds over that period was 11.13% p.a. (in USD). This was made up of “beta” 4.70%, “alpha” 3.00% and fees 3.43%. Bearing in mind our comments above that hedge funds are dynamic strategies we would suggest caution in analysing the split between alpha and beta. Nonetheless, the Ibbotson study showed a breakdown of the sources of return by strategy as shown below.
Sources of Return by Strategy, 1995-2009 (gross of fees, annualised)
The conclusion from the Ibbotson paper is that hedge funds have added value over the time period. However there are a number of further points to make in this regard:
- This is from the perspective of a USD based investor - albeit while one could hedge the returns back into AUD, if the underlying assets are not USD based nor managed to a USD perspective then there is an implicit currency risk being taken. For example, a USD denominated hedge fund that invests in Australian bonds, if hedged back to AUD would be effectively leveraged into the Australian dollar and doubling up on risk.
- The time period represents the period 1995-2009 – previous studies have highlighted the positive returns from hedge funds during the 1990’s but post 2000 the results are more mixed.
- The added value is subject to the assumed “beta” being a) a beta that fits within the overall portfolio and b) a real beta.
- It is worth noting that the fee drag exceeds the alpha produced – i.e. managers are effectively capturing over 50% of the total alpha in fees.
- Most of the return analysis above assumes equal weighting of the underlying strategies not asset weighting. If we were to do the same with say equity index analysis this would add a number of percentage points to the long term returns from equities – however as with hedge funds, implementing such a strategy would be rather difficult.
In examining the return results for hedge funds from the perspective of an Australian investor we have considered:
- the total returns over various time horizons of hedge funds versus broader Australian portfolios and more traditional asset classes2; and
- the change in these returns through time – particularly for the hedge fund universe which has undergone substantial expansion in assets and number of funds.
One of the interesting characteristics we observe about hedge funds has been the substantial change in returns over the last 2 decades. Pre 2000, a period when there were relatively few funds and limited assets (and a universe dominated by global macro), returns where significantly higher than the post 2000 period. This can be observed more readily in the chart below which shows rolling 5 year returns from hedge funds versus a passive balanced3 and capital stable fund4.
We can see quite clearly the deterioration in hedge fund returns as the pre-2000 performance period works its way out of the rolling 5 year data. Over the last 10 years or so total hedge fund returns have been quite poor – in fact generally below that of a passive balanced or capital stable fund from an Australian dollar perspective.
The chart below shows the annualised performance of hedge funds from an AUD perspective (currency translated, not hedged) versus various other asset categories (passive) and a passive benchmark balanced and capital stable portfolio over different time frames to 31 December 2010.
With the exception of global equities and Australian equities over the last 3 years, hedge funds have underperformed more traditional asset categories and more diversified passive portfolios over periods out to 10 years. Additionally given that hedge funds represent “strategies” not actual assets a period of underperformance does not imply a move in relative valuation and the potential for subsequent outperformance (ie. there is no economic reason why mean reversion would occur).
Given the impact of the global financial crisis more broadly on asset class returns, we have also examined the performance of hedge funds versus other investments over the period up to 30 June 2007. This is shown below.
Once again, the themes are relatively similar. Most asset classes had done quite well in aggregate and with the exception of the 10 year returns - which is taking us back into the late 90’s stronger performance period - overall returns for hedge funds have been less than that of other asset classes or more diversified portfolios.
In addition to analysing purely the performance track record we have also considered risk adjusted returns (humour us for a moment that risk equals volatility). The following chart shows the risk and return profile of the asset classes above over the 5 years to 31 December 2010.
As shown in the chart, on a risk-adjusted basis hedge funds have not represented a particularly beneficial allocation in a portfolio (albeit neither have unhedged global equities). Over the 10 year period the results were quite similar.
Diversification and Downside Risk
One of the oft-quoted advantages of hedge funds has been their ability to capture upside and limit downside. This has seen the proponents of hedge funds pointing to their diversification benefits at an aggregate portfolio level.
Diversification benefits can come from 2 main sources:
- tstments (by which we mean the forces that will impact future cash-flows and the pricing of those cash-flows in the market are different to those that will impact other investments in the portfolio); or
- the active management of the strategy tilts or changes the portfolio sufficiently at appropriate points in time to give very different return characteristics of the hedge fund portfolio relative to the total portfolio.
Given the myriad of different strategies and assets used by hedge funds, it is difficult to expect that in aggregate they would satisfy the diversification benefit described in point 1 above – i.e. there are no singular characteristics of the assets that hedge funds own that will make them “diversifying” vis a vis equities or bonds. More likely hedge funds represent a collection of strategies across asset classes that gives the characteristics of a more diversified portfolio (conservative or aggressive) – but not a diversifying portfolio (i.e. a “balanced” fund could be thought of as a diversified portfolio, but not a diversifying portfolio – adding a balanced fund to a balanced portfolio won’t necessarily make it more diversified).
However, the actions of the managers of the strategy could in effect create a portfolio that behaves differently to other portfolios by way of either a different methodology for implementing decisions or outright skill in judging when to hold which assets.
In our view, hedge funds could provide a diversification benefit, but it will be a function of the individual funds specific strategy and the skill of the operators of that strategy.
Nonetheless it is commonly reported that hedge funds aim to capture market upside but limit downside.
This apparent skewness in the return distribution was analysed by (among others) Malkeil and Saha5. Interestingly the results of their analysis covering the period 1995 to 2003:
“…confirms that hedge fund returns are characterized by undesirably high kurtosis and that many hedge fund categories have considerable negative skewness.”
Effectively, Malkeil and Saha refute the assertion that hedge funds capture that upside and not the downside – in fact if anything the effect is the opposite. At this point our observation from the Malkeil and Saha research is that hedge funds exhibit “credit-like” characteristics – ie. generally more smooth returns but with negative skew.
Conducting our 5 year risk and return analysis above using semi-standard deviation as the measure of risk (which only uses the negative return results in the risk measure) gives the following.
As shown, changing the measure of risk to encompass only downside risk does little to enhance the risk/return characteristics of hedge funds relative to other asset classes. This is broadly consistent with the Malkiel and Saha study (albeit over a subsequent time period).
Conducting our own analysis of the correlation through time of hedge fund returns versus a pure credit portfolio or an international equity portfolio reveals the following.
It is apparent from the chart above that while the correlation with equities has varied through time (which looks very much like changes in strategy from our point of view not some underlying beta effect), the correlation vis a vis credit spreads has been, until recently, quite high. We can also observe this by reflecting on the rolling performance of hedge funds vs the rolling performance of the high yield index shown below.
Given the above, it would be our conclusion that while some hedge funds have the ability to be diversifiers in a total portfolio, this is a function of:
- the effective asset class diversification that comes from exposure to a number of different hedge fund strategies, not because hedge funds offer some form of diversifying beta exposure; and
- the implementation of the hedge fund strategy (skill, process, or both) that provides an element of varying market exposures that offers diversification to a broader balanced portfolio.
We particularly note Malkeil and Saha’s comments in relation to hedge funds ability to diversify: “we found that hedge funds have returns lower than commonly supposed. Moreover, although the funds tend to exhibit low correlations with general equity indices—and, therefore, are excellent diversifiers—hedge funds are extremely risky along another dimension: The cross-sectional variation and the range of individual hedge fund returns are far greater than they are for traditional asset classes. Investors in hedge funds take on a substantial risk of selecting a dismally performing fund or, worse, a failing one.” Malkiel and Saha (2003).
To paraphrase: while aggregate returns appear relatively stable they in fact haven’t been that great and for individual funds the volatility is huge. This makes the selection process extremely difficult. The averaging process of analysing groups of hedge funds creates an artificial view of low volatility.
Fees and other implementation issues
One of the more difficult aspects of hedge funds (and in our view the ultimate “category killer”) is the implementation of hedge fund strategies. In general we would observe that:
- Most hedge funds offer limited, if any, transparency making it hard for either aggregators of hedge funds or the overall institutional portfolio to get an accurate understanding of the real risks being taken. Most risk analysis of hedge funds, outside of the fund itself, is focused on historical performance data. The suddenness with which hedge fund returns can deteriorate can make such “risk” analysis misleading.
- For superannuation or other investors to get exposure it is typically via offshore comingled vehicles (which may introduce other operational risks) and where liquidity can also be an issue particularly as a fund could be impacted by the movements of other investors in the vehicle. If the underlying strategies themselves are particularly illiquid this can exacerbate the issue.
- Smaller investors also typically aim to get exposure via hedge fund of fund vehicles. Such vehicles offer another layer of limited transparency, cost and operational layering and most significantly an additional layer of fees.
- Fees in aggregate are extremely high relative to more traditional strategies.
With respect to this final point, one of our greatest issues with the inclusion of hedge funds strategies in a portfolio is the very high level of total fees and the substantial incentive skew to managers of the hedge funds.
Typical fee structures on hedge funds can range from circa 1-2% base fees up to over 3% in some cases. Performance fees are typically 10-20% but up to 50%. Even on a more traditional fee scale of say 1.5% base, 20% performance, the total fees paid by investors can add up to a substantial portion of any purported value add. As noted in the Ibbotson paper, total fees were over half of what they determined as alpha – and in aggregate over 3% p.a. of the total return. This represents an enormous drag on total returns that has to be exceeded for the client to benefit.
By way of example, and to highlight the impact of fees on total returns, we considered what would happen to the return of a typical long-only investment with the addition of a hedge fund like performance fee.
To this end, we took the Schroder Australian Equity Fund for the 15 year period from 30 June 1990 to 30 June 2005 (when overall market returns were quite good) and compared the net return to investors from a “traditional” fee scale – assumed here at circa 40bps p.a. and a hedge fund fee scale of 1.5% + 20% of the outperformance over index.
|“Normal” Fees||“Hedge Fund” Fees|
Source: Schroders, Hedge Fund fee scale used was 1.5% base, 20% performance fee with high water mark. Based on Gross Returns of Schroder Australian Equity Fund with appropriate fees added.
We can see that in the above example the addition of a “normal” 1.5% base/20% performance fee to the return series generates an annualised fee of 3.9%. If we were to then add another layer of fees – potentially including a performance fee – for a fund of fund hedge fund layer the return drag would increase further. Alternatively, if we made the benchmark cash or some cash plus proxy (up to circa 6%) as is the case with many hedge funds, fees would be even higher than shown above.
While some funds may be able to justify a high level of fees, it is our view that in aggregate given the number of hedge funds and the size of the industry it is simply unrealistic to expect that the industry as a whole can generate sufficient returns to justify this level of fees.
As a final comparsion, the chart below shows the rolling 3 year performance of the “Investible” hedge fund index versus the full index used above and the Eurekahedge Hedge Fund of Fund index.
It is clear from the above that the “investible” index lags the full industry index by a considerable margin – on average 4.5% p.a. In addition, the hedge fund of fund index lags the full index by an average of 2.6% p.a. These lags are probably a realistic expectation of the upward bias introduced by using the full index or the additional fee drag from using a hedge fund of fund vehicle.
Our analysis above has shown that while hedge funds have generated positive returns, to the extent that this has been in excess of market risk or what other investments have generated is somewhat dubious.
While hedge funds could potentially offer the opportunity to diversify at an aggregate portfolio level, one needs to be careful that the strategy is actually diversifying as the underlying investments are unlikely to be diversifying (in a static allocation). This means that the real challenge in selecting hedge funds is in finding managers with the appropriate skill to justify the fees and the lack of transparency – this is likely to mean that institutional investors have the odds stacked firmly against them in getting this right.
A more appropriate strategy for a fund looking to get exposure to strategies that are potentially diversifying would in our view be one that:
- offers complete or a high degree of transparency as to the application of the process and the underlying investments to enable a more rigorous estimate of total portfolio risk;
- are more realistic in the fees and potential fees paid to the manager – and significantly ensures that these are not skewed in the favour of the manager (as most performance fees are); and
- utilises a strategy that results in either ongoing exposure to truly diversifying assets (rare in our view) or more likely will vary the exposure to the underlying assets in a way that improves the risk and return of the aggregate portfolio (ie. reduces risk assets when downside risks are high and increases them when downside risk is low).
1Ibbotson, Chen and Zhu, The ABCs of Hedge Funds: Alphas, Betas and Costs, Financial Analysts Journal Jan/Feb 2011.
2Throughout this report Australian equities represented by S&P/ASX200, global equity by MSCI World (ex Aus), bonds by UBSA Composite Bond, high yield by Barclays Global High Yield Index.
3Passive balanced fund has been represented as 35% ASX 200, 25% MSCI World (ex Aus), 5% REIT’s, 15% UBSA Composite Bond, 10% Barclays Global Agg, 10% UBS Bank Bill. All returns unhedged in AUD.
4Passive capital stable fund has been represented as 15% ASX 200, 10% MSCI World (ex Aus), 30% UBS Bank Bill and 45% UBSA Composite Bond. All returns unhedged in AUD.
5Burton Malkiel and Atanu Saha, Hedge Funds: Risk and Return, Financial Analysts Journal, 2005
Opinions, estimates and projections in this article constitute the current judgement of the author as of the date of this article. They do not necessarily reflect the opinions of Schroder Investment Management Australia Limited, ABN 22 000 443 274, AFS Licence 226473 ("Schroders") or any member of the Schroders Group and are subject to change without notice. In preparing this document, we have relied upon and assumed, without independent verification, the accuracy and completeness of all information available from public sources or which was otherwise reviewed by us. Schroders does not give any warranty as to the accuracy, reliability or completeness of information which is contained in this article. Except insofar as liability under any statute cannot be excluded, Schroders and its directors, employees, consultants or any company in the Schroders Group do not accept any liability (whether arising in contract, in tort or negligence or otherwise) for any error or omission in this article or for any resulting loss or damage (whether direct, indirect, consequential or otherwise) suffered by the recipient of this article or any other person. This document does not contain, and should not be relied on as containing any investment, accounting, legal or tax advice. Schroders may record and monitor telephone calls for security, training and compliance purposes.