We’ve spent a great deal of time trying to understand how overall risk levels in the hedge fund industry evolve over time.

A systematic way to assess overall risk is to track the rolling 20 day volatility of the HFRXGL index.  The chart below makes it clear that index volatility has returned to the July lows, and is currently at the lowest level in the past six months.

However, without something to compare it to, we can’t say from this that hedge funds have reduced their market exposures. To do that, we compare the same volatility measure applied to the seven future contracts often used in replication products.  The chart to the left shows the rolling 20 day volatility of each of the key replication factors:  S&P 500, Russell 2000, Eurostoxx, Emerging Markets, 2 Year Treasurys, 10 Year Treasurys, and a USD index.  With the exception of the 2 Year Treasury, realized volatility has been rising in every market through October.

Indeed, the ratio of the volatilities is now half what it was in July. Here is the ratio of the HFRXGL volatility to the average of the market volatilities, showing how much this measure of hedge fund risk has dropped in the past few months.  From this, we conclude that hedge funds, as represented by the HFRXGL Index, have taken risk down since August.

[Thank you to Matt Grayson for the chart below.]

Several times over the past few years, we’ve seen many hedge funds apparently deleverage near a market bottom, only to pile back (late) after the market starts to recover.  The chart below shows what happened last fall, when a hedge fund tracker – which, by definition, won’t deleverage suddenly – bounced back far more than a sample of hedge funds as equities rebounded in October.  (Despite the data issues, we use the HFRX Global Hedge Fund Index here because we can get daily data).

A question that we’ve wrestled with is what’s driving this apparent deleveraging at what, in retrospect, appears to be the wrong time.  After all, hedge fund managers as a group have proved exceptionally adept at managing risk through volatile market conditions.  And for the subset of managers who focus on value stocks, one would think these market dislocations would be an opportunity to add, not cut, positions.

I’ve concluded that there are three interconnected forces that make it difficult for managers to hold losing positions:

  •  First, a byproduct of institutionalization is that investors now have exposure expectations, in addition to strategy/sector/geography classifications.  It’s been taboo for some time for an equity long/short manager to migrate into distressed, even if the opportunity set is more compelling.  But institutional investors and their advisors now seem to place great emphasis on risk stability; hence, it’s harder for a manager to sit on cash as the storm clouds form, but before the storm hits.  Consequently, they’re more likely to find themselves playing defense in the midst of a market drawdown, rather than capitalizing on it.
  •  Second, post-crisis investors used their leverage to force managers to ease liquidity terms – and managers were inclined to do this in lieu of materially cutting fees.  Combined with the institutionalization of the business, this has led to shorter and shorter evaluation periods.   Ask any manager who was flat in 1Q2012, and you’ll likely hear stories about being berated all April by investors for having missed the rally.
  • Third, macroeconomic factors have simply made it more difficult to remain confident in longer term views given the indeterminate downside of a breakup of the euro zone, further sovereign downgrades, the US fiscal cliff, etc.

In the face of this, we find managers struggling to remain focused on core value or to have the fortitude to add to the position.  However, these price dislocations can provide compelling buying opportunities — sometimes at irrational prices.  How many of us would welcome the opportunity to buy a good value stock 25% below where a smart, trusted investor built a significant position?

However, it will be important going forward for smart hedge funds to be selective in how they build their investor bases.  As one manager put it, the “vocal minority can ruin it for everyone.”


“What is the most accurate measure of ‘hedge fund industry’ returns?”

This is an extremely important question for any investor who is seeking to evaluate or benchmark the performance of individual managers.  Consequently, it’s understandably frustrating when numerous firms report different numbers based on proprietary indices and datasets.  In the following I seek to break down this question and propose a solution using a modified version of a widely-available hedge fund index.

To begin with, ideally we would define precisely what we mean by “hedge fund,” aggregate the returns data of each fund (and managed account), and construct a reasonable methodology to use this data to estimate returns.  We’d need to know about every fund out there that falls into our definition, including funds that are now defunct.  We’d want assets under management broken down by fee class and, importantly, information on when each dollar was invested and withdrawn (since high-water marks can result in very different returns after drawdowns).  We might have to think about returns differently for different types of investors:  a $5 million investor might end up with very different net returns than a $5 billion investor.

Clearly, no one has enough data to do this.  What we’re left with instead are various forms of “hedge fund indices” that are put together by firms like Hedge Fund Research, Dow Jones Credit Suisse, Bloomberg and others.  Each firm has a distinct, but overlapping, pool of funds; there are some moderate differences in how they manage the data and construct the actual indices.  Most index providers publish dozens of indices that cover different sectors, geographic regions and, more recently, investment structures.  Not surprisingly, each firm touts its own dataset and construction methodology as the most robust and accurate.  We have some views on this, but this comparison falls outside the scope of this memo.

What we lay out below are some of the strengths and weaknesses of the three major categories of hedge fund indices:  the traditional, non-investable indices of hedge funds; investable versions of these based on subset of those managers; and (also non-investable) indices of funds of hedge funds.  We use three well-known indices from HFR for this analysis.  What we hope to demonstrate is that the non-investable indices of hedge funds of funds, adjusted for fees, provide the most accurate (or least-biased) representation of industry returns.  But first, we outline a few issues with hedge fund indices in general.


Every hedge fund index and database will have its own set of biases.  This is, unfortunately, unavoidable.  There are three broad issues that affect each of them.

First, reporting outside a fund’s investor base is voluntary.  Managers that do report generally have discretion on which funds to report, which share class to highlight, and in which category or sector they choose to be classified.  It’s been argued that there is rigorous screening on these points, but we haven’t seen it.  What we have seen are funds that only report lower fee share classes – presumably to show higher returns – and others that show high fee shares classes – presumably to give more room to negotiate fees.  Also, managers can choose which, if any, databases they’ll report to, so each index will have somewhat different constituents.  Some databases permit funds to delay reporting for up to a few months – in other words, the manager can wait until December to decide whether to report a difficult September.  Some large funds refuse to report altogether or change their reporting methods (Paulson stopped reporting intra-month to one of databases during 2011 after the results kept leaking to the press).

The second issue is attrition rate.  Based on the HFR databases, around 15% of funds cease reporting every year, while another 15% or more take their place.  Imagine how you’d think about the S&P 500 if 75 companies dropped out every year.  We do know that the new companies tended to outperform the index prior to joining, but they quickly gravitate to the mean going forward (to its credit, HFR does not allow those pre-index returns to distort its indices; at least one other index provider still permits this).  In other words, the new entrants look a lot like the other funds, so this doesn’t seem to cause serious distortions in the results.  It’s much harder to know what happens to the funds that drop out; we assume that the majority of funds drop out due to poor performance, and we’ve been able to make some rough statistical estimates on this point.

A third issue is comparability among the indices.  As noted above, not all funds report to each index, so the constituents don’t fully overlap.  This becomes more of an issue with the sector-specific indices, where the number of funds is lower:  in other words, you’ll see less divergence between two industry-wide indices where 1,500 out of 2,000 funds overlap than a health-care specific index where 30 out of 80 overlap.  Construction methodologies also differ:  some equally-weight results, which over-emphasizes the returns of smaller funds, while others asset-weight (specifically, the last time I ran a screen, the average fund in the HFRI database had $40 million in AUMs).  Screening and selection criteria for specific indices may be quite different as well.


There are three broad categories of indices.  We discuss each of these below.

–           Non-investable indices.  These are indices of hedge funds in which managers voluntarily report their results.  Think of it like the S&P 500, except the companies tell you what they returned every month.  When you read about that the “hedge fund industry gained x% last month,” it was most likely one of these.  Non-investable indices have been criticized for various forms of data bias:  backfill, survivorship, construction and others.  Suffice it to say that, from what we’ve seen, the construction and reporting of these indices has improved over time.  The non-investable HFRI Fund Weighted Composite includes over 2,200 funds and is equally-weighted (a clear construction bias issue).  Statistically, though, the one bias that really matters is reporting bias: managers can wait to report a bad month for some period of time to see if they recover so, presumably, some managers will have terrible months and simply never report them.  How do we know this is a real issue?  For the past two years, the “stragglers” have caused downward revisions of the initially reported (“flash”) monthly numbers by 20 bps on average, or 240 bps per annum; to further prove the point, in bad months (May 2010 and July-September 2011) the average downward revision has been over 75 bps.  Given that 15% or so of managers drop out every year, and many of them will have had a lousy month or two along the way, this is a real issue.  Because of this bias, we estimate that the index results are overstated by 100-150 bps per annum.

–           Investable indices.  These are indices that were supposed to mimic traditional indices like the S&P 500:  liquid, investable and fee efficient.  The HFRX Global Hedge Fund Index is one of these.  You can think of this as a highly, highly diversified fund of funds (it has over 200 managers), but excluding fund of funds level fees.  The problem is that this particular index has performed terribly relative to the rest of the industry.  Think of an “S&P 200” index that selects the worst companies out of the S&P 500.  The best explanation for the underperformance is that the managers who agree to be included in the index (and its stringent liquidity and other terms) are inferior and therefore desperate for new assets.  Another possibility is that, by definition, these managers are focused on liquid markets where excess returns are scarce.

–           Fund of funds indices.  Also “non-investable,” these indices appear to have the fewest data biases.  The HFRI Fund of Funds Index includes over 600 funds.  The fund of funds results, by definition, reflect asset-weighted returns, and the results have to include even those funds that might drop out of the non-investable indices (funds of funds don’t let their underlying managers wait three months to decide whether to report a bad month….).  Although the drop out rate is high – remarkably, 70% of funds that were reporting five years ago are no longer in the HFR database – we don’t see reporting bias as a significant issue.  The principal issue is the second layer of fees.  Fortunately, this is relatively easy to estimate – 125 bps on average.   Consequently, the most accurate approximation of “hedge fund industry” returns is a non-investable index with estimated fees added back in.  For simplicity, we call this HFI*.


The following chart shows the performance of the HFRXGL, HFRI Fund Weighted Composite and HFRI Fund of Funds indices over the ten year period ending in December 2011.  We’ve added the FoF index, adjusted for estimated fees.  There are two key points to take away from this.  First, the HFRXGL has persistently underperformed the HFRI Fund of Funds index by around 200 bps per annum; however, if you add back fund of funds fees to make them comparable, this figure rises to 300-400 bps per annum (vs. HFI*).  Second, note that most of the outperformance of the HFRI Fund Weighted Index occurs after the crisis.  During 2009, funds of funds were dealing with a flood of redemption requests, gating issues and other problems.  Consequently, they were effectively delevered during the sharp rebound in the markets later that year.  The non-investable indices, on the other hand, did not have similar issues and therefore recovered more.

Here we show the same chart over the past three years.  We see the same pattern of persistent underperformance of the investable index, although in fairness this underperformance has been less severe over the past few years.


It is clear that investable indices have materially underperformed over time.  The question is, why?  One important difference is that the funds that comprise investable indices are likely to be restricted to more liquid investments.  Many would argue that illiquid investments should garner a return premium over time (endowment model); arguably, if other hedge funds have more of their assets in illiquid securities, this would explain some of the long-term performance differential.  While this argument is compelling, we haven’t found a good way to disentangle illiquidity premia from the adverse selection issue described above.  Hence, the jury is still out.

The good news is that some index providers are expanding their offerings into new areas, which will help with comparative analysis in the future.  For instance, CSFB introduced a non-investable index specifically designed to track managed accounts and similarly liquid vehicles.  We hope that these indices will provide valuable information over time on measures like illiquidity premia, and look forward to sharing the results when they become available.

Leonard Mlodinow of CalTech wrote a terrific short book on randomness titled The Drunkard’s Walk (recommend it highly).  My favorite section deals with Bill Miller and the “hot-hand fallacy” – that past returns are not necessarily indicative of future returns.

As you’ll recall, Bill Miller was the portfolio manager for the Legg Mason Value Trust Fund, which was renowned for outperforming the S&P 500, net of fees, for fifteen straight (calendar) years starting in the early 1990s.  Mlodinow does a terrific deconstruction of various biases in the data:  how the weighting system in the S&P 500 depressed reported performance in certain years, how his streak would not have been nearly as impressive if non-calendar year intervals were used, etc.

But what I found most entertaining was his critique of a CSFB estimate of the probability of a manager outperforming the S&P 500 for a dozen consecutive years:  1 in 4,096, 1 in 477,000 or 1 in 2.2 billion.  Needless to say, he mocks the need for three widely disparate estimates.

But then he further and elegantly argues that the true probability is closer to 75%.  The low odds applied if someone had selected Bill Miller in particular at the beginning of 1991 and had calculated the probability that he specifically would have outperformed the S&P over the next ten plus years.  Mlodinow reframes the question by asking, “what is the probability that one fund out of all funds would have outperformed the S&P 500 for any given fifteen year period?”  Here the probability is quite high.

That said, there is something that feels quite non-random about Miller.  He tended to make large bets, and they worked for a long time.  When his streak ended, his performance didn’t revert to the mean, as implied by Mlodinow.  He then subsequently ranked among the very worst large cap managers over the coming five years.   As the Financial Times pointed out,

Bill Miller, a previous Morningstar manager of the decade, who could do no wrong through the 1990s, has struggled with poor performance since 2005. Now his Legg Mason Capital Management Value Trust is ranked last of the 840 funds in its category over the past five years by Lipper, losing 9 per cent annually.

See FT’s very good article on the same subject here.

Funds by definition are not entirely random.  They are managed by individuals with particular investment biases, and those biases interact in complicated ways with a particular manager’s opportunity set.  Warren Buffett was making very different investments in the 1960s than today, yet there are common elements in how he intrinsically approaches any investment prospect.

Where this becomes useful in thinking about hedge funds is that it helps to understand that past performance is not necessarily a good predictor of future performance.  For one thing, as funds succeed, they tend to grow assets quickly and the opportunity set narrows radically.  For another, market forces can be a critical determinant of returns and are too often overlooked as a contributing factor.

It’s also important to remember that managers will have a particular bias – value vs momentum, activism vs. passivity, asset classes, geographic regions, etc.  As the business has become more institutionalized, the flexibility of managers to deviate from a core philosophy has been curtailed, despite the relative attractiveness of different areas.  Consequently, when evaluating a particular fund, it is critically important to understand how these biases interact with the manager’s opportunity set.

The FT has a compelling cautionary tale on how yesterday’s stars can be tomorrow’s disasters.

Broadly construed, statistical models effectively can break down the return stream of a given hedge fund portfolio into a set of weighted exposures to certain market factors (e.g., the S&P 500, Treasurys, etc.).  This approach has clearly demonstrated that hedge funds, as a group, generate the majority of their returns through shifts in asset allocation among major markets.  Even during the “high alpha” pre-crisis period, the vast majority of returns came from increases and decreases in broad market exposures.   Looked at this way, the hedge fund industry is a powerful guide to asset allocation decisions over time.

This consistency of exposures contradicts the belief that security selection drives the majority of returns – while true for individual managers and possible for highly concentrated portfolios, this isn’t the case for diversified portfolios.  What this means, then, is that the information we derive from these models is valuable and useful:  it provides a window into where hedge funds see better returns going forward.  In one striking example, we saw a major shift in the portfolio in early 2011 away from emerging markets and into US equities.  In retrospect, it’s quite clear today that this accurately reflected a reassessment of the relative attractiveness of US equities versus those in developing markets. The chart below displays the respective weights in the portfolio for emerging markets versus domestic equities:

In practice, as we spend a great deal of time looking at historical returns in order to determine what precisely drove returns, we learn a great deal about the characteristics of the strategy or portfolio.  We know that certain strategies retain more consistently stable market exposures over time.  We can study whether a hedge fund allocator shifts exposures aggressively or not.

The common concern is that statistical analysis is “backward looking” and therefore is only marginally relevant to understanding hedge fund whose managers are “forward looking.”

First, the empirical results over the past five years strongly support the position that hedge funds shift exposures slowly enough (they do change, but over months and quarters) that good statistical models do in fact detect the shifts on a timely basis – even though they are “backward looking.”  Some strategies are particularly stable:  equity long-short, event-driven and credit managers, to name a few.  Managers that do shift exposures rapidly – such as CTAs – are much more unpredictable, but usually constitute a minority of industry exposure.

Second, for the majority of hedge fund strategies, it “makes sense.”  Equity long-short, event-driven and credit managers generally build up positions over months and quarters based on incremental research and market information.   Consequently, today’s portfolio reflects a cumulative series of decisions over the past year or more.  Whether they make or lose money this month will largely be the result of accumulated decisions over the preceding quarters. Therefore, the positions today are, in fact, “forward-looking” in that they reflect the collective views of the underlying managers of where returns will be.  The importance of recent returns—and their distillation into market exposures—remain important windows into prospective opportunities in the industry going forward.

Typical hedge fund seed investing is subject to four disadvantages in the current investment climate.

  • First, it is generally well known that, in most circumstances, managers who need seed capital are those who are otherwise unable to raise it themselves.  This creates adverse selection bias of an indeterminate amount.   Post-crisis, those seed investors who were still offering to provide capital were only willing to do so at materially disadvantageous terms – such as 25% perpetual revenue shares – which further drove away better candidates.  A 25% revenue share is roughly equal to a 50% interest in the management company.  Therefore, if a seed investor was willing to provide $40 million – a fairly typical amount – the manager is starting with a management fee base of $600-800k.  This is insufficient today to support building out the infrastructure that most investors expect.
  • Second, the biggest issue today for smaller funds is how to grow from, say, $40 million AUMs to $250 million.  The revenue share makes it uneconomic to hire outside marketing support, especially third party marketers who will bear most of the costs of marketing in return for yet another revenue share.   Since investors generally want to speak to the manager, not marketers, any concerted effort to raise capital draws the manager away from investment decision-making and further reduces the likelihood of generating outsized performance.
  • Third, the pre-crisis investor base for smaller funds – family offices, European private banks, and smaller funds of funds – is still largely absent from the market.  The consolidation of the funds of funds industry means that most funds of funds must invest predominantly in larger funds.  Institutions and their advisors generally shy away from smaller funds due to perceived headline risk.
  • Fourth, a lower return environment means that all-in fees for funds are lower on average than they used to be.  A fund averaging 15% net might take in 5-6% on AUMs; a fund at 5% net two-thirds less.

In light of this, I would contend that providers of seed capital are inadequately compensated in most seed transactions.  To use some simple assumptions, if the seeded fund grows to $250 million in three years, the seeder is picking up an excess return of $1.5-2.0 million per annum (assuming 25% of management fees plus carry of $7.5 million).  While this represents a compelling 4-5% of excess annual return on the original $40 million investment, on a probability-adjusted basis the actual benefit is perhaps one-fifth of the amount, or around 1% per annum.

Most investors require an expected return premium of around 300 bps per annum for locking up money for multiple years.  Given that seeding vehicles require multi-year commitments, I would argue that this simple example shows that in most circumstances, on a probability and liquidity adjusted basis, seeding has negative net present value relative to direct investing.

[Many thanks to Matt Grayson for his help on this.]

Many investors focus on alpha when considering the returns of a particular hedge fund.  As shown below, while alpha at times is a compelling measure of risk-adjusted returns relative to a specified benchmark, investors should be aware of some of its limitations as an analytical tool.

In the examples below, I first show how a non-correlated return stream can have identically high alpha relative to two other returns streams even when one of those returns streams is definitively superior.

Consider this hypothetical returns series:

B has an Alpha of 1.3% per month over A. Indeed, B is a slightly noisy version of A with a constant added, as the plot below shows.

Series C is a slightly noisy constant, and is thus very uncorrelated to A. Its alpha over A is about 1.2% per month.

Since B is very correlated with A, it should be no surprise that C also has an Alpha of about 1.2% per month over B.

The Important point is that even though B consistently outperforms A by 1% per month, C has the identical alpha over both of A and B.

A more subtle example follows.  In the scatter plot below, we show two non-correlated return series.  Each return series has a positive mean return.  Statistically, Series D as an alpha of 0.1 per month over Series E.  Paradoxically, Series E also has an alpha of 0.1 per month over Series D.

The key point here is that alpha is a measure that explains the outperformance of one return stream against another, after adjusting for the amount of return that is explained by the other.  If the return streams are non-correlated, then the explanatory power is lost.

The seminal work on discounted cash flow valuation methodologies arguably is Tom Copeland’s Valuation.   For anyone who wishes to understand how many private equity firms think about financial models and define free cash flow, this is the place to start.

While DCF methodologies are a critically important tool in the determination of value for any business, I’ve been puzzled recently about the way in which smart investors deal with a few of the assumptions that underlie the analysis:

  • Product and business lifecycles have shortened.  DCF models depend on predictability of cash flows out many years.  This calls into question the ability to make accurate predictions of business fundamentals out more than a year or two.  I was startled to hear a few months ago that the iPod was launched around a decade ago (and the iPhone half as recent).  Creative destruction is more active than ever.
  • Macroeconomic volatility compounds the issue by introducing large externalities – i.e., a breakup of the euro zone – that can have an indeterminate impact on forward cash flows.  My good friend, Jim Surowiecki, wrote a book called The Wisdom of Crowds that provides very good examples of how decision making by smaller groups (i.e., politicians) is subject to wider dispersion than those by very large groups (i.e., the invisible hand).   We’re seeing the same thing with Europe and the US fiscal cliff, where it becomes difficult to quantify the downside when we can’t be sure that the key players will reach an economically rational equilibrium.
  • DCFs are highly sensitive to estimates of weighted average cost of capital.  It is clear that there are massive distortions in global interest rates, and equity market premia have effectively been negative for some time.  Consequently, this must have a real effect on the ability to accurately estimate a given company’s WACC.  Given that cash flows implicitly extend for decades, minute changes in the WACC will have a profound impact on present value.

Take the simple example of a firm with $100 of free cash flow that’s growing 5% per annum and we decide that a reasonable weighted average cost of capital is 10% today.  Calculated as a growing perpetuity, the firm should be worth $2,000 today.  However, 70% of the net present value is due to cash flows in years 11 through infinity.

What DCFs don’t incorporate are valuable options embedded in a business.    This was the case with cable companies, where they were erroneously priced as growing perpetuities in the 1990s (even though they needed to continually reinvest in their businesses, and hence free cash flow was depressed).  However, the value of “owning the biggest pipe” into the house was soon appreciated and understood by the market.

When we speak to managers, we prefer professionals who are always willing to question the assumptions underlying the analysis and maintain a healthy skepticism about the results on any single methodology.

Any insights from investors in the investment business – private equity, hedge funds, etc. – on how best to address these limitations would be more than welcome.

Many analysts have written about the scope and magnitude of estimated biases in hedge fund indices:  selection, survivorship and reporting bias.  As it relates to hedge fund indices, selection bias has three forms:  when only funds that are likely to do better than their peers are added, when strong performers are added retroactively, or when the fund sample is not representative of the overall population.  Survivorship bias is a problem when only good funds remain in the index and reporting bias arises when funds tend to report better numbers.  For reasons outlined below, as it relates to the HFRI Fund Weighted Index, the principal issue is a combination of survivorship and reporting bias.

As a starting point, funds that elect to report to the HFRIFWI have, on average, outperformed the index by around 7% per annum in the preceding three years.  Not surprisingly, a manager who has outperformed has more of an incentive to voluntarily report than one with poor performance.

That said, selection bias is not a significant issue in the HFRIFWI for two reasons.  First, those managers don’t do materially better than the index after they’ve joined; they revert to the mean almost immediately.  Second, HFR correctly does not adjust the prior index returns to reflect new additions.  (Unlike HFR, at least one other index provider — Eurekahedge — revises prior index returns as though each new addition had actually been in the index since inception.  This causes prior monthly numbers – e.g., December 2011 – to drift upward over time and effectively renders the data useless.  HFR hasn’t done this for as long as I’ve followed it, and Hedgefund.net told me they stopped doing this in 2005.)

HFR’s indices are, however, subject to two forms of data bias that are worth discussing.  First, HFR constructs its monthly indices as an equally weighted average of the monthly returns of the underlying constituents.  The HFRIFWI currently has around 2,200 funds that range in size from $19,000 to over $21 billion, with a median fund size of $45.7 million.  The smallest and the largest are given equal weighting in each month’s results.  This becomes a real issue, obviously, if smaller funds materially outperform over time and skew the results upward.  Interestingly, though, at least as it relates to the HFRIFWI, we don’t find that smaller funds do in fact outperform, so we tentatively conclude that this is not a material issue (at this time).

For the HFRIFWI, the more relevant bias is a combination of survivorship and reporting biases that is caused by the fact that HFR permits any fund in the index to wait up to 90 days to report its results for a given month.  Due to reporting delays, HFR publishes a “flash” estimate around seven days after the end of the month – generally, 20-25% of the index constituents — and will periodically revise that month’s return for until it’s “locked down” at the end of 90 days.  The practical effect is that managers who performed well in a given month generally report earlier than those who did poorly, and some who did really poorly never report at all.

To put this in perspective, since 2009 the flash number has been revised downward 63% of the time and by an average of 36bps, which equates to an annual downward revision of 4.41%.   The downward revisions tend to be much more pronounced during particularly bad months for hedge funds, which likely is due to a combination of intentional delays by certain managers and difficulty in settling on marks for some positions.  The following chart shows the month-by-month revisions since last summer and shows how much more pronounced the downward revisions are in down months (August and September 2011 and May 2012, which won’t be finalized until next week).  This should give you pause when you see the first reported “flash” results come across your Bloomberg screen.

What this doesn’t show us is the effect of managers who stop reporting altogether. Take the following example:  a manager is down 20% in August 2011 decides to delay reporting.  If the fund recovers in September and October, he/she will report August before the end of November and remain in the index.  However, if it doesn’t recover, the manager may never report August or later months and the index will be somewhat overstated for that period. Given that 15% of funds stop reporting during the course of any given year, this is a real issue.

We’ve found that the best way to estimate this effect is to compare the results of the HFRIFWI with those of the HFRI Fund of Funds index – an equally-weighted index of over 600 funds of hedge funds.  The reason is simple:  funds must report down months on a timely basis to their investors, and we see this flow through the fund of funds results.  We first assume that both the HFRIFWI and HFRIFOF are reasonably representative of the overall industry, then normalize the results for funds of funds level fees (which we ballpark at 125 bps on average).  The difference between the HFRIFWI and the fee-adjusted HFRIFOF should primarily reflect reporting bias.  In the chart below, the differential over the past five and ten years is between 100 bps and 200 bps per annum.

The next question is what this tells us about those funds that stop reporting.  The best that we can do is infer how much they underperform the rest of the index during the period immediately following their last reported result.  If we apply a 15% attrition rate, and assume that this bias accounts for around 150 bps per annum, then the average fund will have underperformed the index by around 10%.  Granted, we don’t know whether this underperformance occurred in a single month or over a longer period, but the magnitude is interesting.  This affirms the view that the industry is fairly ruthless about shooting the wounded.

Over the next few years, I anticipate that a greater number of hedge fund investors will use liquidity tools to enhance portfolio efficiency and risk management.  Broadly speaking, these liquidity tools encompass investment strategies that seek to deliver hedge fund-like returns but with material structural advantages, like daily liquidity, real-time position-level transparency, no headline risk, scalability, and other features.  I’ve spent a lot of time looking at factor-based models that, in essence, seek to understand the current broad market exposures of the hedge fund industry or a portfolio of hedge funds, and to invest in those market factors directly through liquid futures and ETFs.  I’ve found that there are three key things to understand about these strategies:  the value proposition, why past returns are relevant, and why this approach is fundamentally different from passive indexing in traditional assets.


The goal of a well-designed liquidity program is to deliver similar returns to either a hedge fund index or portfolio.  It generally doesn’t work well for an individual manager – and I wouldn’t recommend trying.  That said, the strategy isn’t designed to outperform, but rather to deliver some or all of the returns of a target.  In practice, there will be periods of outperformance and underperformance, but a well-designed program should deliver the majority of the target returns over time.  Therefore, the principal “value” lies in structural benefits, not outperformance.  Put simply, a 10% return with daily liquidity is more valuable than a 10% return with a long lockup.[1]  Flexibility is valuable:  an investor who could exit the markets in late 2008 would have preserved capital and was far better positioned to re-invest in early 2009.

The question for each investor is how to think about this value proposition and how to deploy it in practice.  Each investor will ascribe a different value to benefits like liquidity, transparency, scalability, lack of headline risk, and other features.  An investor who demands a 300 bps (expected) return premium for investing in an illiquid asset will realize 300 bps of “value” if a liquid investment can deliver the same return.  Other investors may ascribe more value to intangible benefits like the absence of K1s or reduced internal monitoring/administrative costs.  A large institution may place value on the reduction in headline risk, real-time position-level transparency, and low all-in fees.  I personally place a high value on the information about how exposures shift over time — a window into where hedge fund managers see opportunities.


All predictive quantitative models are grounded in the premise that the past is representative of the future.  In factor-based programs, these models rely on an assumption that the composition of the industry and the key (market) drivers of returns are relatively constant over time, not that future returns will mimic past returns.  This is an important distinction and the evidence strongly supports this position.

What this means in practice is that it’s easier to approximate the returns of strategies with relatively stable exposures over time – equity long-short, credit, event-driven, distressed.  Others – CTAs and macro – are more variable over time and hence less predictable.

The factor models require a year or two of historical returns in order to estimate the current weights of the target portfolio/index.  In this sense, the models are always “backward looking.”  What we’ve seen over the past five years is that hedge funds, as a group, shift exposures slowly enough (they do change, but over months and quarters) that the factor models are able to detect material shifts on a timely basis.  While this may not be true for individual managers, it does hold true for diversified portfolios.

For more stable strategies, this “makes sense.”  Equity long-short, event-driven and credit managers generally build up positions over months and quarters based on incremental research and market information.   Consequently, today’s portfolio reflects a series of decisions over the past year or more.  Whether they make or lose money this month will largely be the result of accumulated decisions over the preceding quarters.  The factor models pick up on this.  Therefore, the positions today are, in fact, “forward-looking” in that they reflect the collective views of the underlying managers of where returns will be.

Consequently, an ancillary benefit of the liquidity programs is that they can provide a window into where hedge funds see better returns going forward. In early 2011 we saw a major shift in a typical portfolio away from emerging markets and into US equities – it’s quite clear today that this accurately reflected a reassessment of the relative attractiveness of US equities versus those in developing markets.


The principal difference between hedge fund liquidity tools and traditional passive index products is that you can easily and efficiently invest in the securities underlying the traditional index.  Not so with hedge funds.  A decade ago, some firms sought to create ultra-diversified portfolios of hedge funds – akin to buying 150 of the stocks in the S&P500 – as a means to generate an index-like product for hedge funds; Unfortunately, these “investable index” products suffered from data biases that caused persistent underperformance.

In general, any hedge fund liquidity program will have substantially larger tracking error than that of a passive index.  While we think they’ve proven to be highly effective over time, investors should be prepared for more month-to-month tracking error given that the models provide a (good) approximation of exposures over time, but that idiosyncratic factors can cause divergences.  Also, depending on the target, it may be prudent to expect that the liquidity program will provide only a portion of the target returns; however, given the structural benefits and cost of holding cash, this may still be compelling.

The second issue is that investors are more skeptical about the returns of hedge fund indices.  Many analysts have noted data biases that can skew the results upward or downward over time.  The result is a lack of unanimity as to which benchmark(s) provide the best representation of industry returns.

I’ve generally preferred to use actual portfolios as a target – this better aligns the liquidity program with the investor’s chosen hedge fund portfolio.  However, actual portfolios are more likely to change materially over time, and idiosyncratic manager risks can elevate tracking error.  That said, I’ve found that liquidity tools can effectively deliver the vast majority of the returns of a well-managed portfolio and provide a powerful means to improve risk management and flexibility.

[1]       Liquidity per se is difficult to value.  Endowments generally expect a 400 bps return premium for investing in illiquid assets.  We’ve looked at the value of being able to cut risk in portfolios below a certain loss threshold (i.e., down 5%); depending on the strategy, this can result in benefits of up to 300 bps per annum.