In July 2013, we published a paper on merger arbitrage replication entitled Merger Arbitrage Replication: How Effective Are Rules Based Indices?  The paper sought to address a few key questions about alternative risk premia investment strategies that rely on rules-based trading strategies.  At the time, there were few products with live (as opposed to back-filled) track records, and we sought to compare the performance of those products over time.  Based on the data, we reached the following conclusions:

  • The complexity of parameter specification leads to very different results and calls into question whether there is a readily-accessible “risk premium.”
  • There is no evidence that lower costs translate into better returns for investors.
  • Selecting products is as complicated, if not moreso, than selecting managers; consequently, investors are swapping “manager” for “modeling” risk, which may actually increase idiosyncratic risk.

This note updates several of the charts from the original paper.

Recent Performance of the S&P, Credit Suisse and IndexIQ Merger Arb Indices

After underperforming dramatically in 2010 – 2012, the IndexIQ model has materially outperformed since January 2013. 1The relevant conclusion is that outperformance may not persist and that deviation in performance among models remains widespread.

Recent Performance:  S&P Long Only vs Long/Short Indices

The continued variation in performance between the S&P Long Only and Long/Short Indices is troubling:2

Merger Fund Consistently Matches Hedge Funds

The actively-managed Merger Fund (MERFX) continues to deliver returns close to that of the HFR Merger Arbitrage Index, despite underperforming in the first half of 2015 by 236 bps.3Comparable performance implies that hedge fund managers add value relative to the Merger Fund, but that the added value is paid away in higher fees.  Further, given daily liquidity, the Merger Fund materially outperforms on a liquidity adjusted basis (e.g., due to restrictions on accessing capital, investors should demand that hedge funds outperform liquid alternatives by 200-400 bps per annum).

Conclusion

Merger arbitrage generally is considered one of the easiest “hedge fund strategies” to replicate using a rules-based trading strategy.  Merger arbitrage spreads may in fact represent an alternative risk premium – where returns are tied to a combination of merger activity, rates and equity prices.  The appeal of rules-based strategies is that they can lower costs and reduce idiosyncratic risk – simply by “accessing” the risk premium.  However, reality intrudes when different models result in such widely divergent performance.  If investors are simply swapping manager risk with modeler risk, and the models don’t outperform net of fees, then broader adoption will be tempered.

A recent note[1] discussed a potential solution to excessive hedge fund fees, the Investor Aggregation Model, but highlighted near term obstacles, including industry structure and agency issues.  This note discusses a more realistic alternative:  a Core-Satellite Model that builds on the experience of the traditional asset management industry and recent developments in delivering hedge fund returns at low cost and with liquidity.

Twenty years ago, an institutional investor might have allocated to a dozen different mutual funds in order to get exposure to US equities.  Each fund would charge relatively high fees and the underlying portfolios would likely have substantial overlap.  Today, the same investor is much more likely to get core equity exposure through a low cost index fund, and to use more expensive active managers who have expertise in specific segments of the market.  The low cost, index “core” helps to meaningfully reduce overall fees, and the value-added “satellites” can improve returns over time.

A similar transition will occur in the hedge fund space over the coming years.  The historic obstacle has been a dearth of attractive, low cost “core” alternatives to investing directly in hedge funds.  Investable indices – essentially super-diversified funds of funds – suffered from adverse selection bias and, while offering better liquidity, did little to address the cost of investing.  Early hedge fund replication products successfully outperformed many actual hedge fund portfolios, yet adoption was hindered by product complexity and mediocre index/target performance.  Today, there is a much more compelling suite of liquid investment alternatives – ranging from advanced factor models to risk premia strategies – that have been shown to deliver comparable returns to many hedge fund strategies.

As discussed below, the core-satellite model in the hedge fund space will differ from the traditional approach.  Rather than a single “core” like the S&P 500, hedge fund core exposures will likely be broken down into a combination of portfolio level and strategy specific complements.  Further, a well-constructed hedge fund core will not only lower all-in costs, but can improve liquidity, transparency and risk management – lacking in direct hedge fund portfolios.  These additional benefits should factor into how the allocator “values” the core position.

Defining the “Hedge Fund Core”

You’ll find little resistance to the notion that a US large capitalization manager should be benchmarked to, say, the S&P 500.  But what about a hedge fund?  Allocators might compare a direct hedge fund portfolio to, say, the HFRI Fund of Funds index.  However, unlike the S&P 500, the HFRIFOF is neither investable, liquid nor low cost.  Consequently, its utility is for peer group analysis only – like comparing a mutual fund to six hundred of its peers.

So, how should investors approach this with hedge funds?  Importantly, hedge fund portfolios often have different target returns:  a typical multi-manager portfolio will have much more consistent (and positive) equity exposure than one constructed to have a beta around zero.  Second, the liquidity of the underlying strategies matters:  certain strategies – e.g. equity long/short or CTAs – are more “replicable” than others – e.g. illiquid relative value strategies.

Consequently, one investor’s “core” will differ from another’s.  This argues for customized solutions rather than a one-size-fits-all approach.  Well-designed factor-based replication models are very effective at delivering returns comparable to diversified hedge fund portfolios, while individual factor- and risk premia-based strategies can match or exceed the performance of discrete strategies.  The two approaches are discussed below.

Portfolio Core Allocations

Portfolio core strategies are built on the idea that most hedge fund returns are generated by dynamic exposures to major asset classes.  At a basic level, a hedge fund portfolio with a beta of 0.4 is strongly influenced by the performance of equity markets.  Yes, security selection can add value, but it matters a lot more if equities as a whole are up or down 20%.

The key is that hedge fund exposures are spread across asset classes, and these exposures change over time.  Hedge funds were overweight emerging markets pre-crisis – and benefited when the MSCI Emerging Markets index outperformed the S&P 500 by 121% over 2005-07 – but were underweight in 2012-14 when the S&P 500 outperformed by 61%.  These macro bets, or tactical alpha, drive most hedge fund outperformance over time.

To address this, numerous firms (including ours) have used factor-based models to dissect how hedge funds are positioned across major asset classes.  Live performance has demonstrated that the strategies can deliver comparable or better performance than many hedge fund portfolios.  Importantly, since the strategies invest only in liquid instruments – ranging from ETFs to futures to liquid swaps – the portfolios themselves are liquid – hence, there’s no gating or illiquidity risk.  Further, management fees should be 1% per annum or less, so implementation costs are a fraction of investing in hedge funds.

A 30% customized portfolio core might look something like this:

Customized Core portfolio

Given the current sensitivity to fees, the portfolio core approach provides the most “bang for the buck.”  A 30% core allocation can reduce all-in fees by 100 bps or more while meaningfully improving liquidity, risk management and other criteria, as discussed below.

Strategy-Specific Core Allocations

By contrast, strategy-specific core will seek to deliver comparable performance to a specific strategy.  This approach treats the portfolio as a combination of definable investment strategies and seeks to introduce less costly, more liquid alternatives to one or more of them.  Strategy-specific models range from factor-based strategies to alternative risk premia to 13F clones.  An example with an equity long/short and CTA core is shown below:

ELS CTA Core

The underlying strategy will dictate which approach is most effective.  For instance, equity long/short strategies tend to have relatively stable exposures to various segments of the equity markets – US large and small capitalization, international equities, emerging markets, etc. – hence, a well-designed factor model that seeks to capture 80-90% of pre-fee returns can deliver comparable or better performance over time.  By contrast, CTAs shift exposures much more frequently and consistently rely on certain investment strategies, such as trend following; as a result, a well-designed core requires daily data or must employ specific strategies, such as simple trend following models.  In each case, fee reduction can lead to better performance over time.  Two examples are shown below:[2]

ELS CTA Performance

Past results are not necessarily indicative of future results

One important advantage of this approach is the ability to benchmark individual managers.  If an equity long/short manager consistently fails to outperform a low cost, liquid ELS core, there is a clear and direct measure of value add, or lack thereof.

Defining the Benefits

Unlike traditional investments, hedge fund portfolios are more difficult to manage due to illiquidity, opacity, and other issues.  A core-satellite approach helps to address many of these issues:

  • Enhance long-term performance. The fee difference between hedge funds and replication based strategies can be 300 bps or more per annum; as long as the replication based strategies can deliver the majority of the pre-fee returns of the underlying portfolio or strategy, long-term performance will improve.
  • Improve liquidity profile. Greater liquidity has three important benefits.  First, it improves risk management and provides access to liquidity in extreme market circumstances – gating and suspensions, like 2008-09, should never be an issue.  Second, it improves the ability to tactically allocate to strategies as opportunities arise – you don’t have to wait two quarters to allocate.  Third, allocators can manage exposures more accurately – e.g. manage a glide path or maintain a target allocation by “toggling” exposures up and down at rebalancing dates.
  • Maintain exposure to existing funds. Hedge funds can provide valuable information to allocators on market opportunities.  Allocators can maintain exposure to existing managers and hence preserve the information flow.
  • Concentrate due diligence resources. Allocators can concentrate on due diligence and monitoring resources on more idiosyncratic managers.
  • Expand exposure to illiquidity premium. Many allocators feel pressure to maintain exposure to managers, like equity long/short and macro/CTA, where the underlying portfolios are more liquid, and hence less likely to deliver excess returns over time.  An improved liquidity profile in a core allocation alleviates this pressure.
  • More effective benchmarking. The performance of core strategies can shed light on which managers are truly delivering alpha.

Conclusion

The good news is that most consulting firms and institutional allocators are actively exploring lower cost and more liquid complements to investing directly in hedge funds.  For broader adoption, however, the allocator community will have to overcome two issues going forward.

First, there is concern that the core will systematically underperform hedge funds.  Two arguments in favor of hedge funds are security selection and the illiquidity premium.  On the former, 400 bps of fees is tough to overcome over time, especially when shorting no longer delivers, on average, any alpha.  On the latter, many strategies (such the equity long/short and CTA, as shown above) have very little exposure to illiquid assets.

Second, it requires a paradigm shift among allocators.  As one fund of funds CEO recently noted, allocators must become more like investors to justify their fees.  Investors should value enhanced flexibility – the ability to cut risk when necessary, tactically allocate to more attractive strategies as the opportunity set evolves.  The model also provides a much cleaner way to evaluate whether managers are adding value over time.

The key for allocators is to start to view “fees” as a scarce resource.  When fees are simply “above the line,” it is too easy to ignore the impact on performance.  Pay a manager too much and you don’t get it back if he underperforms; instead, the status quo is simply to redeem and replace him with a better performer.  By setting a fee budget, the allocator is forced to think about where fees are and are not justified.  Paying 2/20 (or more) makes sense when the strategy is capacity constrained, less liquid and truly idiosyncratic.  On the other hand, when the typical equity long/short manager or CTA is outperformed by a low cost strategy core, it forces allocators to seek out managers within the space who bring something special to the table.

[1] How to Cut Hedge Fund Fees in Half:  Part One:  The Investor Aggregation Model

[2] Hypothetical performance of a factor-based replication of the estimated gross returns of the HFRI Equity Hedge index and the Newedge CTA index over the past five years.  Note that hypothetical results include estimated trading costs but are gross of fees.

There is a strong argument that, on average, hedge fund fees today are double what they should be.  Over the past five years, the typical hedge fund generated 2.3% alpha relative to a 60/40 portfolio before fees, yet negative alpha after fees.[1]  The difference is roughly the 300 bps or so that were paid in management and incentive fees.  Granted, a 60/40 portfolio generated unusually high returns over the same period and many hedge funds did much better.  However, when most investors are willing to pay 50% of true alpha generation, the notion of paying away more than 100% seems unjust.

The issue is that the fee structure of the industry has not evolved with institutionalization and asset growth.  As veterans of the industry know, the 2/20 fee structure was designed to provide small hedge funds with an incentive to outperform.  A high management fee would cover costs, while performance fees would align interests with investors – an inducement to attract confident, talented managers to deliver exceptional returns.

How times have changed.  Take a representative $10 billion equity long/short hedge fund.    A 2% management fee generates $150 million or more in pure profits; often, the founder gets paid $100 million or more to walk in the door in January.  Similarly, performance fees almost invariably are paid with no hurdle rate.  The same manager – assuming he’s consistently net long 30-40% – was paid an extra $200 million or more in 2013 simply because equities were up 30% plus.  To paraphrase a private equity titan, performance fees with no hurdle are, “after the wheel, God’s greatest invention” – for the managers.

As discussed below, a more equitable solution would be for management fees to decline as assets grow and for performance fees to be paid above a hurdle.  These two changes could save investors 100-150 bps per annum, improve net of fee performance, and better align incentives.

There are two ways to approach this:  a direct and indirect approach.  As discussed below, investors can pursue a model to negotiate as a group and extract lower fees from the 100 or so large funds that have attracted most capital post-crisis.  An alternative solution, discussed in a forthcoming note, is to migrate to a core-satellite model akin to the traditional asset management business – a lower cost core can reduce costs without diminishing overall performance.

The Investor Aggregation Model

Throughout the business world, firms purchase in bulk at reduced costs and pass much of those savings along to consumers.  Think of the impact of Walmart on consumer retail.  Consumers benefit from lower costs, the firm keeps a spread, and suppliers get scale orders.  Efficiency works.

The structure of the hedge fund industry today is not conducive to this:  buyers (investors) are highly fragmented and hence purchase services from (make investments in) suppliers (hedge funds) on a piecemeal basis.  Post-crisis, as institutional investors steered most capital to the 100 or so largest hedge funds, there’s much more concentration on the supplier side.  Investors tend to make new allocations to funds that are doing particularly well, where demand is highest and the track record of delivering excess returns, net of fees, is strongest.

The end result is that large managers have all the leverage in fee discussions.  Perhaps this is why even small discounts are treated as wins.  The largest investors, like sovereign wealth funds, reportedly will write $250 million checks and expect a 25 bps management fee discount – 5-10% of all-in fees.  At a conference last year, a consulting firm acknowledged that $40 billion of (largely nondiscretionary) client capital resulted in fee savings of 10-15 basis points – or 2-4% of the total.  Taking a different approach, some investors opt to invest in lower cost, but less liquid, share classes – those discounts are generally in the range of 15-20%.  However, when you give up valuable liquidity, you take on added risk:  if the fund runs into trouble, more liquid investors might leave you holding the bag.

In this context, even small discounts are treated as wins – paying $90 certainly is better than paying $100.  But what if “fair” is really $50?  The solution, in theory, is for investors to band together to negotiate fees downward.  This would shed light on what is an achievable, and hopefully equitable, fee level.

Where is the Most Bang for the Buck?

The first goal should be for management fees to decline with asset growth – this would benefit both early and new investors.  Excessive management fees, in a sense, are “pure alpha” – the manager gets a check every month regardless of whether the market is going up or down; conversely, when investors overpay, alpha is reduced dollar for dollar.  Performance should improve as well:  excessive management fees can lead to risk aversion since preservation of firm value trumps risk taking.  (Note that some emerging managers now offer a “founders” share class where management fees decline for early investors as assets grow; although later investors will still pay full management fees, this is an excellent step in the right direction.)

Likewise, the second goal should be for performance fees to be paid above a relevant hurdle rate.  For most equity long/short funds, a hurdle of 30-40% of the appropriate equity index is reasonable.  Over 2010-14, this might have saved 100 bps per annum:  in rising equity markets, investors don’t overpay.  On the other hand, in a severe market decline, the manager can be rewarded disproportionately:  200-250 bps extra if he’s flat when the S&P is down 30%.  Good for investors, good for managers.  For lower beta funds, an alternative hurdle like LIBOR plus 400 bps seems reasonable:  after all, do the managers really think they can’t consistently beat this over time?

What are the Obstacles?

As discussed above, the first is industry structure.  Fragmented buyers have little leverage, but a $1 trillion aggregator could dictate terms.  When Walmart places a $1 billion order, no one questions its ability to perform.  Here we have a chicken and egg issue:  no investor or consulting firm controls enough capital to truly shift leverage.  It’s worth noting that there have been several failed attempts at this, such as investable indices, where the absence of capital up front raised credibility issues and led to adverse selection.

A parallel obstacle is the incentive structure for most allocators:

  • Banding together requires ceding some autonomy which most allocators are loathe to do – if investors break ranks, leverage is lost.
  • Current investors have the most leverage — fifty $50 million investors ready to redeem need be taken seriously by even the largest firms. However, disappointed allocators are much more likely to fire an underperforming manager rather than press for changes in the current fee structure.  It’s simply easier to allocate to a new manager who has been performing well, and hence has justified prior fees.
  • Allocators would have to acknowledge that, in many cases, fees are inequitable. For a current manager, this raises thorny questions about why the fund was recommended in the first place.  For new allocations, it requires a determination of what is and is not justified – not easy to do given the cyclicality of performance.
  • In order to negotiate effectively, allocators need to be prepared walk away and, hence, be excluded from certain managers – difficult to do when access to star managers is an important selling point.

The result is that many funds fail to justify the fee structure, yet the only practical recourse for investors is to redeem and try a different fund.  Since performance is often frustratingly cyclical, high fees paid in one year won’t be recouped if performance suffers the next.  As we know, it’s also very difficult to predict which managers will do better going forward.  Both these points argue for a more aggressive push for lower fees irrespective of recent performance.  The obstacles to this, as noted, are significant.

Conclusion

In a sense, there already are aggregators, if not on the scale described above.  Large funds of funds appear to have meaningfully evolved their models post-crisis to better align incentives.  Some now rarely allocate to flagship funds, preferring instead to negotiate special feeders that concentrate on specific strategies and investment teams.  Reportedly, incentive fees for less liquid strategies are paid over several years rather than annually.  This adds credence to the notion that an investor with real buying power – say, a $500 million feeder to start – can make meaningful progress.  It may also explain why some of those funds of funds have performed much better than industry averages in recent years.

Another alternative is for very large institutions to concentrate their investments.  A typical portfolio might contain five large multi-strategy funds whose portfolios substantially overlap.  Rather than spread the investments – say, $200 million to each – would the investor be better off asking each manager to bid on the whole $1 billion?  This would increase concentration risk, but given how diversified the funds are, probably not as much as many would fear.  It would also shed light on how allocators rank the managers:  maybe Manager A is worth it at full fees, but Manager B is equally attractive at half the fees.  A counterargument is adverse selection; however, it’s hard to see how this would be an issue with the largest funds, where there already is an institutional stamp of approval.

Institutional consulting firms are another option.  They serve as gatekeepers for tens, and in some cases hundreds, of billions of dollars of investment capital.  Most mandates, though, are non-discretionary, so it’s not clear how they would wield their allocation authority.  Presumably, they could remove a recommendation on a manager solely due to fees, but it’s much more likely that such a decision would be based on performance (which, of course, is tied to whether the fees make sense).

Given the industry structure and agency issues described above, real progress on this front is unlikely in the near term.  A more realistic alternative, as discussed in a forthcoming note, is a shift to a core-satellite model.

 

[1] HFRI Fund Weighted Composite index vs a 60/40 portfolio consisting of the MSCI ACWI and ten year Treasury Note over 2010-14.  For caveats on the use of alpha as a measure of outperformance, see Lies, Damned Lies and Alpha.

12. May 2015 · 1 comment · Categories: Uncategorized · Tags: ,

There are three kinds of lies: Lies, damned lies and statistics
Benjamin Disraeli

Investors equate “alpha” to outperformance.  A high alpha fund presumably delivers substantial excess returns relative to its benchmark.  True alpha is short hand for manager skill.  As we know, idiosyncratic returns can improve the risk adjusted returns of a diversified portfolio, and hence investors will pay high fees for “alpha.”  Unfortunately, as a general rule, alpha is widely misunderstood and misused.

Statistically, alpha simply is the result of a linear regression between two return streams.  The regression finds the straight line (ordinary least squares) that best fits the time series.  Visually, beta is the slope of the line and alpha is where it crosses the vertical axis.  The calculation was designed to uncover managers who outperform simply by taking on more risk.  A manager who leverages to outperform the S&P in an up year will show a high beta but no alpha.  Conversely, a manager who took less risk yet matched the S&P will have a beta of less than one and may show meaningful alpha.  This analysis makes sense when you have two very similar investments.  For instance, if a US-focused large capitalization mutual fund consistently earns 1% per annum more than the S&P with similar risk (beta), we’d expect to see alpha of around 1% per annum.

As with all quantitative analysis, however, the devil is in the details.  Importantly, alpha is always relative to something else:  change the benchmark and it means something very different.  Use a different time period (window), and the results vary.  Past alpha often does not translate into future alpha.

In this note, we look at two examples that are particularly relevant to hedge fund investors:  where monthly data is imperfect and when the chosen benchmark is not representative of the underlying portfolio.

Imagine that we’re looking for a long-biased fund.  In a screen of potential managers, we find a fund (Fund A) that has roughly matched the performance of the S&P 500 over the past five years – no mean feat given that the vast majority of mutual funds underperformed.  Yet Fund A did this with less risk, as measured by a beta of 0.8 and a lower standard deviation, and delivered impressive alpha of 2.71% per annum.  Summary statistics are below and the calculation of alpha and beta is shown in the scatterplot to the right (note:  the scatterplot shows monthly data net of LIBOR, so alpha will be roughly twelve times where the line crosses the y axis):

Fund A

Most investors would leap at the opportunity to invest.  There’s a catch, however.  Fund A is simply the S&P 500 with moderate smoothing.  While less of an issue, perhaps, with mutual funds, the level of smoothing is comparable to that of many hedge fund portfolios.  Statistically, smoothing depresses beta and causes a commensurate increase in alpha.  In this case, lower risk and “excess returns” are entirely attributable to a statistical quirk.

Now imagine that we are looking for an absolute return manager to add diversification to our portfolio.  To diversify a traditional portfolio, we are seeking a fund with low correlation to equities.  Fortunately, we find a fund, Fund B, that has demonstrated negative correlation to the S&P 500, yet has managed to outperform LIBOR by over 400 bps per annum over the past five years.  With beta of close to zero, the fund has generated alpha of over 500 bps per annum relative to the S&P 500 (see below).

Fund B

Many investors would conclude that Fund B offers valuable diversification benefits.  Low correlation to equities is very valuable in the seventh year of a raging bull market, and returns of 400 plus over LIBOR – while maybe not what we expected pre-crisis – certainly looks compelling given where rates are today.

As shown in the scatter plot, in contrast to the chart for Fund A, the data points are distributed somewhat randomly.  This makes sense:  when the S&P 500 has a good month, it doesn’t mean that Fund B will as well.  The result, though, is that the linear fit that determines alpha and beta hardly inspires confidence – drop a few outliers or change the time period and you can envision the line shifting around.  The concept of “excess returns” then becomes much harder to support.

Sadly, Fund B is simply the Barclays Aggregate index.  Rather than adding a unique and value added return stream, our allocation to Fund B simply has increased our exposure to fixed income – and probably at the wrong time.  Using alpha and beta as evaluative metrics for very different return streams can lead to wildly misleading results.

Contrary to the concept of “excess returns”, alpha is not transitive.  That is, if A has alpha to B, and B has alpha to C, if does not mean that A has alpha to C.  In fact, and perhaps more confusingly, even when A has alpha to B, B can have alpha to A.  While this shouldn’t be an issue when using an accurate benchmark, you can get a sense of the issue when we compare the S&P 500 to Fund B.  As shown below, over the same period the S&P 500 also shows alpha to Fund B:

Fund B and SP 500

In the analysis, as a result of the negative correlation between the two returns, annualized alpha of 22.74% is substantially higher than actual compound annual returns of 15.45%.  Again, this seems to violate the concept of “value added” or “excess returns.”

So, what does this tell us?

First, inaccurate data seriously undermines calculations of whether a manager is adding value or not.  Returns smoothing continues to be an issue in hedge fund returns for unaudited monthly numbers (weekly numbers are worse).  Granted, the issue is less pronounced than during the pre-crisis years, when hedge funds tended to own a higher percentage of illiquid assets.  Even with more accurate pricing, however, there are more subtle issues that can affect whether the reported monthly return is accurate.  For instance, international stocks close at a different time than US markets and futures markets close earlier than physical markets – an upward or downward move late on the last day of the month won’t be picked up in markets that have already closed.  On a manager level, does it makes sense to price a stock holding at the closing price when a fund, such as a 9.9% activist position, cannot actually trade at this price?

Second, metrics like alpha and beta are only useful with accurate benchmarks that reflect the underlying assets of a fund.  Hedge funds complicate the issue by investing in a variety of asset classes and by shifting exposures over time.  This gets into thorny questions about how best to determine the right combination of benchmarks over a given period of time.  There’s no perfect answer here.  We’ve found replication models far better than anything else because they are adaptive and dynamic; as with any good model, though, it’s important to understand the pros and cons.

Third, in many cases, the calculation of “alpha” doesn’t fit the short hand of either excess returns or manager skill.  Moderately negative correlation to a high performing asset – such as the S&P 500 over the past five years – can result in estimates of alpha that exceed actual returns.  As noted, two managers can have alpha relative to each other.

Fourth, results will vary depending on the time period chosen.  One approach is to look at rolling alpha, but this can suffer from the paucity of data points for any given window and results tend to be unstable.  A related issue is that the correlation structure of the market changes over time:  an unexpectedly large drawdown during a market crisis can reveal sources of risk taking (e.g. selling puts) that aren’t apparent in stable environments.

To draw upon Daniel Kahneman’s work, investors have a natural inclination to search for easy answers.  “Alpha” as a concept has great appeal since it allows investors to condense all the complexities of “added value” into a single metric.  It enables us to quantify in simple terms the value proposition of active management:  a low beta manager is less risky, a high alpha manager adds value.  Statistical models are incredibly useful and have meaningfully improved our understanding of pernicious issues like undeclared leverage and hidden correlations.  However, as with many elegant ideas, it’s incredibly important to pay close attention to the assumptions.  Otherwise, we’ve got a big garbage in-garbage out problem.

Of course it’s true until everyone knows it’s true.
Famous economist in response to publication of Stocks for the Long Run (1994)

Everyone likes bargains.  Perhaps this explains the irresistible appeal of the Value benchmark factor.  The original paper by Eugene Fama and Kenneth French in 1992 propagated the model that investors chase the latest highfliers, while boring “value” stocks are overlooked and underpriced.  The key conclusion was that “value” stocks, as defined by low price-to-book ratios, outperform high multiple stocks and that this outperformance tends to be more pronounced during market downturns – a highly valuable property for an investment portfolio. Historical Sharpe Ratio

But what if it’s all wrong?  Not that the original paper was wrong, but that the return “anomaly” has been priced away or fundamentally changed?  A simple review of risk adjusted returns highlights the issue.  In 1963-90, the period studied, the Value factor had a Sharpe ratio of 0.54 – around double that of traditional assets over time.  Since its “discovery,” the Sharpe ratio declined to 0.17 in 1995-2004 and dropped further to 0.03 in the past decade.[1]

What might this tell us?  Most likely, excess returns were arbitraged away as more investors flocked to the space.  Today, “value” investing is the norm rather than the anomaly, like when Warren Buffett ran a partnership in the 1960s and was buying “cigar butts” for less than net cash.  The growth stock mantra of Peter Lynch and the idea of chasing “ten baggers” seems like a distant memory.  Today, a quick screen on Bloomberg lists 942 value-focused equity mutual funds with assets under management of approximately $1.5 trillion.  Add institutional managed accounts and value-focused investors like hedge funds and the number is much, much higher.  Value isn’t overlooked anymore.

What seems obvious is that the process of identifying and buying stocks is very different today.  Imagine a typical investor in the 1970s.  He reads the Wall Street Journal, scans the stock pages and calls his broker.  The broker recommends an exciting growth stock with a great story, which maximizes his chance of getting paid a commission.  It’s hard to see the same broker trying to explain, say, how a failing consolidated steel company is a bargain because the market’s overestimating the cost of pension liabilities.  The oft touted growth bias of the typical investor, then, may be explained instead by agency issues with the advisor which are less relevant in a world with better dissemination of information, the shift to fee-based advisory and other tectonic shifts in the investing landscape.

Another question is whether the composition of “cheap” stocks has changed over time.  One possible hint comes from the correlation of the Value factor to the overall market.  Correlation was decidedly negative in the period studied and the subsequent decade, which bolstered the argument that cheap stocks were “safer”:  they fell less when markets declined, and vice versa.  Over past decade, however, the correlation has been clearly positive.  This is harder to explain and undercuts the argument for Value as a diversifier under Modern Portfolio Theory.[2]

What’s changed?  One possibility is industry or sector bias.  Unfortunately, it’s difficult to tell given the inaccuracy of simple classifications like SIC codes, especially given how much US industry has changed over the past decades (e.g. deconglomerization).  Another possibility is a change in accounting standards – for instance, the elimination of pooling of interests acquisitions under GAAP in 2001.  Did this fundamentally alter the concept of “book value”? Rolling Five Year Correlation

The danger of investing based on academic studies is that our own biases often rise to the surface.  We tend to overlook the fact that published studies invariably have “interesting” results – the pressure to data mine should not be underestimated.  What looks good on paper may have little relevance once translated into an investment strategy – a failure to screen for market capitalization, for example, can undercut the notion of “investability.”  Conclusions are emphasized over assumptions; yet, the latter are equally, if not more, important.  We often hope that such studies reveal an immutable truism about the market and investors – since relying on truisms is easier than messy realities.

When studies like the HML paper become canonical – what investing course doesn’t cover the Fama-French factors? – we cling to the original conclusions even as the world evolves.  The end result is that academic studies can steer investors to strategies that might have worked decades ago, but not today.  The performance of the Value factor “post-discovery” should be a cautionary tale for investors considering the panoply of smart beta and alternative risk premia today.

[1] Based on data available on Kenneth French’s data library: http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/index.html

[2] Market returns from Kenneth French’s data library were utilized.

Investors and managers find themselves at an interesting inflection point in the evolution of the hedge fund industry.  The growth of liquid alternatives has focused attention on what happens when talented hedge fund managers are asked to manage money within the constraints of a mutual fund structure. The results so far are disappointing:  alternative multi-manager mutual funds underperformed the HFRI Fund of Funds index by close to 200 bps in 2014, despite a 200-300 bps fee advantage (and, as noted in a previous paper[1], underperformed by 400 bps on average in 2013 and 2012). This underperformance mirrors the issues with investable indices, which lagged by an even wider margin in the early years.

Constraints are the Achilles heel of talented investors. Despite deep levels of talent and resources, the vast majority of traditional managers underperform lower cost indices over time. The original concept of the hedge fund industry was that with fewer constraints, talented managers could deliver exceptional returns.  Not surprisingly, many of the original distressed debt investors in the early 1990s were merger arbitrage specialists – when the junk bond market collapsed, they moved to capitalize on it.

The institutionalization of the hedge fund industry has introduced a new set of constraints.  Institutional fear of strategy shift relegates most managers to narrow mandates:  a talented equity long/short manager with health care expertise faces an uphill battle explaining to current investors why half the portfolio should be in credit, another sector, or even cash if the opportunity set within healthcare is not compelling. Effectively, most managers are given a hammer and instructed to look for one particular kind of nail.

This paper looks at two less well known examples:  strategy bias in hedge fund indices, which have been overweight underperforming sectors post-crisis, and funds of funds, whose own constraints have led to meaningful underperformance relative to less constrained multi-strategy portfolios.

Strategy Bias in Hedge Fund Indices

The pre-crisis years were relatively easy sledding for hedge fund allocators, such as funds of funds.  In 2003-07, all major hedge fund strategies generated high risk adjusted returns.  Equity Hedge, Relative Value, Event Driven and Macro all had Sharpe ratios of between 1.29 and 2.17 – exceptionally high relative to traditional assets, where long-term Sharpe ratios typically are 0.2-0.3.

Sharpe Ratio of HFRI Strategies

In this environment, an allocator throwing darts at a wall of hedge fund names would have meaningfully outperformed traditional assets.  This led to the development of a “fully diversified” hedge fund portfolio – relatively static allocations across strategies and, sometimes, dozens of managers.  Diversification would minimize manager risk and, in theory, protect against market drawdowns.  Consequently, the portfolios tended to reflect the composition of the overall industry which, in terms of number of funds, has always been heavily biased to Equity Hedge and Macro.   Today, 70% of the HFR Fund Weighted Index consists of Equity Hedge and Macro funds.

HFRI Fund Weighted Index

Why is this?  Because the barriers to entry are lowest – it’s much easier to launch Equity Hedge and Macro funds, so most funds at any given point in time will consist of those strategies.  Further, because reporting to indices is elective and is viewed as a de facto form of marketing, the composition of the database similarly will reflect the composition of the number of funds across the overall industry.

The post-crisis years have been a rude awakening, with a wide divergence in the risk adjusted returns of the same strategies.  The Sharpe ratio of Macro strategies, for instance, dropped by two-thirds from the pre-crisis to post-crisis period, and was less than a quarter that of Relative Value.  The Sharpe ratio of Equity Hedge strategies dropped by half over the same period.

Sharpe Ratio of HFRI Strategies 2010-2014

Opportunity sets change, but many allocators can only capitalize at the margin.

Viewed in this light, the indices have been overweight underperforming strategies for years.  This explains in part why many of the actively managed hedge fund portfolios have been able to persistently outperform the indices.  As the opportunity sets change, they adapt accordingly.

Funds of Funds vs. Multi-Strategy Hedge Funds

Take the performance of the HFRI Fund of Funds index versus the largest multi-strategy funds[2]. The latter, by definition, have open mandates to pursue opportunities wherever and whenever they arise.  Managers with billions of dollars of capital at stake have the right incentives and resources to maximize risk adjusted returns over time.

In the pre-crisis period, risk-adjusted returns between large multi-strategy funds and the HFRI Fund of Funds index (adjusted for the second layer of fees) were comparable.  However, in the past five years, the Sharpe ratio of multi-strategy funds was almost double that of the typical fund of hedge funds.

Stats

What has driven the outperformance by multi-strategy managers in recent years?   In large part, multi-strategy funds have been much better at adjusting their portfolios to capitalize on shifts in the opportunity set.  The following chart breaks down pre-fee performance according to equity beta, tactical alpha (shifts in asset allocation among key markets), and position alpha (primarily security selection and illiquidity premia).

Fees

The results are striking:  despite similar beta exposure, multi-strategy managers generated vastly more tactical alpha and position alpha.  We see this in greater exposure to credit and emerging markets post 2008, more illiquid assets in 2012, and more aggressive shifts to US equity exposure in 2012-13.  By contrast, under pressure from investors who fear a repeat of the gatings and suspensions from 2009, many funds of funds have shifted to more liquid strategies, which have underperformed on a risk adjusted basis.

Conclusion

The implications of this analysis are several-fold.  First, the most sophisticated allocators, including leading funds of funds, have become more adept at shifting exposures across strategies based on feedback from underlying managers.  In analyzing dozens of live portfolios, we’ve seen persistent outperformance of 300 bps or more relative to the indices:  the best allocators are becoming more like multi-strategy funds than index trackers.

Multi-strategy funds have been much better at adjusting their portfolios to capitalize on shifts in the opportunity set.

Second, the perception that hedge fund indices, with their relatively static weights, have generated disappointing performance is valid.  By analogy, it’s as if the S&P was significantly overweight industrials during a tech boom.  In the replication space, we’ve seen how successfully tracking (and in most cases, outperforming) the fund of funds and liquid indices has failed to produce absolute returns commensurate with expectations from the pre-crisis period.

Finally, there are serious questions about the degree to which newly-launched liquid alternative products – with a panoply of new constraints – will continue to underperform investor expectations.  In some strategies, such as equity long/short, the impact should be minimal.  However, in a diversified model – such as that employed by alternative multi-manager mutual funds – the impact of regulatory constraints on performance may be expected to be far more pronounced.

[1] See Performance Drag of Alternative Multi-Manager Mutual Funds

[2] Custom equal weighted index based on the 16 largest funds in the Credit Suisse Multi Strategy Index.

We’re surrounded by investment products that track indices. S&P index funds seek to replicate the performance of the S&P 500 index – easily accomplished by simply buying the constituent stocks in designated weights. Other indices are more difficult to track – for example when the product invests in futures to approximate spot market returns (GSCI) or acquires only a subsample of index constituents (Barclays Ag).

A new generation of indices promises to emulate more complicated investment strategies, such as currency carry, volatility and roll trades. Investment banks now offer institutional investors an array of derivative products tied to such indices, and asset managers are packaging them into ETFs and other fund products.

One problem, however, is that newly created indices tend to overstate historical, hypothetical performance. From a commercial perspective, there’s little point in launching a new index if the pro forma returns are unattractive; consequently, there’s a strong incentive to adjust the calculation methodology until the results look favorable.

Further, unlike mutual funds, indices can be created and published with minimal disclosure of key information, such as when the index went “live” and what assumptions are made about trading and other costs. The combination can mislead investors who may expect actual net of fee fund returns to match hypothetical gross of fee index returns.

A case study is the PowerShares Multi-Strategy Alternative Portfolio fund (LALT), an active ETF launched at the end of May 2014. This Fund seeks to match or outperform the Morgan Stanley Multi-Strategy Alternative Index (Bloomberg ticker MSUSLALT), comprised of a combination of risk premia strategies designed to deliver absolute returns.

Unrealistic historical index returns

On Bloomberg, the Index data begins on 1/1/2003.  Given the start date, it is possible that the Index was launched sometime in 2013 with roughly ten years of backfilled data.  Unfortunately, there is no requirement to differentiate between backfilled and live results, and neither the LALT prospectus nor Bloomberg sheds any light on when the Index went “live.”

The backfill thesis is supported by historical performance.  The following chart shows the Index returns for the ten years preceding the launch of LALT against the S&P 500 and HFRIFOF index.

The Danger of Indices - 1

Over the decade, the Index “returned” 6.83% per annum with an annualized standard deviation of 2.93% and a Sharpe ratio of 1.64.  The maximum drawdown – during a period that covers the Great Financial Crisis – was only 2.33%.  The Index “delivered” almost 90% of the return of the S&P 500 with one fifth the volatility.  Annual performance was almost 350 bps higher than that of the HFRI Fund of Funds index, which has limited data bias and generally represents live performance. The following table provides some summary statistics:

The Danger of Indices - 2

If the Index represented actual performance, it would rank among the best performing hedge funds over the past decade.  In fact, the risk adjusted return (Sharpe ratio) was better than 97% of all hedge funds in the HFR database over the same period.  Only three live hedge funds had smaller drawdowns. Plus, unlike investing in illiquid and expensive hedge funds, the performance in theory was achievable at low cost and with daily liquidity.

Disconnect between hypothetical and live returns

Prior to May 2014, the Index would have outperformed 97% of hedge funds. Since then it has lagged hedge funds by 600bps.

Since launch, however, both the Index and Fund have failed to meet these high expectations – to say the least.  Each is down approximately 8% since May 2014 – underperformance of 600 bps versus the HFRXGL (daily investable hedge fund) index, which itself tends to underperform the HFRIFOF index by 100-200 bps per annum due to adverse selection bias.  The drawdown over the first seven and a half months is more than triple the hypothetical drawdown over the ten preceding years – during a time when the S&P has risen 8%.

The Danger of Indices - 3

Looking at the live returns, it appears that the Fund and Index were hurt when the Swiss Franc decoupled from the Euro on January 15, 2015.  This underscores the backfill issue:  while the Index “sidestepped” any major adverse market events over the past decade, both the Index and Fund walked into a proverbial propeller seven months after launch.

This issue is particularly timely given the plethora of complicated risk premia products introduced by investment banks over the past two years.  Most indices created recently will be subject to the same backfill bias highlighted above.  A live index, the Merrill Lynch Foreign Exchange Arbitrage Index, is down over 6% in January – but will currency carry indices launched in the future show better pro forma results?  And will investors appreciate this distinction?

Most investment bank indices are subject to the same backfill bias.

 In order to better align investor expectations with likely performance, indices should be subject to the same rigorous disclosure requirements as funds:  investors should know when the index went “live,” which performance is hypothetical, and what assumptions are made about costs and expenses.  Otherwise, the tendency to publish only successful indices will persist.

As of mid-month October, the S&P 500 was down over 5% and the MSCI World was down 6%.  In this context, drawdowns among hedge funds have been unexpectedly large.  Before fees, the HFRX Global Investable index was down over 4%, while the Equity Long/Short and Event-Driven sectors were down 5% and over 7%, respectively (note that the reported losses are lessened by the reversal of accrued performance fees).  The average alternative multi-manager mutual fund (generally with 0.2 to 0.3 equity beta targets) was down 3% net of fees.

What explains the underperformance?  A significant portion likely is due to position crowding, which occurs when many hedge funds hold similar positions.  In good times, additional buying can support stock prices and contribute to excess returns.  For instance, the GS VIP index, which tracks positions in which hedge fund managers have a significant stake, outperformed the S&P 500 index by around 400 bps (per annum) from 2009 to September 2014.  Performance like this is used to support the thesis that hedge fund managers add value over time through stock selection.

In periods of market stress, however, those same positions can underperform significantly as hedge funds cut positions simultaneously.  The table below shows the GS VIP performance during the market drawdowns of 2008, 2011 and the current year:

In each of these drawdowns, widely-held positions declined by 30-50% more than the market as a whole.  While gross underperformance in September through mid-October has not been nearly as pronounced as earlier periods, the data suggests that these positions will further underperform if the markets decline further.

Another form of position crowding occurs when hedge funds invest in a common theme.  This year, many event-driven managers have owned stocks that are takeover candidates due to tax inversion arbitrage; a recent shift in the regulatory environment led to price declines in numerous such positions (most recently Shire, which purportedly caused over $1 billion in losses for hedge funds last week alone).  Anecdotally, many hedge funds also have outsized exposure to oil and gas producers, an implicit bet on high oil prices; it remains to be seen if the recent decline in oil prices has caused outsized losses here as well.  Likewise, the sudden drop in 10 year Treasury yields last week has also been blamed on hedge funds scrambling to cover short positions.

It is quite possible that the concentration of capital among the largest hedge funds will exacerbate this going forward.  Further, numerous investment products now clone long positions of prominent hedge funds (from recent 13F filings) and investors regularly piggyback on positions held by their hedge funds. This additional capital may amplify both upside and downside performance in the quarters and years ahead.

Position crowding is analogous to (or maybe a form of) illiquidity risk.  Alpha can quickly turn negative in periods of market stress, as we saw with illiquid hedge funds during 2008.  Looked at another way, low beta funds became high beta when markets declined.  The same is true for position crowding.  Hedge fund investors who are seeking to protect against downside moves may need to factor this into overall portfolio construction.

Institutional hedge fund investing is entering a new era.  Generation one entailed investing through funds of hedge funds, which offered manager selection, access, diversification and better liquidity in an era when hedge funds were more opaque and less well understood.  The crisis, however, revealed that many funds of funds were running a dangerous asset-liability mismatch as over 80% of redemptions in early 2009 came from funds of funds.  Madoff and other frauds highlighted the risk of failures to adequately diligence certain strategies, which further heightened risk aversion in the coming years.

Generation two has been a disintermediation of funds of hedge funds through direct allocations to a diversified portfolio of single managers.  Driven by institutionalization and dissemination of knowledge about manager selection and due diligence, three quarters of investors today allocate directly.  This has helped to reduce the all in cost of investing by 100 bps or more.

Generation three involves a shift to a core satellite model, where the core allocation to a given strategy or sub-strategy is implemented through low cost, liquid alternatives, including dynamic or alternative beta programs.  A good analogy is long equity investing, where two decades ago an institution might have selected two dozen active managers – each at relatively high fees – but now utilize low cost index funds/ETFs to obtain core exposures and concentrate resources on identifying higher value added, more idiosyncratic “alpha” satellite funds.  Generation three promises to materially drive down all in fees by 200 bps or more, improve liquidity and risk management, and enable investors to concentrate resources on higher alpha opportunities.

Why is this happening today?  There are four principal drivers:

  • Numerous liquid alternative products have established track records of delivering comparable returns to higher cost, illiquid hedge fund portfolios.  This contrasts with investable hedge fund indices, which have materially underperformed due to adverse selection and investment constraints.
  • With lower returns, institutions face increasing pressure to reduce all in fees.  The flow of capital to only the largest firms has prevented a material reduction in fees, even among larger investors.  Whereas the focus of the past five years was on disintermediating fund of funds level fees, attention has shifted to the 2/20 structure.
  • A better understanding of the underlying drivers of performance, such as more sophisticated factor/risk premia models, has demystified the drivers of returns and demonstrated that many funds are overpaid for providing beta-like performance.
  • Capacity issues for larger investors (e.g. CalPERS).

There is little question that the hedge fund industry will continue to grow. Citibank estimates that traditional hedge fund assets will rise to almost $5 trillion by the end of 2018, or roughly five times the total just a decade ago, as institutions diversify away from low yielding fixed income investment and seek to meet high long term return targets.  The rise of liquid alternatives could bring another $1 trillion into the industry as retail investors make meaningful allocations for the first time.

With growth comes the need to evolve.  The traditional investment business has shifted to a model of “pay less for beta, pay up for alpha.”  With more innovative tools at their disposal, institutional investors can better manage their portfolios to enhance net of fee returns, lower costs, improve liquidity and reduce risks.