Hedge fund investing has always been something of a paradox: potentially powerful diversification benefits, but at the price of illiquidity, egregious fees, opacity, embarrassing blow ups and often insurmountable minimums. Since the early days of the business, there have been six waves of products designed to preserve the benefits while addressing the drawbacks.
Broadly speaking, these efforts have focused on four issues: diversification, investor access, liquidity and fees. With hedge funds, diversification connotes spreading bets across strategies and managers to reduce volatility, minimize blow ups and limit unexpected drawdowns. Access seeks to address high minimums and expand the pool of potential investors. Liquidity and fees have risen to the fore with institutionalization, regulatory pressure and lower overall hedge fund returns.
Today, these products typically are grouped into the broad category of “liquid alternatives.” Hundreds of billions of dollars have flowed into hundreds of products, reflecting the dire need of allocators for client-friendly solutions in regulated vehicles like UCITS funds. Regrettably, few have achieved the objectives noted above. In many cases, aggressive and unrealistic marketing, followed by disappointing results, left allocators disillusioned and frustrated. The predictable result is an industry-wide crisis of confidence. Despite this, there are a few isolated successes. Arguably, the most successful approach has been the simplest, yet also the most controversial as it challenged both the orthodoxies and business models of many professional hedge fund investors. This note chronicles the six waves to provide historical context for today’s allocators and a road map for creating an opportunity to improve client outcomes going forward.
- Wave 1: Funds of Hedge Funds
- Wave 2: Investable Hedge Fund Indices
- Wave 3: Hedge Fund Replication (v1)
- Wave 4: Multi-Manager Mutual and UCITS Funds
- Wave 5: Global Multi-Asset Class Funds
- Wave 6: Alternative Risk Premia
Fund of hedge funds might seem like a strange starting point: they are, after all, illiquid hedge funds that themselves invest in other hedge funds. In the 1990s, hedge funds were a new, esoteric phenomenon and few family offices had the network or resources to identify and evaluate them; further, most high net worth investors had insufficient capital to invest directly in a dozen or more funds. Consequently, funds of funds were created to address diversification and (to a degree) access, albeit at a steep price: most charged both management and incentive fees above and beyond the high fees charged by the hedge funds themselves.
In the 2000s, as databases of hedge funds proliferated and prime brokers connected hedge funds with investors, access became less valuable. Diversification, however, remained important in the wake of high-profile blow ups like Long Term Capital. Consequently, most funds of funds sought to deliver “index-like” returns by investing in 40-50 managers.[i] Product proliferation soon led to competition on liquidity terms: most funds offered monthly liquidity even though hedge funds themselves were lengthening redemption periods.[ii] Some bank-sponsored funds sought to amplify returns with leverage.
The Great Financial Crisis was devastating to the funds of funds business. First, Madoff and similar frauds raised serious questions about the rigor and effectiveness of the research process. Second, asset-liability mismatches proved disastrous, with widespread gating and suspension of funds. Third, consultants soured on funds of funds and disintermediated them by building direct investment portfolios for institutional clients. Finally, many leveraged products were unceremoniously unwound at the depths of the bear market.
With twenty years of data, we can conclude that funds of funds overall did successfully offer “index-like” exposure. That said, alpha from fund selection was essentially non-existent and failed to compensate for the additional layer of fees. The chart below — of the HFRI Fund of Funds (HFRIFOF) index against the HFRI Fund Weighted Composite (HFRIFWI) index since 2000 – underscores high correlation, but at the cost of net-of-fee underperformance of approximately 1.7% per annum:
[i] “Index-like” was a positive feature when asset class returns were higher; see discussion of post-GFC changes in views on fund selection versus broad exposure.
[ii] During the 2000s, many hedge funds were not registered with the SEC and were reluctant to do so. One potential way to circumvent registration rules was to extend lock up periods to three or five years to more closely align with private equity. Subsequent changes in regulation obviated this.
First launched in the early 2000s, investable hedge fund indices sought to take the fund of hedge funds model one step further: to provide diversification and access, but with daily, not monthly, liquidity. Instead of investing in illiquid hedge funds, investable indices were designed to invest only in specially created (and daily liquid) managed accounts run by the same hedge fund managers that populated non-investable indices like the HFRIFWI. The goal was, in effect, to create a super-diversified, daily liquid funds of funds that investors could use as a core allocation to “hedge funds” – akin to an S&P 500 index fund – which could then serve as the industry benchmark. As with funds of hedge funds (and in contrast to traditional index products), fee reduction was an afterthought given strong overall hedge fund returns net of fees.
While compelling in theory, the products ran immediately into a practical issue: few credible hedge funds were willing to offer managed accounts. The result was a fatal adverse selection issue: out of the gate, some investable indices underperformed actual hedge fund indices by 5-6% per annum. The chart below shows one such product, the HFRX Global Investable Hedge Fund Index (HFRXGL) versus the HFRIFWI from 2004 through 2007.
Clearly, few investors would sacrifice half or more of performance for modestly better liquidity terms. Consequently, the products never achieved any meaningful traction, but they did serve as a warning about the potential costs of porting hedge fund strategies into liquid vehicles.
The clear issues with investable hedge fund indices prompted a new approach in 2006-07. The workaround was that, rather than seek to invest with hedge funds, replication sought to mimic the key drivers of returns using only a handful of liquid financial instruments. On paper, such a strategy could provide diversification (synthetically) by replicating a hedge fund index, offer daily liquidity, yet also be priced at a fraction of the fees charged by funds of funds or even hedge funds themselves.
Based on this promise, we entered the market and launched the first factor-based hedge fund replication fund in May 2007. Our initial focus was on helping funds of hedge funds to address the obvious asset-liability mismatch, but the potential for regulated funds was clear. The pushback from funds of funds was immediate. Replication simply was viewed as an existential threat: if hedge funds could be replicated at low cost with a handful of liquid futures contracts, why would anyone invest in funds of funds with two layers of fees and illiquidity?
As the first wave of products was being launched, the GFC hit. Hedge funds declined more than expected; that plus the prevalence of frauds, single fund blow ups, gating/suspensions and forced liquidations cast a dark shadow over the industry. Ironically, this should have been replication’s shining moment: drawdowns were shallower (the absence of illiquid assets and single stock positions is a positive in a liquidity crisis), frauds were irrelevant, liquidity was never an issue, and headline risk was non-existent. In the chart below, we show the performance of a straightforward five factor replication model built by Merrill Lynch against the hedge fund and funds of funds indices referenced above:
To our surprise, the GFC caused a step backward in thinking about hedge funds: many consultants and allocators who previously had endorsed funds of funds now railed against the latter’s “index-like” approach. Instead, they argued that henceforth their clients should only invest in the “top quartile” or even “top decile” of hedge funds – which generally meant funds that happened to have performed well through the GFC. To more seasoned allocators, this was a self-serving example of seeking to cast blame for an allocation that underperformed expectations. More problematically, though, it was wishful thinking: a portfolio of funds that happened to have preserved capital during 2008 was no more likely to (and almost invariably did not) outperform going forward. Further, given recent drawdowns, many institutional allocators dropped beta targets and virtually ensured that, in a period of near zero interest rates and given high fees, hedge fund portfolios were destined to deliver returns in the low single digits.
For replication, this meant that the natural constituency – pension funds with tight fee budgets and sensitive to liquidity – were steered toward direct, expensive, illiquid hedge fund portfolios. Meanwhile, although some registered funds were launched, the broader criticisms of hedge funds, plus the abstract nature of replication itself, limited broader adoption.
The next wave started in 2012-14 and consisted of multi-manager registered (mutual or UCITS) funds. The foremost objective here was to expand access to retail clients and wealth managers. The products were designed to offer diversification (albeit 8-12 managers, not hundreds), daily liquidity, lower fees (at 2.5-3.0% per annum, exorbitant but half actual funds of hedge funds), and low minimums. Some viewed it as an opportunity for the funds of funds business to rise, phoenix-like, from the ashes of the GFC: even modest allocations across retail portfolios could lead to hundreds of billions of dollars of inflows.
In contrast to the experience with investable indices, many credible hedge funds now were willing to run strategies in managed accounts. On paper, this seemed to solve the adverse selection bias problem. The products, however, contained a different, non-obvious risk: systemic performance drag.[i] Fund marketing materials highlighted the credentials of top tier hedge funds, yet failed to articulate how the stringent rules in registered funds would impact performance. In a 2014 paper, we concluded that the costs likely would far exceed any fee/cost savings, and hence the products were likely to materially underperform actual hedge funds. Using a small sample and early performance figures, we predicted systemic underperformance of 300 bps to 400 bps per annum net of fees – which proved to be accurate.
While the products did materially improve access to hedge fund strategies, it was a pyrrhic victory. Muted hedge fund performance plus systemic performance drag meant clients quickly grew frustrated paying 2-3% in fees for low single digit returns. As of today, perhaps three quarters of such funds have been liquidated.
Replication had asked the question, “what are hedge funds actually doing?” Given the re-examination of hedge funds, allocators now asked, “what do we wish they would do?” For many, the answer was a beta of 0.2 or less (down from 0.3 to 0.4 pre-GFC) with “cash plus 5%” returns. This was, in essence, akin to a top quartile, low beta diversified hedge fund portfolio.
Coincidentally, a growing category of “global multi-asset” funds promised a similar performance profile. As with actual hedge funds, the ethos was that returns often diverge meaningfully across asset classes and markets for sustained periods, and a seasoned team of talented investors should be able to shift presciently into the right asset classes, geographic regions and sectors at the right time. Like replication, the pitch rested on the idea that factor rotations drive alpha. Unlike hedge funds, the products were offered in funds that were low cost, daily liquid and accessible to both retail and larger investors.
The poster child was Standard Life and its Global Absolute Return Strategy. As shown in the chart below left, several years of exceptional performance following the GFC reinforced the argument for superior absolute returns with client-friendly terms. Not surprisingly, geometric asset growth followed, and the family of products reportedly exceeded $90 billion at its peak, arguably the single largest “hedge fund” product ever.
Unfortunately, as shown in the chart above, subsequent performance failed to match expectations. This called attention to an easily overlooked problem with hedge funds: single manager risk. Although invested in hundreds or thousands of individual positions, a single manager fund – in this case, a relatively low risk but unconstrained discretionary macro strategy – is itself highly risky. Each manager has a house view, investment biases, etc. that create a concentration of ideas and investment themes; given investment latitude and the absence of a clear benchmark, dispersion of returns within any hedge fund category is an order of magnitude greater than in traditional asset classes. A well-constructed hedge fund portfolio, by contrast, has clear diversification across strategies, sub-strategies, idea generation, net exposure, investment horizon, etc. – which helps to both increase predictability and reduce risk. In this case, allocators learned that single manager products may appear “index-like” for discrete periods, but that this typically is due to correlation and not causation.
Around mid-decade, a new wave of products emerged: alternative risk premia. Given the awkward combination of high fees and low single digit returns on direct institutional hedge fund portfolios, allocators and consultants were under growing pressure to identify lower cost options and to improve returns. For many, factor replication remained an existential threat and the “factor rotation as alpha” model overlapped too closely with their own asset allocation decisions. Into this void stepped investment banks and quant managers, who argued that – rather than replicate hedge funds exposures – they could build trading strategies to extract pure “hedge fund alpha.” The theory was that most hedge fund alpha was generated through simple trading strategies that, themselves, offered excess returns due to “permanent” market structure or behavioral reasons. On paper, these “alternative” risk premia were irresistible: zero correlation to equities with target returns of cash plus 6%, plus daily liquidity and low fees.
We wrote as early as 2015 that there were several problems with this thesis. Almost all data was back-tested and there was little evidence that hedge funds employed such strategies in a static manner, and even less that they explained much hedge fund alpha. Although marketed as “harvesting” permanent risk premia, we concluded the products were risky, leveraged quantitative macro funds masquerading as less risky index-like, “replication-based” solutions. On the positive side, fees were lower and liquidity terms were favorable.
Today, the space is largely viewed as a (costly) failed experiment. The chart below shows the performance of the SocGen Multi Alternative Risk Premia index vs. the HFRIFWI for the five years through 2020. The drawdowns of several flagship products were far worse, one prominent fund suspended redemptions and imposed dilution pricing in early 2020, and assets have fled the space.
One positive outcome was that allocators began to focus more rigorously on the concept of fee reduction to improve returns.
Our research over the past fifteen years has highlighted the difficulties of meeting the four objectives of diversification, access, equitable fees and daily liquidity. From hard experience, most experienced allocators today are sensitized to adverse selection bias, recency bias, systemic performance drag, single manager dispersion, asset-liability issues, and related risks. We call these the “landmines” of liquid alts.
On the positive side, the evidence is strong that factor replication can be a highly effective and reliable strategy to get “exposure” to hedge funds while side-stepping these landmines. When we recently surveyed the top performing multi-strategy UCITS funds – a category that includes multi-manager funds, alternative risk premia and replication – most of the top funds over the past three and five years employed replication. That said, it is worth noting that some replication practitioners have introduced unnecessary complexity (or their own single manager risk premia strategies) that has fostered confusion and often appears to have hurt performance.
Building on early success, after the GFC our focus shifted to how to use “fee disintermediation” to deliver more alpha to clients. Our research during the early 2010s showed that, while that hedge funds continued to generate meaningful alpha, simply too much was paid away in fees. Hence, our approach evolved to focus on replication of pre-fee, not net-of-fee, returns. Our theory was (and is) that replication of 90% of pre-fee returns at low cost is highly likely to consistently outperform actual hedge funds when, on average, only half of pre-fee hedge fund returns inures to clients. Properly implemented, then, replication has evolved from “index-like” to “index-plus”, since the hedge fund index is, by definition, after fees. Ironically, this approach of “alpha through fee reduction” mirrors progress made by the largest institutional allocators, who have tended to outperform when they negotiate significant fee reductions in return for large, sticky allocations. The promise of the second generation of replication products, then, is the ability to offer a broad range of investors exposure to more hedge fund alpha with essentially all of the client-friendly features of UCITS and similar registered funds. After twenty years of fits and starts, we may finally be upon a solution that solves the hedge fund paradox.