Everything Investors Know About Hedge Funds Is Based on Flawed Data

Researchers have found that performance is far higher than people think once the data includes the largest institutional hedge funds. “Every single piece of research needs to be revisited.”


Illustration by II

If the best performing small-cap companies hadn’t publicly reported their returns for decades, investors would question everything from how much they allocated to these stocks to the validity of academic research showing that small caps outperform large-caps.

Well, that’s essentially what has been happening with hedge funds for years.

Most hedge funds above $1 billion in assets don’t share their data to the commercial databases — it’s voluntary after all. But without these funds, hedge fund performance looks dismal. Once these hedge funds are added to the data set, average performance rises by more than two percentage points.


Academics, institutional investors, advisors, and others rely on return and other data to determine correlations between strategies, how much money they should optimally put into hedge funds broadly and into different strategies, and the role they can play in a portfolio. Using benchmarks to compare managers to peers, broad market indices, and judge an allocator’s ability to select top performing funds also is an ingrained practice in every investment sector.

“The hedge fund space has always been challenging because the data academics have used is sourced from a handful of different commercial data providers and we know there are serious problems there,” said Christian Lundblad, a finance professor at University of North Carolina Kenan-Flagler Business School and one of the authors of a paper published by Kenan-Flagler’s Institute for Private Capital in February that explores the impact of the missing data.

“So we wanted to just ask and answer a really simple question which is ‘what is the universe of institutional quality hedge funds?’” Once that is known and the data collected and analyzed, Lundblad says other questions can be addressed, such as “‘what are the real return properties for that space?’ ‘And what does that mean for diversification opportunities for big institutional investors?’ ‘What does that mean for risk-adjusted performance and where alphas might be hiding?’ and ‘what does that mean for benchmarks?’

“We know important funds are missing, we know some funds are included that aren’t real funds,” added the finance professor. As for ‘real funds,’ the paper notes that commercial databases include “many funds that are not ‘real’ primary hedge funds. On average these are small and include funds-of-one, managed accounts, redundant share classes, feeder funds and even private asset draw-down funds.”

The limitations of existing benchmarks are well-known, but the study now has proved and quantified the problem.

The research finds that commercial databases and their indices don’t include historical return information on better performing institutional quality funds. The research out of UNC Kenan-Flagler found that 152 managers with at least $1 billion are not available in databases. In a footnote in the paper, the authors say, “We use the same set of commercial databases as Barth et al (2023).” That study used data from HFR, EurekaHedge, BarclayHedge, Preqin, and Bloomberg databases.

Specifically, the PivotalPath Composite index, which includes information on 200 hedge funds above $1 billion that are not available in any commercial database, outperformed a composite of these databases by 2.4 percent on an annualized basis, according to the study’s three authors: Greg Brown, a professor of finance at Kenan-Flagler and research director of IPC, Lundblad, and Wiliam Volckmann, research associate at IPC.

“…These hidden hedge funds had higher alphas, lower exposure to market risk factors, and less flow sensitivity to performance,” according to the study. The funds over $1 billion in assets that are exclusive to PivotalPath returned 9.5 percent.

The researchers also found that the 2.4 percent outperformance consists entirely of alpha relative to an equity factor.

If hedge funds are performing far better than most investors — and their boards of directors — believe then they may have far less in hedge funds than they should once a fuller universe of funds is analyzed.

“When everyone thinks the performance of hedge funds is far less than in reality, it’s hard to make the case that allocators should invest. When the HFR 10-year returns of 3.5 to 4 percent are plugged into any optimizer or asset allocation model, no matter what you do, basically the argument is not to invest in hedge funds,” said Jon Caplis, CEO of PivotalPath, which only shares information with institutional investors. “Why bother? There’s too much risk, a lack of transparency, high fees, and at the end of the day, it’s not worth it.” Caplis started an index working group at the end of last year that includes some of the world’s largest institutional investors and hedge fund managers (PivotalPath will make its indices available to hedge fund managers for free ).

One sign of interest in the topic was a standing room only breakfast the Index Working Group held last week in New York, said Michelle Noyes, head of the Americas for the Alternative Investment Management Association.

“It’s not a new observation that [benchmarks] are flawed,” said Noyes. But having a different data approach will allow AIMA to better represent the industry to policy makers and others. “We want to be able to express the reality that investors experience when they use hedge funds in their portfolio. If you are using a number that excludes a vast majority of the industry’s AUM that’s not going to be the case,” she added.

Lilly Knight, co-head of investment management and member of the Investment Committee at K2 Advisors, said “it just feels like constant droning of how bad the industry is when we have clients that are really pleased with the outcomes they’ve had. There’s this endless battle… I have to defend it all the time, but clients are happy, so which is it?”

Knight added that it’s an important time to address the issue of faulty benchmarks. Hedge funds are once again operating in a normal interest rate environment with dispersion and investors want portfolio protection. “This could be a really great time,” but not if the same numbers are fed into an asset allocation, she said. “We’re at a time when we really need people to understand what hedge funds do,” she said.

Without good benchmarks, investors and managers are operating in a vacuum.

Doug Haynes, former president of Point72 who is launching a new hedge fund called Norias in July, said he wants to make sure he’s constantly improving his fund’s operating model. Without benchmarks that reflect the entire industry, it’s hard to do that.

Haynes, whose new hedge fund combines fundamental research with systematic execution, says a robust benchmark is “a source of intelligence” that can point managers to areas that are more likely to yield return. Understanding Norias’s performance relative to older models, such as multi-manager and single manager, can help him spot areas of advantage and disadvantage.

“What history shows in other industries is that if you benchmark, you often redefine your view of how good, good can be,” said Haynes, who is managing partner of Norias Research Group.

The hole in the data raises a huge issue about decades of academic findings on hedge funds. As Lundblad says, “We just know that the data we’ve been using is kind of wrong, or at least incomplete and biased in certain ways. And that just means that a lot of academic research is imperfect.”