Generating alpha is what active fund managers are paid to do. The problem is, when a manager succeeds, more people are persuaded to invest. And the larger the assets under management, the harder it becomes to continue generating alpha. LARRY SWEDROE has been looking at the latest evidence.
Hedge fund assets under management (AUM) have increased to more than $5 trillion, up from approximately $300 billion 25 years ago. Given the research findings that there are diseconomies of scale in mutual funds (driven by increased transaction costs and/or less active share), an interesting question is: Has the dramatic increase in AUM negatively impacted hedge fund returns?
My September 24, 2021, TEBI article, Hedge Funds: The End of an Era?, presented the evidence demonstrating that the increase in AUM was not justified by performance, which had deteriorated over time. The deterioration had increased sharply post 2009, and alphas had been negative since.It also presented the findings from research demonstrating that there were no good predictors allowing investors to identify future alpha-generating hedge funds. And finally, it discussed why performance had deteriorated.
In our 2020 book, , Andrew Berkin and I provided four explanatory factors for the deterioration in hedge fund performance—factors that also explain the deterioration in the performance of active managers in general:
Academic research has been converting what was once alpha into beta.
The pool of victims that can be exploited has been shrinking, as individual investors directly own a shrinking share of public equities.
The competition has been getting tougher — today’s active managers are far more skilled and have far greater resources.
The supply of dollars chasing alpha has increased — in their 2021 study, Nicolas Bollen, Juha Joenväärä and Mikko Kauppila found evidence of decreasing returns to scale at the industry level, helping to explain the decline in hedge fund performance.
New evidence
David Forsberg, David Gallagher and Geoffrey Warren, authors of the study Capacity Constraints in Hedge Funds: The Relation Between Fund Performance and Cohort Size, published in thesecond quarter 2022 issue of the Financial Analysts Journal, formed peer “cohorts” using return correlations (with a threshold of 0.75) to determine if there was a negative relationship between fund returns and cohort AUM. If no other funds exceeded the threshold, the fund formed its own cohort (it ran a relatively unique strategy).
Fund performance was measured both relative to the hedge fund sector to which the fund belonged and using the seven-factor model (bond trend-following factor, currency trend-following factor, commodity trend-following factor, bond market factor, credit risk factor, equity factor and size factor) of Fung and Hsieh, which has been shown to explain a significant portion of hedge fund returns and has become the standard tool in the hedge fund literature.
The researchers' database was drawn from Hedge Fund Research and eVestment and covered 7,406 funds over the period January 1997-March 2016. Funds were assigned to one of seven broad hedge fund sectors, including directional traders, fixed income, macro, multi-process, other, relative value and security selection. To eliminate backfill bias, they only included returns after the date a fund was added to the database. Following is a summary of their findings:
The median cohort AUM of $886.0 million was 12 times the median fund AUM of $74.0 million and 4 times the 75th percentile for fund AUM of $221.0 million.
More than 75 percent of fund-quarter observations involved funds with at least one other fund in their cohort.
Aggregate AUM at the investment strategy level was negatively associated with hedge fund performance.
A one-standard-deviation increase in cohort size was associated with a 1.86 percent decrease in annualized alpha under the seven-factor model, and a 44 basis point decrease in annualized returns adjusted for mean sector performance.
A doubling of cohort size would have decreased sector-adjusted returns by 12.4 basis points per annum and decreased factor-adjusted returns by 52.7 basis points per annum.
Consistent with the proposition that greater competition encourages individual funds to accept inflows that occur in response to strong past performance (notwithstanding that greater AUM may dampen future performance), the performance-flow relation was stronger in the presence of close peers.
There was a weaker relation between performance and inflows where funds faced less competition — cohort structure likely influences the propensity to accept assets. With a unique strategy, a fund could reject assets to enable continued alpha generation which it shares via the performance fee. However, with competition, if a fund rejects the assets, a competitor could accept them anyway. By accepting the assets, at least the fund earns its management fee.
Larger fund size was associated with reduced returns for funds that formed their own cohort — diseconomies were more clearly related to fund size for those funds that did not face close competition due to pursuing a relatively unique strategy.
Single cohort funds generated significant seven-factor alpha but did not necessarily outperform their sector in terms of unadjusted returns.
Their findings were robust to the choice of a 0.75 correlation threshold — similar results were found with cohorts formed using correlation thresholds of 0.70, 0.80 and 0.90. However, the magnitude of the negative coefficient on cohort size increased with the correlation threshold (capacity constraints may be better identified under higher correlation thresholds).
An interesting finding was that a majority of the hedge funds exited the sample by the end of the sample period, highlighting the importance of addressing the issue of survivorship bias. Another finding (one that should be expected) was that fund flows were dependent on past fund returns.
Their findings led Forsberg, Gallagher and Warren to conclude: “Our analysis documented a significant negative relation between performance and cohort size, consistent with the existence of capacity constraints for hedge funds primarily operating at the strategy level.”
Investor takeaways
The overwhelming body of evidence demonstrates that active investing, whether in mutual funds or hedge funds, has been a loser’s game. It's a game that is possible to win but whose odds are so poor it’s not prudent to try. One reason for this is that the markets are highly (though not perfectly) efficient. Another is that the evidence demonstrates that success in generating alpha contains the seeds of its own destruction: fund flows, which are negatively related to the ability to generate alpha, follow performance.
These findings provide important insights for investors in both hedge funds and mutual funds. They should consider avoiding investing in actively managed funds that belong to cohorts that have already reached significant size, as they are less likely to perform well in the future. Instead, if you are going to invest in an active strategy, which I don’t recommend, consider investing only if a fund has a unique style and commits to not growing its assets beyond a certain size.
© The Evidence-Based Investor MMXXIV. All rights reserved. Unauthorised use and/ or duplication of this material without express and written permission is strictly prohibited.