Hedge fund hold ’em☆
Introduction
Studies have generally cast managers’ personal pursuits as sources of value destruction. Whether it is distraction arising from executives’ excessive golfing (Biggerstaff et al., 2017), fund managers’ romantic relationships (Lu et al., 2016), or managers’ hobbies, purchases, and indiscretions correlated with excessive risk-taking and fraud (Cain and McKeon, 2016; Brown et al., 2018; Griffin et al., 2019), the evidence suggests personal pursuits are largely value destroying for shareholders and investors.1 Yet, at odds with this, at least in the investment management context, is proficiency in one particular personal pursuit: poker playing. Poker skills are widely believed to be correlated with success in investment and trading.2 David Einhorn, a fund manager, likens poker to investing, stating both “are games of incomplete information…. You have a certain set of facts and you are looking for situations where you have an edge…” Another fund manager, Steve Cohen attributes poker as “the biggest determinant in [his] learning to take risks.”3 On the other hand, anecdotally, the distraction and shirking noted in the literature regarding personal pursuits in general has also been associated with card-playing. James Cayne, CEO of Bear Stearns, was characterized as being “asleep at the wheel” during the crisis that led to the firm’s demise, as he joined a “crucial call late, because he was busy playing bridge …”4 In this study, we try to resolve this tension of whether poker, as a personal pursuit, is a distraction for hedge fund managers, or whether success in poker is correlated with success in investing.
More so than other games like chess or bridge, poker has been singled out by both practitioners and academics as being subject to similar vicissitudes due to luck as investing in the markets. For example, Kevin Zollman, a game theorist and Professor of Philosophy, echoes David Einhorn’s sentiment, contending that poker is a “game of incomplete information,” and thus more akin to the markets than a game like chess, where information is complete, or even bridge, where much of the uncertainty can be resolved through logical deduction and “educated guesses.”5 Thus, we focus on poker, rather than some other games, and we study two questions: (1) Is there a correlation between skill in poker and skill in hedge fund management; and (2) How do investors respond to visible evidence of skill in poker in fund managers?
We attempt to empirically answer these two questions by looking at the poker tournament winnings of hedge fund managers, obtained from Global Poker Index’s Hendon Mob website. We examine whether fund managers who have won cash prizes (i.e., have “cashed”) in poker tournaments perform better in their funds. We find that hedge fund managers who have “cashed” in at least one poker tournament significantly outperform managers who have not in the Hendon Mob database,6 with monthly return and alpha measures that are about 10 bps higher than their non-poker playing peers. Furthermore, we find managers cashing in tournaments with more entrants, larger buy-ins, and ultimately those winning higher cash prizes exhibit even better fund performance than the overall sample of poker tournament winning managers. These results are robust to using four alternative measures of skill proposed in and Kacperczyk and Seru (2007), Kacperczyk et al. (2008), Sun et al. (2012), and Berk and Binsbergen (2015).
While these results are suggestive of poker skills being correlated with fund management skills, we recognize it is possible that reverse causality may be driving our results: fund managers who perform well in their funds may have more money with which to enter more and larger tournaments. To address this, we conduct two additional analyses. First, we consider fund managers who have won poker tournaments prior to starting their funds and find that they also outperform other similar types of funds started at the same time. Second, although we only observe cashes, or situations in which managers place in approximately the top 10% of the tournament entrants who are paid cash prizes, there is still a distribution of how our managers finish in this top 10%. If reverse causality is driving our results, and managers are simply playing in more tournaments after a run of good performance in their funds, and cashing more often simply because they are playing in more tournaments and getting lucky once in a while, there should be no correlation between fund performance and where they finish, conditional on them cashing.7 We test this empirically, separating manager cashes based on their percentile finish in the tournament into above median and below median percentile finish and exploring the manager performance across these two groups. We find that managers who cash in positions better than the median cashing position (which, in our sample is the 5.9th percentile) perform significantly better than managers who cash in positions worse than the median cash. These results are unlikely to simply be a function of managers entering more tournaments after doing well and cashing more often due to luck. Rather, these results are consistent with skill in poker being correlated with skill in fund management.8
We next explore why poker tournament winning fund managers outperform. We focus on two transferable skills that managers who have done well in poker tournaments may exhibit in their investing. Extant studies show that winning poker players are more patient and less susceptible to behavioral biases such as the disposition effect. For example, Siler (2010) shows that “patience is rewarded” among winning poker players, and these players tend to avoid “overweight[ing] frequent small gains vis-à-vis occasional large losses.” We argue these traits may help their investing performance, with increased patience manifesting in lower turnover and lower susceptibility to the disposition effect manifested in a lower ratio of gains realized to losses realized. These channels would also be consistent with findings in the financial literature suggesting lower turnover and less susceptibility to the disposition effect are beneficial for investment performance (e.g., Odean, 1998; Barber and Odean, 2000). Using holdings level analysis, we find that poker winner managers are less susceptible to the disposition effect than non-poker winner managers. Additionally, we also find that poker winner managers are more patient, demonstrating significantly lower turnover.
We find that fund managers experience an economically and statistically significant increase in net inflows after cashing in a poker tournament compared to matched peers.9 Their average monthly inflows in the year following the tournament win are 1.2%–1.8% higher than average net inflows to their matched peers who do not experience poker tournament success. Investors appear to be responsive to poker wins, deploying additional capital to managers who win poker tournaments. Whether this is due to investors proactively considering fund manager’s poker skills in their investments, or whether this is due to managers themselves incorporating poker skills into marketing opportunities, such as interviews,10 or networking opportunities at the poker tables, there appears to be a clear link between success in poker tournaments and subsequent fund flows. Examining subsamples, we find that the documented increases in net inflows are concentrated in subsamples of managers who (1) have media coverage of their poker playing (1.9%–2.0% increase in monthly net inflows), (2) win cash prizes higher than the median tournament win in our sample, i.e., place better in tournaments, win in larger tournaments, or cash in tournaments with higher buy-ins (1.5%–1.8% increase in monthly net inflows),11 and (3) who experience poker tournament wins in two higher profile tournament series, the World Series of Poker (WSOP) or the World Poker Tour (WPT), which are more likely to be televised and thus recognized by the general public (1.9%–3.5% increase in monthly net inflows). These subsamples all rely on certain types of poker tournament wins as being more visible (media coverage, larger wins, higher profile tournaments) or providing a stronger signal of poker skill (larger wins, higher profile tournaments) and our results are consistent with more visible and more difficult tournament wins being rewarded with higher flows from investors.
Finally, a natural question arising from these results is how do these hedge fund investors fare after poker tournament cashes? To answer this, we use a matched sample to examine how fund performance is affected by poker tournament cashes. Our results suggest that performance, measured either as raw returns or Fung and Hsieh (2004) seven-factor alphas, decreases significantly after cashes, and such funds underperform their matched peers to the tune of 40–70 bps a month. This decrease in performance, along with the increased net inflows to these funds suggest that this may be a function of the decreasing returns to scale highlighted in Berk and Green (2004). Given the magnitude of the decline in performance, we also entertain the possibility that distraction may be driving some of the relatively poor fund performance after tournament wins. We explore further by looking at the link between underperformance and net inflows, as well as measures of fund activity, such as active share and index holdings after tournament wins.
We examine how performance changes for funds that obtain above median levels of increases in net inflows after tournament cashes compared to funds that obtain below median levels of increases in net inflows. Consistent with decreasing returns to scale, the decrease in performance is concentrated in funds that are receiving the highest net inflows.12 We also find that the R-squares of fund returns with respect to standard hedge fund return factors increase significantly after poker cashes. Examining holdings, we also find funds increase their holdings in index constituents and decrease their Cremers and Petajisto (2009) active share measure. Finally, we also find that funds with high inflows exhibit the highest incidence of increased R-squares, higher index constituent holdings, and lower active share measures. Together, these results suggest that managers who experience higher net flows after a tournament win deploy their capital in non-alpha generating investments, such as index constituents and known factors, reducing fund performance going forward.
Together, these findings suggest that decreasing returns to scale are likely to be responsible for at least some of the deteriorating fund performance after tournament wins. While overall distraction seems unlikely to be a significant contributor, given the tight link between increased flows and underperformance, it is possible that distractions arising from fund marketing in conjunction with tournament success, such as networking with other players at the poker tables, or spending a lot of time giving interviews, may also contribute to the poor performance, as well as account for the tight link between underperformance and increased flows.13
Section snippets
Literature review
Our study contributes to several areas of the literature. Most directly, our study adds to the growing literature on using personal characteristics to gain additional information on manager performance. While studies examine both corporate and fund managers, our study is most closely linked to those examining managers’ personal pursuits. For example, Biggerstaff et al. (2017) examine CEO golfing to measure leisure time consumption and find evidence consistent with shirking behavior. To our
Sample construction
We hand-collect data on hedge fund managers poker wins from the Hendon Mob website, the largest live poker database.15
Are skill in poker and hedge fund management correlated?
While the summary statistics provide some suggestive evidence that poker skill is correlated with skills for running a hedge fund, in this section, we formally analyze this using regression analysis. We estimate multivariate logistic regressions to study the risk and performance characteristics of funds with poker-winning managers. We present the results in Table 2. We run cross-sectional regressions where the dependent variable is an indicator variable for the fund manager having won a poker
Fund flows after managers win a poker tournament
We next examine investors’ response to fund managers’ tournament wins. If investors believe that poker skill translates into skill in hedge fund management (as noted anecdotally and documented in the evidence above), they may increase allocations to hedge funds when fund managers win a poker tournament. To analyze this question, we use a matched sample approach to account for possible variations in the fund flows that are unrelated to our event of interest. We create two matched samples. For
Alternative explanations
In this section, we consider several alternative explanations for our results. There are several alternative, intuitive explanations that may be driving some of the links between poker and hedge fund management. First, we may expect hedge fund managers who play poker to have a higher tolerance for risk, which in turn may lead to better performance due to a positive risk-return tradeoff. We find no evidence for this as funds managed by poker-playing managers do not exhibit higher return risk. On
Conclusion
Our findings suggest that skilled poker players are, on average, better fund managers. Additionally, investors seem to be responsive to managers’ poker successes and increase investment in the funds of managers who demonstrate poker skill by cashing in a poker tournament. This may be the result of investors being cognizant of the link between poker skill and fund performance, or it might be the result of managers using their poker success as a marketing opportunity by networking at the poker
References (38)
- et al.
Measuring skill in the mutual fund industry
J. Financ. Econ.
(2015) - et al.
An econometric model of serial correlation and illiquidity in hedge fund returns
J. Financ. Econ.
(2004) - et al.
IQ, trading behavior, and performance
J. Financ. Econ.
(2012) - et al.
Limited attention, marital events and hedge funds
J. Financ. Econ.
(2016) - et al.
Pilot CEOs and corporate innovation
J. Financ. Econ.
(2017) The Impact of Venture Capital Screening
(2020)- et al.
Information Sharing through Informal Interactions: Evidence from the Golf Course
(2018) - et al.
Trading is hazardous to your wealth: the common stock investment performance of individual investors
J. Finance
(2000) - et al.
Mutual fund flows and performance in rational markets
J. Polit. Econ.
(2004) - et al.
FORE! an Analysis of CEO Shirking
(2017)
Fooling some of the people all of the time: the inefficient performance and persistence of commodity trading advisors
Rev. Financ. Stud.
Sensation seeking and hedge funds
J. Finance
On persistence in mutual fund performance
J. Finance
CEO personal risk-taking and corporate policies
J. Financ. Quant. Anal.
A tangled tale of training and talent: PhDs in institutional asset management
Manag. Sci.
Wall Street and the Housing Bubble
Are some mutual fund managers better than others? Cross-sectional patterns in behavior and performance
J. Finance
Learning about mutual fund managers
J. Finance
How active is your fund manager? A new measure that predicts performance
Rev. Financ. Stud.
Cited by (2)
Running a mutual fund: Performance and trading behavior of runner managers
2022, Journal of Empirical FinanceCitation Excerpt :Cohen et al. (2009) find that managers take larger bets on firms they are connected to through their social networks and earn significantly higher returns. More recently, Lu et al. (2020) show that hedge fund managers who do well in poker have a higher level of performance. We add to this literature by exploring the differences in the performance and style of a specific group of fund managers, namely long-distance runners.
- ☆
We have benefitted from helpful comments and suggestions from Brandon Adams, Vikas Agarwal, Nick Bollen, Aaron Brown, Rohit Chopra, Philip Howard, Russell Jame, Juha Joenväärä, Marie Lambert, Kevin Mullally, Adam Reed, Tom Shohfi, Reid Walker, and conference and seminar participants at the University of Central Florida, the Universidade Nova, the University of Virginia, Virginia Commonwealth University, the Magnolia Finance Conference, the 2019 FMA Conference, and the 12th Paris Hedge Fund Research Conference. We are especially thankful to Hamilton Bailey, Parker Steed, Will Stewart, and Luqi (Emma) Xu for excellent research assistance.