Impact of Size and Age on Hedge Fund Performance: evestment Research Division April 2014

Transcription

1 Impact of Size and Age on Hedge Fund Performance: evestment Research Division April 214

2 Table of Contents Methodology... 2 Size and Age Indices: Number of Funds... 3 Size and Age Indices: Cumulative Performance and Distribution of Monthly Returns.. 4 Size and Age Indices: Annual Returns... 5 Size Universe: Number of Funds and Average AUMs.. 6 Size Universe: Performance and Risk Statistics... 7 Size Universe: Performance, Risk Statistics, and Percentile Returns.. 8 Age Universe: Number of Funds and Longevity... 9 Age Universe: Performance and Risk Statistics... 1 Age Universe: Performance, Risk Statistics, and Percentile Returns 11 Cross Sectional Analysis by Size Cross Sectional Analysis by Age. 15 Cross Sectional Analysis with Secondary Size Ranges.. 18 Conclusions... 2 About evestment

3 Methodology The hedge fund screening process for institutional investors and investment consultants encompasses a multitude of qualitative and quantitative components. Among the most frequently searched criteria are assets under management (AUM) and longevity. 1 Impact of Size & Age on Hedge Fund Performance: uses these two variables to provide insight into performance trends for funds of various sizes and ages. The report draws from the records of over 24, investment vehicles available in the evestment alternatives database and parses the data into: (1) size and age indices encompassing all hedge funds for monthly performance information; (2) size and age universes comprised of hedge funds that survived (those with more than 1 monthly returns within yearly periods) for annualized performance and risk statistics. For the size indices and universes, funds were required to have AUM and performance information for analysis. For the age indices and universes, only performance information was necessary. For cross sectional comparisons, AUM and performance data was required. All available fund records were utilized for the indices analysis and the evestment alternatives database retains all fund records irrespective of their active or inactive status to mitigate survivorship bias. Historical returns for funds with a solid track record and for those with poorer performance streaks are also captured because performance since inception is a requisite for funds listing with evestment, which aids in offsetting the effects of backfill, or instant-history, bias in the analysis. Hedge funds were divided into three primary size and three primary age ranges: Size Range Age Range < $25 million < 2 years Medium $25 - $999 million Mid-Age 2-5 years Large $1 billion > 5 years Some investors may prefer examining hedge funds in narrower size bands; therefore, additional secondary size ranges were provided for cross sectional analysis. Hedge funds within each size and age classification were rebalanced annually to more closely coincide with investor timeframes which ordinarily extend beyond monthly and quarterly periods. Fund sizes were determined by their last twelve month (L12M) maximum master AUM, or L12M maximum rolled up share class AUM if the master data was missing, because the typical fund screening process begins several months in advance of an allocation decision. Duplicate AUM data across master and share classes were removed along with highly correlated performance data. Fund ages were determined by their performance length as of the December preceding the start of each annual period. Only unique fund records were kept. If share classes belonging to this unique fund record had AUM and performance information outside of its reported bounds, the additional information was appended to the record to more accurately capture assets and longevity. Funds with less than $1 million in AUM to start each annual period were not analyzed within that period. Additional steps were taken to cleanse the AUM and performance records of funds. A fund s maximum AUM was compared to its average and median AUM to identify potential data input errors. AUM values were corrected when possible, or otherwise removed to preserve a clean sample. Particular attention was placed on larger funds and those with maximum AUMs exceeding 1 times their average and medians. Performance records with zero, very low, or very high standard deviations were also examined, corrected when possible, or removed in an effort to offer a more refined analysis. The $25 million threshold to denote a small fund was chosen to align with a theoretical minimum operating AUM to cover expenses. A recent industry survey found that, without incentive fees or additional capital injections, managers should have somewhere between $25 million and $375 million in AUM to cover the expenses associated with running a fund. 2 The $1 billion threshold to denote a large fund was chosen because funds of this size and above can typically absorb the allocations large institutional investors make without the institution becoming the fund s dominant investor. Larger funds also have the resources to spend more on accommodative infrastructures for larger allocators. The young, mid-age and tenured thresholds were chosen from experience in the management of an alternatives database with managers and investors communicating their perceptions of age. The 5 year threshold to denote a tenured fund is very close to the 5 th percentile value of fund age. 1 Based on data from evestment Advantage, a product which offers a view into how investment consultants and institutional investors use evestment Analytics to evaluate managers for investment mandates. 2 According to Citi Prime Finance s 212 Hedge Fund Business Expense Survey 2

4 Size and Age Indices: Number of Funds Size and age indices were constructed using the equal weighted average of the performance returns of funds within their respective groups during each month. Any fund reporting a monthly performance figure was included within its respective index for that month, provided the fund met the criteria outlined in the methodology. The compositions of the size and age groups were rebalanced annually. The number of small funds reporting performance at the start of each year rose from 1,396 in January 23 to a peak of 2,431 in January 27. funds experienced declines in the ensuing years to a trough of 1,265 in January 213, a decline of % from their high. The number of medium sized funds reached its apex in January 28, at 564, then oscillated for a several periods before settling at 488 in January 213, a decline of % from the peak. Large funds also experienced an oscillation similar to the medium sized funds, but parted ways in January 213 as the constituent count reached 262 funds, surpassing a prior high of 251 set in January 28. A gradual decline has befallen the smallest hedge funds over the analysis period while the proportion of medium and large funds comprising the size group has ticked upward with the exception of a drop in January 21 coinciding with an annual rebalancing period. As of December 213, medium funds comprised 25.94% of the size group and large 14.19%, their heftiest shares to date. The hedge fund industry has matured over the years. The number of young funds reporting hit a high of 2,273 in December 26, mid-age funds 1,829 in January 28, and tenured funds 2,213 in January 212. The highest proportion of young funds (48.63%) dates back to December 23, mid-age funds (34.8) less distantly to January 28 and tenured funds (54.49%) most recently to January ,5 3, 2,5 2, 1,5 1, 5 6, 5, 4, 3, 2, 1, Figure 1: Stacked chart of monthly index fund counts and monthly index compositions by fund size with annual rebalancing from January 23 to December 213 Figure 2: Stacked chart of monthly index fund counts and monthly index compositions by fund age with annual rebalancing from January 23 to December

5 Size and Age Indices: Cumulative Performance and Distribution of Monthly Returns Among the three sizes indices, the small index had the highest cumulative return from January 23 to December 213, at 122.1%. The medium index came in second at 92.86% and the large index third at 82.32%. From January 29 through December 213, however, the cumulative return of the small fund index was the lowest of the three, while that of the medium fund index was highest. During this 5 year period, the small fund index returned a cumulative 42.5, the medium fund index 49.29%, and the large fund index 44.. Relative to the medium and large indices, the monthly returns of the small fund index most closely resemble a standard normal distribution. The frequency of monthly returns falling in the 2% to 4% range was higher for the small versus medium and large index. Among the three age indices, the young had the highest cumulative return from January 23 to December 213, at 21.56%. The mid-age index came in second at % followed by the tenured index at %. The young fund index also outperformed on a cumulative basis during the 5 year period ending December 213, but the tenured index did slightly better than the mid-age index. During the past 5 years, the young index returned a cumulative 64.93%, the mid-age index 47.21%, and the tenured index 48.29%. The distribution of monthly returns for the young, mid-age, and tenured indices were very similar based on their skew and kurtosis values, but the performance of the young index was usually higher within the analyzed ranges. In the 61 months where the young, mid-age, and tenured indices overlapped within the to 2% range, the young index outperformed the mid-age in 85.2 and the large in 73.77% of these monthly periods. Figure 3: Cumulative index performance by size and age from January 23 to December Figure 4: Distribution of monthly index returns by size and age from January 23 to December Skew Kurtosis Skew Kurtosis 6 5 Medium Mid-Age Large

6 Size and Age Indices: Annual Returns Prior to 29, the small fund index consistently outperformed the medium and large indices in annual returns, by as much as 7.43 points (versus large in 23) to as little as.77 points (versus medium in 25). The small index even outperformed the medium and large indices during the global financial crisis in 28, perhaps due to liquidity advantages, but also possibly due to the representation of managed futures strategies among smaller funds. All three indices were down in the months of September and October 28, but the medium and large declined by more. In September 28, the small index fell -5.39%, the medium -6.91%, and large -6.44%; in October 28, the small index fell -5.93%, the medium -7.74%, and large -7.32%. We revisit the data in figure 1 and mention that there were 22.19% fewer small funds reporting returns in December 28 versus the number that started reporting in January 28, compared to 12.41% fewer medium funds and 1.76% fewer large funds, increasing the likelihood that many small funds lowest returns were not captured in the indices. Figure 6: Annual index returns by size and age from January 23 to December % 13.42% 13.23% % 6.21% 9.13% 11.68% % % 9.32% 6.68% 12.82% 9.8% 9.84% % % 13.6% 12.32% % 9.51% % % % -2.42% -7.58% % % 22.12% 2.53% 23.9% % % 9.1% 7.78% % % -2.58% -2.42% % % 5.87% 6.24% 8.91% % % 8.73% 7.4% 9.66% 8.31% 9.38% Figure 5: Annual index returns by size and age from January 23 to December The dominant annual performance streak of By annual returns, the young index has small funds, however, ceased in 29 and has consistently outperformed the mid-age and been inconsistent since, suggesting that smaller tenured indices, while the mid-age funds have lost the performance momentum outperformed the tenured in 6 of 11 years. they enjoyed in earlier years relative to their er hedge funds may outperform for many medium and large peers. reasons. Delaying a launch due to difficult In 29, both the medium index (22.12%) and large index (2.53%) outperformed the small index (18.54%). The small index led again in market conditions may play a role, as well as a flood of funds targeting emerging opportunities. Willingness to accept more risk may also play a 21, but was third in 211, 2.7 points behind meaningful role for some as a substantial the medium index and 2.86 points behind the portion of earnings will likely come from large. In 212 it edged out the medium index slightly, by 12 bps, yet trailed the large by 25 bps. While finishing first in 213, the small index only led by a narrow 17 bps margin. incentive fees needed to continue operations and develop strong comparable track records. Lastly, employees of already successful, larger hedge funds firms may start their own funds, bringing with them valuable experience and a proven successful skillset. 5

7 Size Universe: Number of Funds and Average AUMs The following information pertaining to the size universe uses a subset of the data from the size indices only funds with more than 1 months of returns in an annual period because annualizing performance and risk statistics for funds with a limited number of returns becomes a speculative undertaking. The number of small hedge funds with more than 1 months of performance within a year rose from 1,324 in 23 to a peak of 2,11 in 27. Beginning in 28 and continuing throughout ensuing periods, the number of small funds began to decline, hitting a trough of 1,8 in 213 just as the number of large funds reached a new peak of 231 in the same period. Medium funds hit a high of 52 in 29 and were down to 429 as of 213. Average small fund AUM declined by 5.52% in 21 from the prior year, perhaps because robust performance gains in 29 lifted the bigger small funds into the medium category to start 21. However, average AUM has risen by 23.9 since. Money seems to have flowed primarily to the largest hedge funds, even during the global financial crisis, because average AUM has not declined in a single year since at least 23, while the average AUMs of small and medium sized funds have fluctuated. The best periods for asset growth for the average large fund, in absolute terms, were 211 (YOY average AUM increase of $4.8 million) and 212 (YOY average AUM increase of $447.7). Since 23, average large fund AUM has risen by 146.8, compared to an average small rise of 39.77% and average medium rise of only 5.74%. The largest YOY percentage change in average AUM for medium funds was 2.41%, from 211 to 212. During this period, small AUM expanded by 16.9% and large by 12.82%. Between the 2 th and 8 th percentiles, the AUM values of medium funds have remained relatively stable compared to the small and large groups. The 8 th percentile AUM value for small and large funds peaked most recently in 213 but for medium sized funds the high was in 29. 2,2 2, 1,8 1,6 1,4 1,2 1, Figure 7: Annual fund count by size with annual rebalancing for funds with >1 months of performance in respective year Figure 8: Average annual AUM by size for funds with >1 months of performance in respective year in USD millions Sma ll $ 49.4 $ 53.1 $ 54.9 $ 54.8 $ 55.7 $ 58.3 $ 58.9 $ 55.7 $ 59.2 $ 68.8 $ 69. Me dium $ $ $ 481. $ 48.9 $ $ $ 56. $ $ 495. $ 56.9 $ 52.4 La rge $ 1,628.2 $ 1,825.4 $ 2,16.5 $ 2,269.3 $ 2,565. $ 2,782.9 $ 3,2.2 $ 3,92.1 $ 3,492.9 $ 3,94.7 $ 4,19.2 Figure 9: AUM percentile ranks by size for funds with >1 months of performance in respective year in USD millions $14 $12 $1 $8 $7 Medium $4,5 $4, $3,5 Large $8 $6 $3, $6 $5 $2,5 $4 $2 $4 $2, $1,5 $ $3 $1, 65th - 8th 5th - 65th 35th - 5th 2th - 35th 65th - 8th 5th - 65th 35th - 5th 2th - 35th 65th - 8th 5th - 65th 35th - 5th 2th - 35th 6

8 Size Universe: Performance and Risk Statistics The disparity in the average annual returns of small funds versus medium and large funds was highest in 23. The average small fund with more than 1 months of performance had an annual return of 22.54%, the medium fund 13.92%, and the large fund By median returns, the largest disparity was in 28, when small funds outperformed medium funds by 5.88 points and large fund by 8.42 points. Average annual returns of medium funds were higher than large funds in 6 of the past 11 years; however, large have outperformed medium in each of the past three years, albeit by narrow margins of 13 bps, 3 bps, and 9 bps. While small funds outperformed large funds in 9 of the past 11 years on an average basis, this figure drops to only 7 of 11 on a median basis. Similarly, the outperformance of small over medium funds drops from 9 to 8 years. Large funds exhibited the lowest average and median standard deviations on an annualized basis in every year except in 24 (medium average was lower by 3bps) and 21 (medium fund median was lower by 33 bps). funds exhibited the highest average and median standard deviations in every year except 213, coming in second to medium funds (medium average was higher by 32 bps). On an annualized risk-adjusted performance basis, as measured by the Sharpe ratio using the 1-year Treasury constant maturity rate in each annual period as the riskfree rate, the average small fund outperformed the average medium fund in 7 of the past 11 years. The four years in which the average medium fund outperformed were all sequential and post global financial crisis (29, 21, 211, and 212) % 16% 14% 12% 8% 6% 4% 2% Figure 1: Average and median annual returns by size for funds with >1 months of performance in respective year Return Average Figure 11: Average and median annualized st. dev. by size for funds with >1 months of performance in respective year St. Dev. Average Figure 12: Average and median annualized Sharpe ratios by size for funds with >1 months of performance in respective year 3 Sharpe Average 2.5 Sharpe Median % 16% 14% 12% 8% 6% 4% 2% Return Median St. Dev. Median 7

9 Size Universe: Performance, Risk Statistics, and Percentile Returns In 23, 24, and most recently in 213, the average and median large fund had a higher Sharpe ratio relative to both its small and medium peers. Outlier funds with positive performance and very low volatilities provided an annualized Sharpe ratio of.24 for the average medium fund in 211. Some of the medium funds with positive Sharpe ratios in 211 also employed less conventional strategies: securitized credit/asset based lending, distressed, life insurance policies, short-term fixed income, volatility, and even energy royalties strategies. From 23 to 27, at least 8 of all small, medium, and large hedge funds with more than 1 months of returns reported positive annualized performances based on percentile rankings. The 2 th percentile values for small funds were also lower than that of large funds in every year since at least 23. It is important to keep in mind percentile information is based solely on funds with more than 1 month of returns within an annual period. Figure 13: Average and median risk and return statistics by size for funds with >1 months of performance in respective year Average Annual Returns Annualized St. Dev. Annualized Sharpe 22.54% 1.71% 1.59% % % 1.54% -4.72% 6.71% 1.68% Medium 13.92% 6.26% 9.93% 11.83% 1.54% % 25.84% 9.11% -2.43% % Large % 7.21% 1.16% % 22.94% 8.16% -2.31% 6.38% 7.98% 1.41% % 9.97% 11.13% 19.16% 14.21% 12.43% 12.93% 1.41% 9.28% Medium 6.71% 6.48% 7.61% 7.86% 9.32% % 1.37% 8.74% 9.6 Large 5.43% 6.51% 7.1% 7.3% 9.17% 16.78% 11.74% 9.14% 9.23% 7.7% 5.99% Medium Large Median Annual Returns Annualized St. Dev. Annualized Sharpe 15.81% 8.7% % -9.78% 12.67% 8.47% -3.44% 5.61% 8.81% Medium 1.99% % 11.1% % 15.84% 7.78% -1.39% % Large % 1.88% % 7.89% % 7.58% % % % 1.28% 9.84% 1.13% Medium % 6.58% 6.27% % 9.32% 7.41% 8.23% % Large 4.6% 4.31% % 7.52% 13.63% 9.19% % 5.59% 5.9% Medium Large Figure 14: Percentile ranks of annual returns by size for funds with >1 months of performance in respective year Medium -3-4 Large 65th - 8th 5th - 65th 35th - 5th 2th - 35th 65th - 8th 5th - 65th 35th - 5th 2th - 35th 65th - 8th 5th - 65th 35th - 5th 2th - 35th 8

10 Age Universe: Number of Funds and Longevity The following information pertaining to the age universe uses a subset of the data from the age indices only funds with more than 1 months of returns in an annual period because annualizing performance and risk statistics for funds with a limited number of returns becomes a speculative undertaking. One month minimum longevity values for young funds represent funds launching in December prior to the start of each annual period. The number of young hedge funds with more than 1 months of performance within an annual period declined from a peak of 1,61 in 27 to 492 in 213. Mid-age funds fell from a high of 1,552 to 897 in the same period but tenured funds have risen, though reaching a peak of 1,918 in 211. Investors have possibly become more hesitant to allocate to funds without longer track records following the global financial crisis. All three age groups saw their numbers rise from 23 to 27. In 28, 9.88% fewer young funds, with more than 1 months of returns, reported from the prior year and in 29, 3.16% fewer mid-age funds reported; the declines have continued in each year since. funds with more than 1 months of returns expanded up through 211 but declined by 4.9 and 8.23% in the ensuing periods. Since the beginning period of the analysis, the number of young funds reporting with more than 1 months of returns in an annual period has declined by % and mid-age by -4.27%. The number of tenured funds, however, grew by 99.17%. While the longevity data for young and mid-age funds is confined based on the parameters of the analysis, the average, median, and maximum reporting lengths of tenured funds have remained steady or continued rising in all periods since at least 23, showing a level of resilience in this reporting segment of the industry. 2,2 2, 1,8 1,6 1,4 1,2 1, Figure 15: Annual fund count by age with annual rebalancing for funds with >1 months of performance in respective year Figure 16: Average and median longevity by age for funds with >1 months of performance in respective year Average M edian M id-age M id-age Figure 17: Longevity data for funds with >1 months of performance in respective year. Top point on high-low line represents the maximum number of monthly returns reported by a single fund and bottom point represents minimum; top line on squares represents average number of monthly returns by group and bottom lines represent the median; *reading of average and median lines reversed Mid-Age

11 Age Universe: Performance and Risk Statistics Annual returns of the average young fund, with more than 1 months of performance within a year, were higher than those of the average mid-age and tenured fund in every year since at least 23. Annual returns of the median young fund were higher in every year except 23 (lower than both) and 213 (lower than tenured funds by 23 bps). For the average young fund, the largest outperformance versus the average mid-age fund was in 29 (5. points) and versus the average tenured in 28 (7.43 points). The figure versus tenured funds rises to 1.17 points in 28 when comparing median annual returns. Mid-age and tenured funds have largely gone back and forth in outperformance over one another in terms of average and median annual returns, albeit mid-age funds have one additional year in their favor. Performance volatility generally appears to increase with age. The average mid-age fund had a higher annualized standard deviation than the average young fund in 7 of 11 years, while the average tenured fund exhibited a higher standard deviation than the average mid-age in 9 of 11. Survivorship bias may be responsible for a smoothing of earlier returns, if younger, more volatile funds have lower survival rates than their mid-age and tenured peers. Annualized risk-adjusted performance, as measured by the Sharpe ratio using the 1-year Treasury constant maturity rate in each annual period as the risk-free rate, tends to wane with age. The average tenured fund only outperformed the average mid-age in 29, although the median tenured fund managed to generate a higher Sharpe ratio than the median mid-age fund in 5 of 11 years. The average and median young fund displayed higher Sharpe ratios than both its mid-age and tenured peers throughout the entire timeframe examined % 16% 14% 12% 8% Figure 18: Average and median annual returns by age for funds with >1 months of performance in respective year Returns Average Returns Median Figure 19: Average and median annualized st. dev. by age for funds with >1 months of performance in respective year 16% St. Dev. Average St. Dev. Median 14% 12% 8% 6% Figure 2: Average and median annualized Sharpe ratios by age for funds with >1 months of performance in respective year 2.5 Sharpe Average 2 Sharpe Median

12 Age Universe: Performance, Risk Statistics, and Percentiles The average young fund had positive annualized Sharpe ratios in 28 and 211 two years in which the average hedge fund suffered losses. Responsible for this were outlier funds with positive performance and very low volatilities employing niche strategies: life insurance policies, distressed commercial real estate opportunities, commercial receivables, and volatility arbitrage strategies. Their relative youth may have also saved them from getting entangled in the plummeting global markets of late 28. funds could have refrained from making large directional bets with the markets hitting new highs, and launches could have been designed around the idea that a crisis was looming. Since 23, the 8 th percentile values for the annual returns of young funds were consistently higher than the 8 th percentile values for mid-age funds; the young fund value was also higher in 9 of 11 years when compared to tenured funds. In 29, the 2 th, 35 th, 5 th, 65 th, and 8 th percentile values for tenured funds were all higher than those of mid-age funds. Figure 21: Average and median risk and return statistics by age for funds with >1 months of performance in respective year Average Annual Returns Annualized St. Dev. Annualized Sharpe 23.94% % 15.71% 16.43% -5.98% 26.6% 13.2% % Mid-Age % 9.53% 14.14% 12.72% -1.67% 21.7% 1.27% -2.58% 8.37% 9.94% 22.43% 9.96% 1.21% 12.74% % % 6.68% 1.51% 9.53% % % % 1.58% 11.61% 9.12% 8.38% Mid-Age % 8.99% 8.87% 1.57% 18.9% % % % % 1.37% 18.67% 13.66% % % Mid-Age Median Annual Returns Annualized St. Dev. Annualized Sharpe 14.82% 9.9% 9.54% 11.99% % 16.49% 1.11% -.66% 7.56% 8.43% Mid-Age 16.17% 7.67% 7.66% 11.71% 9.68% % 8.28% -1.31% 7.26% 8.41% % 7.31% % % 15.99% 8.49% -3.18% 5.98% 8.66% 6.94% 7.1% 7.43% 7.82% 8.66% 13.24% 9.83% 8.1% 8.64% 7.1% 6.66% Mid-Age 7.52% % % 14.14% 9.86% 9.2% 9.41% 6.79% 6.78% 8.71% % 7.57% 8.54% 15.34% % Mid-Age Figure 22: Percentile ranks of annual returns by age for funds with >1 months of performance in respective year Mid-Age th - 8th 5th - 65th 35th - 5th 2th - 35th 65th - 8th 5th - 65th 35th - 5th 2th - 35th 65th - 8th 5th - 65th 35th - 5th 2th - 35th 11

13 Cross Sectional Analysis by Size The following cross sectional information pertaining to the size and age universes uses a subset of the data from the size and age indices funds with more than 1 months of returns in an annual period because annualizing performance and risk statistics for funds with a limited number of returns becomes a speculative undertaking. The percentage of small funds which are young has steadily declined since the first year of the analysis, from 36.78% in 23 to 7.14% in 213. Similarly, the percentage of medium young funds also peaked in 23 at 14.5 to an eventual low of 4. in 213. The waning of young funds across all sizes since 23, paired with a consequential increase in the percentage of tenured funds, is interpreted as the maturation of an industry which only several years ago was still primarily funded by funds of hedge funds (FoHFs) and high net worth individuals (HNWI). In the 1 years since 23, the largest allocators to hedge funds have increasingly become institutional investors. The percentage of large funds which are tenured has increased from 61.76% in 23 to 78.3 in 213. For each annual period, the percentage composition of large tenured funds was higher than that of medium tenured and small tenured in their respective size groups. Amassing $1 billion or more in assets is usually a lengthy and formidable undertaking; it is no wonder that relative to its small and medium size peers, the annual universe composition of large tenured funds grew higher with the passing of time. Figure 23: Annual universe composition by size and age for funds with >1 months of performance in respective year; * denotes <8 funds in sample % 34.37% % 34.41% 29.13% % 33.41% 3.91% % 31.94% % 35.96% 35.18% % 36.37% % 35.34% % 54.12% % 28.56% % 26.69% 66.17% Medium % 4.6% 47.73% % % % 36.34% 52.86% % % % 28.16% 63.83% % 23.56% 7.67% % 22.66% 71.68% % 77.39% Large %* % %* 37.33% %* % % % % 24.19% 69.89% % 66.22% %* % %* 17.14% 8.71% % % % 18.18% 78.18% %* 19.91% Medium Large 12

14 Cross Sectional Analysis by Size Among the small size funds, the average young fund has outperformed the average mid-age (tenured) fund in 7 of 11 (6 of 11) years; the best annual outperformance was by 5.12 (6.7) points in 27 (211). The small mid-age fund has trailed the small tenured in 7 of 11 years. Among the medium size funds, the average young fund has underperformed in 8 of 11 (7 of 11) years relative to the average mid-age (tenured) fund; its lowest underperformance was by (-9.63) points in 28 (29). The average medium mid-age fund has led the average medium tenured fund in 6 of 11 years. Among the large size funds, seniority tended to produce winners as the average tenured fund outperformed the average mid-age fund in 7 of 11 years; the largest outperformance was by 6.74 points in 29. Comparisons with large young funds cannot be made for the full analysis period because of sample restrictions. However, it is interesting to note that in 28 one of the few periods with an adequate sample the large young funds were the worst performers by a significant margin. Yet youthfulness, at least among the small and medium size segments, generally exhibited the lowest annualized volatility profile (data displayed on next page). Within the small and medium groups, as a fund increased in age from young to mid-age to tenured, so too (usually), did its annualized volatility. What occurs in the large size segment is less clear because of sample restrictions. Among the small size funds, young funds produced the best risk-adjusted returns, as measured by the Sharpe ratio, relative to the mid-age (tenured) funds in every year except 212 and 213 (213), while the mid-age funds had higher Sharpe ratios than tenured funds in every year except 24 and 21. In the medium and large size segments, however, this youthful dominance was not as pronounced (data displayed on next page). Figure 24: Average annual returns by size and age for funds with >1 months of performance in respective year; * denotes <8 funds in sample % 22.81% 24.43% % 9.94% 1.27% % 8.64% 1.24% % 14.23% % 1.97% 11.43% % % % % % -4.42% -6.8% % 6.47% % 8.1% 12.13% Medium % % % % % 1.67% % % % % % % % 29.58% % 8.33% 9.17% % -.29% % % 6.27% 8. Large %* 14.39% 14.48% %* % %* 5.48% 8.2% % % % 6.8% 11.77% % % %* % %* 7.94% 8.17% % -.14% -2.78% % 7.76% 5.92% %* 5.93% 8.43% Medium Large 13

15 Cross Sectional Analysis by Size Figure 25: Average annualized st. dev. by size and age for funds with >1 months of performance in respective year; * denotes <8 funds in sample Figure 26: Average annualized Sharpe ratio by size and age for funds with >1 months of performance in respective year; * denotes <8 funds in sample % 1.66% % % % 9.54% % 1.76% 11.31% % 19.36% % 14.16% 14.98% % 13.9% 12.77% % 13.89% % 8.98% 11.51% % 9.52% Medium % 7.27% % % 7.39% 7.88% % 6.79% 8.48% % 9.44% 9.29% % 15.77% % 13.31% % 9.2% 1.48% % 1.99% % 6.23% 9.73% % % Large %* % %* % 25 2.* 7.27% 7.7% % 7.33% 6.93% % 8.94% % 17.12% 16.24% %* 12.87% 11.44% * 1.18% % 8.54% 9.57% % 7.23% %* 5.64% 6.4% 18% 16% 14% 12% 8% 2 2 Medium Large Medium Large * * * * * * Medium Large 14

16 Cross Sectional Analysis by Age The following cross sectional information pertaining to the age and size universes uses a subset of the data from the size and age indices funds with more than 1 months of returns in an annual period because annualizing performance and risk statistics for funds with a limited number of returns becomes a speculative undertaking. The percentage of young funds which are small has declined in every year since 24, while in the same period the percentage of young medium sized funds has risen every year except 24 and 211. Among the mid-age group, the percentage compositions of small, medium, and large funds all exhibited ebb and flow patterns during the 11 year analysis. Typically, a year or two of declines were followed by the occasional year or two of increases, likely the result of funds being reclassified from young to mid-age and mid-age to tenured. Mid-age small funds did show a substantial 6.34 point rise from 29 to 21, bringing their percentage composition to just above their the all-time 23 high within the context of this analysis. The trend among the tenured group included a general decline in the percentage composition of small funds offset by increases in medium and large during the past 11 years. Keeping a hedge fund operating for over 5 years certainly increases the prospect of receiving new allocations from investors, but also presents some additional chances to generate performance-based AUM growth. Figure 27: Annual universe composition by age and size for funds with >1 months of performance in respective year; * denotes <8 funds in sample % %* % 5.32%.33%* % 6.38%.4* % 6.57% 1.21% % 1.47% % 7.82% 1.47% % 9.16% 1.22%* % 1.3%.91%* % 9.54% 3.44% % % %* Mid-Age % 13.54% 2.4% % 14.44% 4.26% % % % 17.26% 3.89% % 17.26% 4.71% % 19.41% % 18.76% 5.11% % 14.52% % 3.31% % 7.89% % % % 4.19% % % % 21.27% 9.87% % % 22.99% % 24.94% 13.12% % % % % % 26.19% 13.69% % 28.14% 15.34% Mid-Age 15

17 Cross Sectional Analysis by Age Among the young funds with more than 1 months of returns in an annual period, the average small fund has outperformed the average medium size fund in 8 of 11 years; the best annual outperformance was by points in 28. The young large fund sample is limited, but the available data for 28 shows young small funds outperforming large peers by robust points. In the mid-age segment, the small size story is less pronounced. Mid-age small funds outperformed the midage medium funds in only 6 of 11 years, and the best annual outperformance, 11.6 points, occurred early on in 23. Among the tenured funds, the small funds outperformed both the medium and large funds in 8 of 11 years, but the best outperformance also occurred early on in 23. Further, the small tenured funds underperformed both medium and large funds in the rebounding market period 29. On an annualized basis, young small size funds were more volatile than young medium size funds except, interestingly, in 28 (data displayed on next page). Mid-age small size funds were more volatile than both the medium and large in every year. Mid-age medium funds were less volatile than large funds in 7 of 11 years. In the tenured group, the small size funds generally produced the lowest risk-adjusted returns, as measured by the Sharpe ratio (data displayed on next page). Annualized Sharpe ratios for medium size funds were higher than the small funds in 8 of the past 11 years and for large size funds in 9 of 11. Figure 28: Average annual returns by age and size for funds with >1 months of performance in respective year; * denotes <8 funds in sample % 13.89% 11.9%* % 5.56% 9.62%* % 8.97% 5.22%* % 9.78% 12.58% % 7.6% 1.19% % %* % 11.38% 9.62%* % -2.39% -.67% % % 9.19% 11.8%* Mid-Age % % % % 1.67% 5.48% % % % % % % % 2.63% % 7.94% % -.29% -.14% % 8.54% 7.76% % 6.27% 5.93% % 15.54% 14.48% % 5.92% 8.21% % % % 9.97% % 1.22% 11.77% % % % 24.33% % 8.17% % % % % % % Mid-Age 16

18 Cross Sectional Analysis by Age Figure 29: Average annualized st. dev. by age and size for funds with >1 months of performance in respective year; * denotes <8 funds in sample Figure 3: Average annualized Sharpe ratio by age and size for funds with >1 months of performance in respective year; * denotes <8 funds in sample % %* % 3.87%* % 2.* % 8.42% 6.99% % 9.2% % 21.27% 23.7% % %* % 6.88% 9.6* % % 6.16% % 7.47%* * * * * * * Mid-Age % 6.22% % 7.39% 7.27% % 6.79% 7.33% % 9.44% 9.74% % 15.77% 17.12% % 1.93% 12.87% % 9.2% 1.18% % 9.3% 8.54% % 6.23% 6.36% % % 18% 16% 14% 12% 8% 6% 4% 2% Mid-Age Mid-Age Mid-Age % 6.4% % % % 7.7% % 6.93% % 9.29% 8.94% % % 13.31% 11.44% % 1.48% % 1.99% 9.57% % 9.73% 7.23% % 1.57% 6.4% 18% 16% 14% 12% 8% 6% 4% 2%

19 Cross Sectional Analysis with Secondary Size Ranges In 213, the average tenured fund with $5 to $99.99 million in assets generated the highest annual return, at 13.79%. In 212, it was the average mid-age fund with $25 to $ million in assets by returning 9.59%. And in 211, it was the young fund with $1 to $9.99 million, posting a 3.82% return. In 211, a year in which the average hedge fund was in the red, young funds with $1 to $9.99 million, $1 to $ million, and mid-age funds with $5 to $ million in assets were the only three groups to remain in the black. Among the small fund secondary size ranges, young funds with $1 to $9.99 million and $5 to $99.99 million had higher annual returns than their mid-age, same size counterparts in 8 of the past 11 years. Also posting higher annual returns in 8 annual periods were: (1) young funds with $1 to $9.99 million versus their tenured peers and tenured funds with $1 to $24.99 million and $25 to $49.99 million versus their mid-age peers. Figure 31: Average annual returns by age and size (in USD millions) for funds with >1 months of performance in respective year; * denotes <8 funds in sample A N N U A L R E T U R N S M id-age m-9.99m 28.13% 14.94% 13.48% 14.67% 24.16% % 3.82% 6.53% 2.6 1m-24.99m 19.32% 13.31% 11.11% 15.88% 13.17% % 1.27% -.29% % 25m-49.99m 13.94% 8.91% 14.59% 12.71% 11.27% % % m-99.99m 16.2% % % -5.41% 23.17% 11.19% -2.58% 7.97% 3.7 1m m % 11.94% % % 16.9% 1.24%.46% 3.92% m m % 7.73% 7.69% 1.2% % % 7.81% 9.98% M edium 5m m 11.7%* -8.27%* 1.28% 1.34% 2.7* % 8.76% -3.18%* 5.78% 5.22%* 75m m 17.5* 6.7%* 11.88%* 16.67%* %* 27.34%* 8.4*.94%* 8.* 1.66% Large 1b+ 11.9%* 9.62%* 5.22%* 12.58% 1.19% %* 9.62%* -.67% 9.57% 11.8%* 1m-9.99m 28.3% 1.59% 9.19% 14.33% 12.72% -4.66% 17.14% 1.71% -4.32% 4.34% 11.76% 1m-24.99m % % % 17.49% 11.26% -5.43% 9.48% 11.24% 25m-49.99m % 18.78% 9.74% % 21.24% 11.46% % 8.7% 5m-99.99m 22.13% 13.24% 9.64% 11.12% % 8.68% -2.93% 6.58% 5.67% 1m m % 9.51% 14.11% 1.82% % 22.82% % 7.47% 7.34% 25m m 9.81% 6.81% % 12.39% % 19.94% 7.74% -.14% 9.59% 5.41% M edium 5m m 17.24% % 11.39% 13.34% -1.56% % % 7.18% 75m m % 13.89% % % 24.14% 6.79% % 1.11% Large 1b % % % % % -.14% 7.76% 5.93% 1m-9.99m 26.76% 12.71% 12.91% 11.13% 14.67% % 11.92% % 1.58% 1m-24.99m 34.8% 13.52% 7.17% 13.71% % 22.41% 11.37% -5.92% 9.32% 13.74% 25m-49.99m 25.84% % 9.24% 1.19% % 2.89% 11.47% -4.89% 8.53% 1.59% 5m-99.99m 22.11% 9.99% 9.84% 12.91% 11.38% % 25.13% 1.44% -6.41% 6.53% 13.79% 1m m 16.64% 6.91% % 1.68% -12.8% 21.27% 11.48% -5.82% % 25m m 14.18% 6.54% 1.64% 13.43% 11.19% -15.9% 27.94% 9.58% -3.21% 5.87% 8.36% M edium 5m m 16.26% 7.98% 7.26% % % 31.16% 9.69% -3.43% 6.72% 8.34% 75m m 19.76% 1.1% 9.61% 1.53% 1.7% % 6.48% -2.58% 2.74% 7.14% Large 1b % 8.21% 8.2% 9.97% 11.77% % 24.33% 8.17% -2.78% 5.92% 8.43% Figure 32: Universe composition by age and secondary size for funds with >1 months of performance in respective year; from left to right: young, mid-age, and tenured b+ 75m m 5m m 25m m 1m m 5m-99.99m 25m-49.99m 1m-24.99m 1m-9.99m b+ 75m m 5m m 25m m 1m m 5m-99.99m 25m-49.99m 1m-24.99m 1m-9.99m b+ 75m m 5m m 25m m 1m m 5m-99.99m 25m-49.99m 1m-24.99m 1m-9.99m 18

20 Cross Sectional Analysis with Secondary Size Ranges Among the medium fund secondary size ranges, annual performance was predominately mixed across the three ages. funds with $25 to $ million posted higher annual returns than their mid-age, same size counterparts in 7 of 11 years; all other age comparisons yielded at most 6 years in favor of one fund over another. When comparing the small fund secondary size groups by age, the annualized standard deviation increased, relatively, as funds got older in nearly every instance. The highest number of occasions where the opposite was true is for mid-age funds with $1 to $9.99 million, which had less volatility than their young peers in 4 of 11 years. Survivorship bias may be responsible for a smoothing of earlier returns, if younger, more volatile funds have lower survival rates than their mid-age and tenured peers. Similarly, when comparing the medium and large fund secondary size groups by age, fund volatility also tended to increase as funds got older. The exception was for large tenured funds which had lower volatilities in 6 of 11 years relative to large mid-age funds. The risk-adjusted performance, as measured by the Sharpe ratio using the 1-year Treasury constant maturity rate in each annual period as the risk-free rate, of the medium fund secondary sizes favored the mid-age over the tenured group by a slight margin. Mid-age funds with $25 to $ million had a higher annualized Sharpe ratio in 6 of 11 periods compared to tenured funds, in 7 of 11 periods for those funds with $5 to $ million, and in 6 of 11 periods for those funds with $75 to $ million. Figure 33: Average annualized st. dev. and Sharpe ratios by age and size for funds with >1 months of performance in respective year; * denotes <8 funds in sample S T A N D A R D D E V I A T I O N S H A R P E R A T I O M id-age M id-age m-9.99m 11.39% % 12.14% 15.8% 21.94% 16.87% 1.68% 1.73% % 1m-24.99m % 1.13% 1.69% 17.91% 11.49% 1.69% 9.43% 8.6% 1.17% 25m-49.99m 7.23% 7.53% % % 1.46% 8.48% % 5m-99.99m 6.88% % 8.23% % 9.46% 8.97% 8.43% 9.47% 1m m % 7.72% 7.72% 9.1% 16.86% % 6.33% 7.73% m m 6.2% % 7.44% 8.59% 23.42% 1.42% % % M edium 5m m 4.7* 12.86%* % 5.68%* 18.16% 9.91% %* 7.44% 5.33%* 75m m 9.7* 6.11%* 3.14%* 14.3* 12.72% 11.73%* 13.8* 4.86%* 8.2%* 5.8%* 4.71% Large 1b+ 4.97%* 3.87%* 2.* 6.99% % 1.88%* 9.6* 5.92% %* 1m-9.99m 13.42% 14.42% 12.57% 11.69% 11.96% 21.8% 17.24% 17.17% m-24.99m 12.51% 1.16% 9.16% % 19.41% % 11.97% 9.73% 8.54% 25m-49.99m 1.18% 8.53% 8.43% % 2.7% 14.27% 12.93% 14.18% 9.41% 8.89% 5m-99.99m 7.82% 8.42% 8.22% % % 12.16% % 1m m 7.53% 6.71% 7.98% % 18.53% 12.94% 1.89% % m m 5.93% 6.72% % 9.93% 16.7% 11.72% 8.98% % 6.4 M edium 5m m 7.46% 5.93% 6.42% 6.2% 8.61% 14.83% 9.29% 9.58% 9.66% 6.57% 6.33% 75m m 4.29% % 8.72% 16.7% 1.39% 7.37% 8.6% % Large 1b % 7.33% 9.74% 17.12% 12.87% 1.18% 8.54% 6.36% 5.64% M edium 1m-9.99m % 14.11% % % 17.11% 13.62% 12.41% 1m-24.99m 13.76% 13.6% 1.82% % 2.14% 15.99% 13.44% 15.17% 14.17% 1.81% 25m-49.99m 12.13% 1.58% 12.3% 9.12% 1.73% % 13.22% 13.67% 12.14% 9.32% 5m-99.99m 11.38% 1.19% 9.76% 9.48% % % % 1m m 8.22% 8.2% 8.84% % % 11.62% 12.3% 9.57% 8.32% 25m m 6.66% 6.68% % 11.38% 8.78% 7.62% 5m m 7.86% 6.24% % 9.43% 22.24% 12.66% 9.96% % m m % 8.49% 7.92% 14.58% % 11.28% 9.66% Large 1b+ 6.4% 6.67% 7.7% 6.93% 8.94% 16.24% 11.44% % 7.23% 6.4% m-9.99m m-24.99m m-49.99m m-99.99m m m m m M edium 5m m 1.94* -.25* * * * 75m m 1.43* 1.11* 2.57*.81* * 2.6* 1.63*.13* 2.35* 1.77 Large 1b+ 1.57* 2.42* 1.3* * 1.81* * 1m-9.99m m-24.99m m-49.99m m-99.99m m m m m M edium 5m m m m Large 1b m-9.99m m-24.99m m-49.99m m-99.99m m m m m M edium 5m m m m Large 1b