Funds of Hedge Funds

This study also uses data on funds of hedge funds in its analysis. The use of this

sample is important to highlight the differences between this study and the study

by Chen (2009). Chen (2009) particularly includes funds of hedge funds in the

sample with other hedge funds in the multivariate analysis of his study. Applying

the same criteria as for hedge funds to select funds of hedge funds from the TASS

database provides a sample of 763 funds of hedge funds of which 2 funds do not

report minimum investment. Thus, 761 funds of hedge funds are used in the

cross-sectional analysis of risk and performance of funds of hedge funds.

Table 6 presents the descriptive statistics for funds of hedge funds. The statistics are very similar to that of hedge funds presented in Table 2. For instance, the av-erage management fee 1.47 % for funds of hedge funds is approximately the same as for hedge funds. However, the incentive fee is 9.76 % (18.85 % - 9.09 % = 9.76 %) lower for funds of hedge funds. The risk is significantly lower for funds of hedge funds when measured using the standard deviation, downside volatility, MVaR, and VaR. This is likely the result of diversification in different hedge funds. However, the average Cornish-Fischer estimate is more negative for funds of hedge funds implying a fatter left tail of the return distribution of a fund of hedge funds. It is also interesting that this estimate is lower for idiosyncratic re-turns than for market-based rere-turns, which is the opposite for hedge funds. The interesting question remains as to the cause of this distinguishing characteristic of funds of hedge funds.

The statistics in Tables 2 and 6 also suggest that the performance statistics are weaker for funds of hedge funds than for hedge funds. This result, however, can be related to fewer biases in the funds of hedge funds data (see Fung et al. 2002b), therefore it is obvious that funds of hedge funds cannot be judged to perform bet-ter without further evidence.

Table 7 presents the correlation statistics between the variables of the characteris-tics of funds of hedge funds. The correlation statischaracteris-tics for the complexity of the derivative strategy of funds of hedge funds is fairly similar to that of hedge funds.

However, the complexity does not have statistically significant correlation with

management fees and the correlation with the natural logarithm of the size of a

hedge fund is also weaker. These statistics imply that the complexity would not

be as costly for funds of hedge funds.

Table 6. Descriptive Statistics for Funds of Hedge Funds

This table presents descriptive statistics and the Jarque-Bera (JB) test statistics of the variables used in the cross-sectional analysis for funds of hedge funds of this study. The sample includes 761 funds of hedge funds used in multivariate analysis of this study. See Table 1 for definitions of the variables.

LNSIZE LNAGE IFEE MFEE MIN RESTRICTION Mean 17.22 7.70 9.09 1.47 552451.40 57.75 Median 17.23 7.65 10.00 1.50 250000.00 60.00 Maximum 21.74 9.29 30.00 6.00 25000000.00 180.00

Median 0.00 1.78 0.65 -3.38 -4.00 Maximum 60.00 28.45 5.08 -0.05 85.94 Minimum 0.00 0.22 -3.29 -63.37 -90.88 Std. Dev. 5.34 2.13 0.49 4.99 7.25

Skewness 3.69 4.34 -0.23 -4.21 -0.84 Kurtosis 26.30 38.92 19.89 35.44 64.26

JB 18931.48*** 43307.63*** 9051.67*** 35611.90*** 119070.10***

CF DD SKEW EXKURT SHARPE

Skewness 1.78 4.29 -0.85 5.11 0.41 Kurtosis 48.04 36.88 11.61 37.48 5.83

JB 64720.79*** 38716.41*** 2444.30*** 41017.28*** 274.77***

SHARPED APPRAISAL ALPHA SCF ICF

Table 7. Correlation Statistics for Fund of Hedge Funds Characteristics

This table presents a list of correlation statistics for the characteristics of funds of hedge funds.

The probability statistics on the right hand side of the correlation statistics indicate significance of correlation based on the t-statistics.

IFEE HMARK 0.31 0.000 PERCAPITAL LEVERAGED 0.14 0.000 MFEE LNSIZE -0.08 0.021 PERCAPITAL MIN 0.05 0.197 MFEE LNAGE 0.10 0.007 PERCAPITAL LOCKUP -0.04 0.225 MFEE HMARK -0.24 0.000 PERCAPITAL RESTRICTION -0.09 0.011 MFEE IFEE 0.21 0.000 PERCAPITAL AUDIT 0.14 0.000 LOCKUPP LNSIZE -0.01 0.703 OPENENDED LNSIZE -0.04 0.308 LOCKUPP LNAGE -0.07 0.043 OPENENDED LNAGE 0.11 0.002 LOCKUPP HMARK 0.22 0.000 OPENENDED HMARK -0.25 0.000 LOCKUPP IFEE -0.02 0.640 OPENENDED IFEE -0.06 0.094 LOCKUPP MFEE -0.16 0.000 OPENENDED MFEE 0.07 0.053 LOCKUPP LEVERAGED -0.10 0.006 OPENENDED LEVERAGED 0.12 0.001 LOCKUPP MIN 0.12 0.001 OPENENDED MIN -0.10 0.006 AUDIT HMARK -0.10 0.005 COMPLEXITY LEVERAGED 0.22 0.000 AUDIT IFEE -0.02 0.555 COMPLEXITY MIN 0.01 0.853 AUDIT MFEE 0.00 0.940 COMPLEXITY LOCKUP -0.15 0.000 AUDIT LEVERAGED 0.08 0.023 COMPLEXITY RESTRICTION -0.08 0.023

AUDIT MIN 0.07 0.047 COMPLEXITY AUDIT 0.09 0.014 AUDIT LOCKUP 0.02 0.591 COMPLEXITY PERCAPITAL 0.10 0.007

AUDIT RESTRICTION -0.05 0.144 COMPLEXITY OPEN 0.16 0.000 COMPLEXITY OPENENDED 0.11 0.002

Table 8 presents the correlation statistics for the complexity of the derivative

strategy of funds of hedge funds and their performance and risk measures. The

statistics suggest that the complexity is negatively correlated with the

Cornish-Fischer expansion of residual returns. Thus, the complexity seems to be

associ-ated with the fatter left tail of the residual return distribution of a fund of hedge

funds. The statistics for VaR and the standard deviation suggest that the

complex-ity is associated with less risk. The complexcomplex-ity is also negatively associated with the standard deviation of residual returns. The performance ratios instead do not seem to be correlated with the complexity. All in all, the statistics support Hy-potheses 3 and 5b and the support for Hypothesis 5b is related to idiosyncratic returns of funds of hedge funds. Hypothesis 4 is not supported.

Table 8. Correlation Statistics for the Complexity of Derivative Strategy, and Fund of Hedge Funds Risk and Performance

This table presents a list of correlation statistics for the characteristics of hedge funds. The prob-ability statistics to the right of the correlation statistics indicate the significance of the correlation based on the t-statistics.

SKEW COMPLEXITY -0.08 0.029 VAR COMPLEXITY 0.07 0.050

MVAR COMPLEXITY 0.03 0.385 EXKURT COMPLEXITY 0.05 0.204

SSTDEV COMPLEXITY -0.09 0.011 SSKEW COMPLEXITY 0.01 0.724 MEAN COMPLEXITY 0.00 0.953 D COMPLEXITY -0.06 0.114 SKURT COMPLEXITY 0.01 0.797 RSKEW COMPLEXITY -0.10 0.006 RKURT COMPLEXITY 0.04 0.291 RSTDEV COMPLEXITY -0.04 0.241 APPRAISAL COMPLEXITY -0.03 0.393 ALPHA COMPLEXITY 0.00 0.891 SHARPE COMPLEXITY -0.02 0.583 SHARPED COMPLEXITY -0.03 0.473

Figure 2 presents an illustration and the summary statistics for the complexity of the derivative strategies of funds of hedge funds. The distribution of the complex-ity is clearly different from that of hedge funds as illustrated in Figure 1. The complexity peaks at the high and low values of the variable meaning that there are relatively many funds of hedge funds using either a highly complex derivative strategy or not using derivatives at all.

The mean value for the complexity variable is also 109 % higher for funds of

hedge funds than hedge funds. This statistic is also consistent with those

pre-sented in Table 5 for hedge funds which invest in other funds. Thus, funds of

funds use more complex derivative strategies than hedge funds and have a

differ-ent profile of the use of derivatives which motivates one to investigate these funds

separately.

This figure presents the number of different types of derivatives used by a fund of hedge funds for the sample of funds of funds. This figure also presents descriptive statistics for the variable COM-PLEXITY.

Figure 2. Number of Different Type of Derivatives Used by Funds of Hedge

Funds

In document Do investors benefit from the use of options and complexity of derivative strategy of a hedge fund? (sivua 90-95)