Suggestion for further research

This thesis analyses a new approach to hedge fund replication and to the knowledge of the author is the first of its kind. The analysis is thorough; however there is always a possibility to improve the study with more depth and other aspects.

It would be interesting to include other variables that capture the dynamics of hedge funds. The VIX index is a variable that has been discussed by many studies. The index is starting to mature and there exists already sufficient data to include it into the regression model. Further, and more importantly, the trading volumes for the VIX index have risen. The index can soon be described as a liquid instrument.

Also, other alternative risk exposures could be tested. As mentioned, hedge fund strategies often invest in alternative beta, and to capture the alternative risk-exposure one needs alternative variables. An example on an alternative variable is an art index.

Also, the study includes data only until September 2012. The financial markets are constantly developing and therefore new data can yield different results and other findings. Therefore, new data could be included and the development post-crisis could be analyzed in more detail to answer if the model works.

Finally, the study only focuses on managed futures hedge funds. Even if managed futures are one of the most replicable hedge fund strategies, other strategies could be analyzed to increase the robustness of the study. A large variation exists between different strategies, and the chosen variables can explain one strategy better than another.

9 CONCULSION

This thesis has contributed to the on-going discussion as it has shown that it is possible to increase the risk-adjusted returns with screening the data. However, the thesis is not able to support that these replicators earn consistent higher risk-adjusted returns than other assets. However, there is indication that HP replicator is able pre-crisis to beat some of its benchmark indices.

The thesis is divided in four different parts. The first part analyses and classifies hedge funds into two groups based on their performance. The results show that there exists difference between high-performance funds and the aggregated hedge funds measured with risk-adjusted returns.

These groups are thereafter analyzed with a seven-factor risk model to obtain the risk-exposures for the individual hedge funds. The model includes the following factors: (1) USD: the U.S. Dollar index return; (2) BOND: the return on the Barclays intermediate corporate Bond (AA) index; (3) SP500: The SP500 total return, (4) CREDIT: The spread between Barclays intermediate corporate Bond (BAA) index and the Barclays U.S. 5 year treasury index; (5) MORTAGE: The spread between GNME mortage index and the Barclays U.S. 5 year treasury index; (6) COMMODITY; the return of SP GCSI Commodity index; and (7) SMB; The spread between small and large company spreads (Fama-French factor). It shown that hedge fund risk-exposure can be analyzed, but that most variables and models are insignificant. However, it is shown that SP500 and the Commodity variable show significant results.

The third part of the thesis creates a replicator product based on the obtained risk-exposures. The replicator products obtain monthly returns for the whole time period between January 2004 and September 2012, which is further divided in a pre-crisis (January 2004 to August 2008) and post-crisis (September 2008 to September 2012) time periods. It is shown that there exists variation in the risk exposures between these two groups and between these two different time periods.

The final part of the thesis compares the performance and replication quality for the replicators to benchmark return. The performance is shown to be higher for HP replicators compared with the All replicator, and the differences are even greater in the sub-periods. Also, the HP replicator is able to beat some of the benchmark indices pre-crisis risk-adjusted returns. However, the HP replicator is unable to consistently beat

the benchmark indices, due to low post-crisis performance. Overall, the replicators’

replication quality was comparable, however somewhat lower than that of previous studies. Also, there exists a large difference between the replication quality between pre and post-crisis. The thesis shows pre-crisis replication quality consistent with previous research, but low post-crisis replication quality. The assumption is that low post-crisis replication quality is because the financial markets work differently during unusual time periods.

As discussed, hedge funds aim to produce risk that differs from traditional risk, such as tail risk, liquidity risk, and credit risk. Investors are well off for bearing part of this risk, as this able them to diversify their investment and therefore maximize their return to risk ration. This thesis is able to produce a replicator product that maximizes the return earned through replicators.

To conclude, the thesis is unable to produce a replicator product that beats the markets consistently. The implication of this study is that replication performance can be enhanced by data selection and that there exists differences between post and pre-crisis periods. This indicates that the hedge fund replicators should include more hedge funds with high return to the sample when they create their replicator index and that hedge fund replication is not feasible during financial crises.