CHAPTER 6 Hedge Fund Classification

The principal idea behind classifying hedge funds is the ability to convert a set of hedge funds returns into groups so as to be able to visualise clear clustering and style group boundaries. For the fund of hedge fund (FoHF) manager this is indispensable – since clustering and classification are observable ways of scientifically studying hedge fund return data so that an empirical set of estimates can be produced for further hypothesis testing where necessary. As such, similarities emerge from the data – patterns from what would otherwise look very noisy to the naked eye. These patterns allow us to visually check a fund's grouping – to see how similar it is with respect to peers within a stated style. The main technique supported by MATLAB® is the dendrogram (Mantegna and Stanley (2000), Lhabitant (2004)). The dendrogram (“dendro” from the Greek meaning “tree”) is the industry standard method of analysis of the natural hierarchy or taxonomy of a data set. This chapter allows us to apply the tools available in MATLAB® to delve deeper into the structure of hedge fund returns and gain a better understanding of their grouping and classification.

6.1 FINANCIAL INSTRUMENT BUILDING BLOCKS AND STYLE GROUPS

In the Eurekahedge database as of November 2011, there were 13,674 hedge funds – a huge number of funds considering that there are only eight main “building blocks” or instrument groups available for trading. Table 6.1 and Figure 6.1 show these ...

Get Hedge Fund Modelling and Analysis using MATLAB now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.