Thesis (MBA (Business Management))--University of Stellenbosch, 2007. / ENGLISH ABSTRACT: Private equity has been dominating the news recently both as an asset class and a tool to facilitate market efficiency in South Africa. Many private equity funds have had good returns in recent times. However, many pension funds still shun away from this asset class, mainly due to lack of understanding of the risk inherent in this asset class, and in its potential overall risk diversifying characteristics. Regulation 28 of the Pension Fund Act, prescribes maximum allocation of five percent to unlisted assets and that includes private equity as an asset class. One of the reasons why investors don't understand this asset class is the lack of comparable data relative to the listed asset classes such as listed equities and listed bonds. From these asset classes, people get data on a daily basis and they are able to understand the daily prices and values of their portfolio. They can also try to value the individual stocks within their portfolio by using variables such as price earnings ratio and dividend yields, as that information is freely available. This research tries to use the listed private equity companies to establish a Private Equity Index. The criteria used are the same as the ones used by Bilo, Christophers, Degosciu and Zimmennann (2004) when they created a Listed Private Equity Index (LPE), which is made up of 114 global listed private equity companies mainly from Europe and North America. The South African Index was created from six companies, and even though the sample looks small, the results were good enough to justify further studies on the same line and to encourage more listings of these types of companies. The Index was then analysed using modem portfolio analysis and optimisation tools to understand the risk and return characteristics of this asset class and how it behaves when blended with other asset classes. The results suggest that this asset class is under allocated and there is a need to create a more representable index even on an IRR basis. / AFRIKAANSE OPSOMMING: Privaataandele het onlangs die nuus oorheers, beide as 'n bateklas sowel as 'n instrument om die markdoeltreffendheid in Suid-Afrika te fasiliteer. Talle privaataandeelfondse het onlangs goeie opbrengste opgelewer. Baie pensioenfondse skram egter steeds weg van hierdie bateklas, hoofsaaklik weens 'n kennisgebrek van die risiko wat inherent is aan die bateklas en die potensiele risikoverskansing wat die bateklas bied. Regulasie 28 van die Pensioenfondswet vereis dat 'n maksimumtoewysing van vyf persent in ongenoteerde bates bele mag word en dit sluit privaataandele as 'n bateklas in. Een van die redes hoekom beleggers nie hierdie bateklas verstaan nie, is weens 'n gebrek aan vergelykbare data met die genoteerde bateklasse soos genoteerde aandele en genoteerde verbande. Uit hierdie bateklasse verkry mense inligting op 'n daaglikse basis en is hulle in staat om die daaglikse pryse en waardes van hul portefeuljes te verstaan. Hulle kan ook probeer om die afsonderlike gevolge binne hul portefeuljes te begryp deur veranderlikes soos prys inkomsteverhouding en dividendopbrengste te evalueer, soos wat die inligting beskikbaar raak. Hierdie navorsing poog om genoteerde privaataandeelmaatskappye te gebruik om 'n privaataandeelindeks daar te stel. Die maatstawwe wat gebruik word, is dieselfde as die wat Bilo, Christopbers, Degosciu en Zimmennann (2004) gebruik het toe bulle 'n Genoteerde Privaataandeel-indeks (GPA) daargestel bet, waarin 114 genoteerde privaataandeelmaatskappye wereldwyd uit hoofsaaklik Europa en Noord-Amerika gebruik is. Die Suid-Afrikaanse Indeks is saamgestel uit ses maatskappye en hoewel die steekproef klein blyk te wees, was die eindresultaat voldoende bevredigend om verdere navorsing volgens dieselfde riglyne te regverdig en om die notering van meer sulke maatskappye aan te moedig. Die indeks is vervolgens ontleed aan die hand van moderne portefeulje-ontledings tegnieke om die risiko en opbrengste van hierdie bateklas te verstaan en hoe dit reageer wanneer dit met ander bateklasse vermeng word. Die resultate dui daarop dat hierdie bateklas onvoldoende onderskryf is, maar dat daar 'n nodigheid bestaan om 'n meer verteenwoordigende indeks selfs op 'n interne opbrengskoers grondslag daar te stel.