Are unit trust performances inflated at quarter-ends

Steyn, Esther (2003-12)

Thesis (MBA)--Stellenbosch University, 2003.

Thesis

ENGLISH ABSTRACT: Suspicion has long existed that the performance of unit trusts is artificially inflated at quarter-ends, as those are the periods focussed on in the press. Studies investigating this have been done in the USA, but to date no similar study has been done in South Africa. This study examines the daily net asset value (NAV) performances of South Africa's domestic equity funds for the period from 1996 to 2002, as obtained from the MoneyMate database. Specifically, returns in excess of the benchmark index at quarterends are compared to excess returns at other month ends. The benchmark index performance is subtracted to correct for movement in the market. Domestic equity funds are divided into nine sectors and in much of this study, the funds in a sector are examined collectively. Results show that unit trust performance does peak at year-ends and quarter-ends, with a subsequent drop at the start of years and quarters. Furthermore, more funds beat the JSE All Share Index at year-ends and quarter-ends, than is normally the case. In contrast, this pattern was not repeated in the percentage of funds beating zero. Following these results, tests were done to establish whether a relationship exists between an end-of-quarter positive return and a subsequent beginning-of-quarter negative return. The suspected relationship was confirmed specifically in the funds from the period from 2000 to 2002. From this study, it would appear that inflating unit trust performances at quarter-ends is not yet extinct in South Africa. Advice to potential investors would therefore be to invest at the beginning of a quarter and to sell at the end.

AFRIKAANSE OPSOMMING: Daar bestaan lank reeds die vermoede dat die opbrengste van effektetrusts op kwartaaleindes opwaarts bestuur word, siende dat dit die opbrengste is wat dekking in die media geniet. Daar is al in die VSA studies gedoen wat hierdie vermoede ondersoek, maar tot op hede was daar nog geen soortgelyke studie in Suid-Afrika nie. Hierdie studie ondersoek die daaglikse opbrengste van die netto batewaarde van Suid- Afrika se binnelandse effektetrusts vir die tydperk 1996 tot 2002, soos verkry van die databasis MoneyMate. Die verskil tussen opbrengste van die effektetrusts en die indeks op kwartaaleindes is vergelyk met die verskil op ander maandeindes (die verskil tussen opbrengste is geneem om bewegings in die mark in ag te neem). Plaaslike effektetrusts word in nege sektore verdeel, en in die grootste deel van hierdie studie word die effektetrusts in 'n sektor as 'n geheel ondersoek. Resultate toon dat effektetrusts opbrengste wel pieke toon op jaareindes en kwartaaleindes, en dan weer val in die begin van jare en kwartale. Verder is daar ook meer effektetrusts wat beter vaar as die JSE Alle Aandele Indeks op jaareindes en kwartaaleindes as wat normaalweg die geval is. In kontras word hierdie patroon nie herhaal as gekyk word na die persentasie van effektetrusts wat nul klop nie. Na aanleiding van hierdie resultate is toetse gedoen om vas te stel of daar 'n verband bestaan tussen positiewe opbrengste op kwartaaleindes en die daaropvolgende negatiewe opbrengste aan die begin van kwartale. Hierdie verband is veral gevind vir die tydperk 2000 tot 2002. Uit hierdie studie wil dit blyk dat, in Suid-Afrika, die opbrengste van effektetrusts op kwartaaleindes wel kunsmatig verhoog word. Raad aan voornemende beleggers sou dus wees om aan die begin van 'n kwartaal te koop, en aan die einde te verkoop.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/53290
This item appears in the following collections: