The Tracking Error of JSE-Listed ETFs and their Determinants

Akinsete, Akintunde
Journal Title
Journal ISSN
Volume Title
ABSTRACT This study examines the tracking errors of exchange-traded funds (ETFs) listed on the Johannesburg Stock Exchange (JSE) and the factors that influence them. Despite the increase in the popularity of ETFs as an investment instrument judging by the more than $1.75 Trillion and 12.6 Billion in asset under management in the United States and South Africa respectively (Hill, Dave, & Matt, 2015; South African Reserve Bank, 2006), there is a dearth of academic research on the subject in South Africa. The few existing academic papers concentrate on equity tracking ETFs while other asset classes such as commodities, bonds and property ETFs have largely been ignored. This study, therefore, looked into the tracking errors and their determinants for JSE-listed ETFs across several asset classes. The tracking error for each ETF in the sample population was calculated using four different methods and the effect of frequency of observation on tracking error was also examined by comparing the tracking error calculated from daily return data with monthly return data. The relationship between tracking error and identified fund characteristics such as fund age, fund size and total expense ratio (TER) were tested using ordinary least square (OLS) regression models to examine the effects of these hypothesised determinants on tracking errors. The research results indicate that JSE-listed ETFs exhibit tracking errors and was found to be higher than those exhibited by ETFs in more developed markets like the US, but similar in magnitude compared to the tracking error exhibited by other emerging market ETFs. The frequency of the returns data used (daily, weekly or monthly) was also found to have a moderating effect. With respect to the determinants of tracking errors, the research found that the age of the fund and its total expense ratio (TER) influenced tracking errors. Contrary to expectation and to the findings in other comparative studies, there was no clear relationship between fund size and tracking error. The study attributed this to the structure of the South African ETF industry. Many of these funds fall under the umbrella of well-established South African financial institutions and as such, are under one strategic group, thereby possessing similar economies of scale and bargaining power (Porter, 1980), which appears to renders the effects of a fund’s size redundant
M.B.A. Thesis
Johannesburg Stock Exchange, Exchange traded funds.