MBA & MM Theses

Permanent URI for this collectionhttps://wiredspace.wits.ac.za/handle/10539/9084

For information on accessing MBA & MM Theses content please contact Khomotso Chipu via email : Khomotso Chipu or Tel(W) : 011 717 3638
Alternatively contact Patience Mpitsa via email : Patience Mpitsa or Tel (W) : 011 717 3635

Browse

Filters

20111

Settings

Author: search.filters.author.Lishman, Ryan MarkSubject: search.filters.subject.Share prices

Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Predictability of share price returns on the JSE
    (2011-05-11) Lishman, Ryan Mark
    This study addresses the issue of shar eprice predictability on the JSE Securities Exchange (JSE) over the 1989 to 2009 period. Th eoverall predictability of the JSE in terms of market indices and individual share prices is examined. The predictability in prices of portfolios formed by ranking shares according to a number of easily observable benchmarks is also investigated. Finally, the study identifies leading indicators of the returns of these ranked portfolios. Using a variance ratio test, the study finds little evidence to reject the hypothesis of uncorrelated increments in the JSA All Share Index or in the returns of individual shares. The study does, however, reject the hypothesis that returns of portfolios ranked according to market capitalization, dividend yield, earning yield, industry and trading volume are serially uncorrelated. Furthermore, the strong rejections of these hypotheses retain their statistical significance in the presence of heteroskedasticity in portfolio returns. In line with the results of previous research, the study finds that serial correlation in portfolios increases as the markets capitalisation and average trading volume of its constituents decrease. Portfolios containing high dividend and earnings yield shares are also found to have a high degree of serial correlation. When ranked according to the industry in which their constituents are classified, portfolios of industrial and retail shares show significant levels of serial correlation. Analyses of he load and lag characteristics of portfolios show that the returns of portfolios consisting of large capitalisation and well traded shares lead those consisting of small capitalisation and thinly traded shares. The economic and staistical significance of the results show a significant deviation from the commonly held view that the JSE is an efficient market. From the point of view of an investor seeking a potential avenue of excee return, the challenge lies in constructing a trading strategy and a portfolio with sufficient serial correlation to be robust to the costs incurred from frequent trading.