Thesis (MBA (Business Management))--University of Stellenbosch, 2007. / AFRIKAANSE OPSOMMING: Verskeie navorsingsprojekte is oor die jare gedoen ten opsigte van die voorspellingsmoontlikhede van kontantvloei en winste, met teenstrydige resultate. Daar is egter weinig navorsing gedoen oor die verhouding wat tussen winste en kontantvloei bestaan. Hierdie navorsingsverslag beoog om ondersoek in te stel na die verhouding tussen kontantvloei en winste, dus om te poog om te bepaal watter een die drywer is, maar ook om te bepaal of die een veranderlike ingespan kan word om vooruitskattings ten opsigte van die ander te kan doen.
Aangesien finansiële tydreekse die meeste van die tyd nie-stasionêr is, moet dit in ag geneem word wanneer die kousale verwantskap tussen die twee veranderlikes bepaal word, asook wanneer regressie-analise met die oog op vooruitskatting gedoen word. Daar word egter vermoed dat die aspek van stasionariteit weinig aandag geniet in menige navorsing wat ten opsigte van finansiële tydreekse gedoen word.
Die feit dat weinig tydreekse stasionêr is, is bevestig deur te toets vir die bestaan van eenheidswortels in die veranderlikes. Die beste resultate vir stasionariteit is verkry deur die tweede verskille van die veranderlikes te bereken. Daar kon egter nie met sekerheid vasgestel word of winste kontantvloei dryf of andersom nie. Die gevaar van skyn-korrelasie is ook bewys, aangesien 'n groot aantal pare veranderlikes beduidende korrelasies tussen mekaar aandui, maar wanneer hulle stasionariteit en kousaliteit in ag geneem word, is weinig van die pare veranderlikes kousaal verwant aan mekaar. Die toets vir ko-integrasie is ingespan om steun te verleen by die regressie-analise en vooruitskatting van die tydreekse. Die regressie analise van die geko-integreerde tydreekse het in die meeste gevalle 'n hoë R2 en aangepaste R2 gelewer. Die vooruitskattings was egter teleurstellend onakkuraat. / ENGLISH ABSTRACT: Throughout the years a variety of research projects have been done about the predictive ability of cash flow and earnings, with contradictory results. However, limited research has been done about the relationship between cash flow and earnings. The aim of this research report is to investigate the relationship between cash flow and earnings, thus attempting to determine which one is the driver, but also to investigate the ability of one variable to predict the other.
Because financial time series are non-stationary most of the time, this fact has to be taken into account when the causal relationship between the two variables is determined, as well as when regression analysis is done with forecasting in view. It is, however, suspected that the fact of stationarity has been neglected in much of the research that has been done on financial time series.
The fact that very few time series are stationary has been established by testing for the existence of unit roots in the variables. The best results for stationarity were obtained by calculating the second differences of the variables. It could not be established beyond doubt whether earnings cause cash flows or vice versa. The danger of spurious correlation has been proved, because a vast number of pairs of variables indicates a significant correlation with one another, but when stationarity and causality are taken into account, only a few pairs of variables are truly significantly correlated to one another. The test for co-integration was used to assist in the regression analysis and forecasting of non-stationary time series. The regression analysis of most of the co-integrated variables resulted in a high R2 and adjusted R2. The forecasted values, however, were disappointingly inaccurate. / cmc2010
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/801 |
| Date | 12 1900 |
| Creators | Bezuidenhout, Annelise |
| Contributors | Hamman, W.D., Nel, D.G., University of Stellenbosch. Faculty of Economic and Management Sciences. Graduate School of Business. |
| Publisher | Stellenbosch : University of Stellenbosch |
| Source Sets | South African National ETD Portal |
| Language | en_ZA |
| Detected Language | Unknown |
| Type | Thesis |
| Rights | University of Stellenbosch |
Page generated in 0.0085 seconds