Testing for the reliability of the Nairobi Stock exchange indices

Abstract

The stock market indices are today a vital tool for both economists and actors in the financial world. The multiplication and the very importance given to the indices raise the question of their accuracy and of the reliability of the methods that are used to construct them. The study begins an investigation on these questions by developing a background on the understanding of the indices. The objectives of the study were to determine how well the NSE-20 share index represents market performance, to determine how well the AIG (EA) 27 share index represents market performance to test the relationship between the NSE-20 index and AIG-27 index and to test the existence of biases in the two indices. The results show that the NSE indices have biases; the AIG-27 share index is more representative index than the NSE-20 share index in measuring the market performance and that a relationship between the NSE indices exists. The MCAP and TVA affect both in almost a similar magnitude. The study developed a comprehensive theoretical framework. This is because index numbers are defmed and computed to pro-v.i.de solutions to practical problems and there can be no measurement without theory and much of what is vague and ambiguous in index numbers practices can be traced back to the lack of good theoretical basis. The study comprised a population of all the 52 companies registered and actively trading their shares as at June 31, 2006 at the NSE. The data set consisted of the NSE and AIG reports for the companies that traded for the period 1999 to 2006, both years inclusive. The study used secondary data only. The base date was dictated by the time in which the AIG Global investment company has computed the AIG-27 share index. The analysis was tied to each objective to enable accurate conclusions.