Effects Of Macroeconomic Variables On Equity Prices Of Commercial Banks Listed At The Nairobi Securities Exchange In Kenya

Abstract

A number of studies have been undertaken to identify the factors influencing stock prices in different stock markets. The extant literature available strongly supports the movement of stock price as a consequence of firm specific factors (micro economic) such as dividend, book value, earnings etc. This study sought to establish the effects of macroeconomic variables on equity prices of commercial banks listed at NSE in Kenya. The main objective was to analyze the effects of macroeconomic variables on equity prices of commercial banks listed at NSE in Kenya. The selected macro-economic variables included inflation rate, exchange rate and interest rate. The study followed descriptive research design. Secondary time series data was utilized for the period between 2008 and 2014. The researcher used a time series data analysis approach. First, the data was tested for stationarity. Differencing was done to attain stationarity and subsequently analysis was done using the first lags. Lag order selection was then undertaken. This procedure indicated that the optimum number of lags is one. The researcher then undertook a Johansen cointergration test on the study variables. Since there were no cointergration vectors, a Vector Autoregressive (VAR) was fitted to the study data. The VAR model showed that only lagged Forex rates had a significant effect on other indicators. The influence of other lagged indicators wasn’t significant. This underscores the important of the currency market in determining macroeconomic trends and movements of the stock market in Kenya. Further, granger causality testing showed that the Forex rates unidirectionally granger cause all other time series, i.e. the mean share price, interest rate, and inflation. This implies that the foreign exchange rate is an importance driver of other macroeconomic fundamentals in Kenya. Finally, the researcher tested for residual autocorrelation. The residuals were found to be uncorrelated – an indication that the model was correctly specified.