Analysing the effect of treasury bill rates on stock market returns using Garch model

Abstract

This paper examines the relationship between the returns of ordinary shares listed at the Nairobi Stock Exchange (NSE) and the Treasury Bills Rate using GARCH Analysis. Existing studies in Kenya on the relationship of factors affecting the returns of assets in the NSE have used various methods, mostly Ordinary Least Squares (OLS) Regression and have yielded inconclusive results. This study recognizes the unique characteristic of financial series data that makes common analysis techniques like OLS regression unsuitable for generating effective forecast models. The study systematically examines the returns of the various market segment returns within the NSE for these characteristics so as to build a basis for using GARCH analysis. Finally, it compares the results obtained using OLS regression with the results obtained using GARCH analysis techniques. The study concluded that in keeping with theory, Treasury Bill Rates have a significant impact on the asset returns of the various market segments, the NSE – 20 Share Price Index and all market returns as a whole. The behaviour of the returns of assets on the NSE can be better explained by considering the volatility of previous periods. The study found that GARCH analysis gives a better explanation for the relationship between Treasury Bill Rates and asset returns than OLS regression in every market segment. Furthermore, the explanatory power becomes stronger as we consider the effect of previous variances on the current observations.