Speaker
Description
ABSTRACT
Economic time series analysis has mostly been done in time domain and frequency domain independently. In an attempt to analyze a time series in both domains concurrently, wavelet analysis is used as an alternative approach to other classical approaches. The objective of the study seeks to analyze the time-frequency relationship between stock returns on three different sectors on the Ghana Stock Exchange (GSE). The study employs daily closing returns data of GCB Bank Ltd, Produce Buying Company (PBC) and Unilever Ghana Limited (UNIL). Correlations between stocks are determined by measures of wavelet technique such as wavelet power spectrum and wavelet coherency as alternative to Pearson's linear correlation. Results form the wavelet coherency showed pairs of returns correlation changing over each scale with respect to diverse time horizon.
