Predictive ability of selected asset pricing models on the Nairobi stock exchange.
Abstract
This study is on the predictive ability of three
pricing models on the Nairobi Stock Exchange
objective was to identify which of the selected
highest predictive ability.
selected asset
(NSE). The main
models had the
The 57 companies quoted on the NSE as of February 1989 were
surveyed. The focus was on companies that have ordinary shares in
their capital structure. These companies were stratified into two
groups:' the actively traded and the non-actively traded
companies. Though there were sixteen companies in the former
group, only twelve were studied.
The major primary sources of data were the NSE daily price lists,
and the annual reports and accounts of the companies. The data
obtained from these sources were monthly prices; and the earnings
per share (EPS), dividends per share (DPS) and capital employed
respectively.
Forecasting of future sh~re prices, DPS and EPS was done via the
Box-Jenkins (1976) method of time-series analysis. The forecasted
values for DPS and EPS were discounted and capitalized
respectively for five separate periods to give rise to predicted
share prices for the first five months after each company's
financial year end. Each of the prices obtained was compared to
the actual price for the same period. The differences between the
• two prices were subjected to t-tests. From the results of these
tests, it was generally concluded that under the prevailing
conditions in the NSE, no model qualified as a good predictor of
share prices.
Citation
Master of Business AdministrationPublisher
School of Business, University of Nairobi