| dc.description.abstract | Studies such as that conducted by Grinblatt et aI., (1984), have proven that Stock dividends
are more than mere cosmetic changes. They appeal to investors because of their
psychological value, tax benefits and because they signal prospects of higher profits in the
future. This study, by refining the Lev and Lakonishok's (1987) study, empirically
investigates the characteristics of the stock dividend firms quoted at the Nairobi Stock
Exchange. It also documents a model for predicting the likelihood of a firm to issue stock
dividends.
The study applies Discriminant Analysis to construct the predictive model. The
characteristics of the firms are derived from the group means. Classification of the firms
into a group is based on the number of times a firm has made the distributions. Those that
have never made the issue between 1991 to 1999 are classified in group 0, the ones that
have issued once in group I, twice in group 2 and thrice in group 3.
The model when expressed in a quadratic function correctly predicts 82% of the firms into
their true groups. The results, based on an examination of fourteen financial statement
derived variables, also indicate that the firms that have never made the distributions have
higher dividend payout ratios, dividend yield, return on investments and a higher percentage
• of capital reserves in the total reserves. This is so because these variables are reduced in size
in the event of an issue. The firms that have made the issue·s twice or more times have the
highest changes in cash from operations, earnings, growth in earnings, shareholders' funds
and total reserves as well as the lowest returns on investments. Re-capitalization increases
the investment thus increasing the earnings realized and consequently, the growth in
earrings.
The frequency in the issues is related to the firm size. Financial and industrial sectors have
the highest concentration of the distributions. Some of the characteristics so established
vii
confirm the Lev and Lakonishok's (1987) findings.
The study proves that only two (total reserves and dividend yield) out of the fourteen
variables are significant for prediction purposes. This means that managers may be using
non-quantitative considerations in deciding whether or not to issue bonus shares and thus
indicating that there exists a gap between finance theory and practice in the issue of Stock
Dividends in Kenya.
-. | en |
