THE DETERMINANTS OF CORPORATE DIVIDEND POLICIES IN PAKISTAN

—AN EMPIRICAL ANALYSIS

By ANEEL KANWER
Apr 28 - May 04, 2003

In a world where there are no obligations to pay dividend on common stock, why then the companies pay dividend. Dividend pay out decisions is among the basic policy choices confronting corporate financial officers. How much to pay is still an open issue.

Over the past several decades, finance scholars have engaged themselves in extensive theorizing about factors that might be important in determining a firm's dividend policies. In developed countries the decision between paying dividend and retaining earnings has been taken seriously by both investors and management, and has been subject of considerable research by economists in the last four decades.

The factors influencing dividend policies might include profit after tax, liquidity of the firm, reserve position of the firm, changes in equity holding, turnover of shares, and size of the firm. The decision to retain, reinvest or pay out after tax earnings in form of cash or stock dividend is important for realization of corporate goals — the maximization of the value of the firm.

We try identifying the factors influencing dividend pay out decisions of the firms listed at Karachi Stock Exchange. The study also attempts to describe variations in dividend payout behavior within different industries; but the reasons for those variations are not discussed.

DATA

Table 1
AVERAGES OF VARIABLES IN VARIOUS INDUSTRIES (1992-98)

Sector Name

Dividend Yield

Market to Book Ratio

Sales

Surplus

Share Price

Fuel & Energy

0.03

3.05

17,488

1,950

96

Chemicals

0.03

1.69

1,382

389

94

Cement

0.03

6.60

1,227

326

70

Others

0.03

2.34

677

74

56

Engineering

0.03

1.40

982

106

39

Other Textile

0.03

1.49

944

230

37

Paper & Board

0.04

1.50

606

188

34

Tobacco

0.04

1.84

5,156

248

31

Cotton Textile

0.02

1.75

634

49

25

Vanaspati & Allied

0.01

3.48

454

(176)

24

Sugar & Allied

0.03

0.94

762

164

24

Transport

0.01

(0.50)

11,490

1,767

21

Jute

0.01

6.71

430

83

10

The study considering the criteria for good economic and business research undertakes two business cycles (about 8 year). It covers period from 1992-97. We include 317 companies in our analysis. The choice of the companies is based on the availability of data.

The data for the listed companies is obtained from the publication of State Bank of Pakistan named as "Balance Sheet Analysis Joint Stock Companies Listed on Karachi Stock Exchange". Information on Stock Price was obtained from publication of State Bank of Pakistan named as "Index of Stock Prices".

The sample statistics are presented in Table 1.

Share Price is highest in Fuel and Energy Sector, while in the period 1980-86 the highest share price was observed for Sugar and Allied Industry [Mohammed Nishat, Nighat Bilgrami, 1994]. Break up of share price analysis reveals that there is significant difference among the share prices in a sector for different years.

Averages of Dividend Per Share and Market to Book Ratio also reveal significant information. In all the sectors Market to Book Ratio is observed to be greater than Dividend Per Share except for two sectors namely Transport and Chemicals.

Sales and Surplus in Sectors Transport and Fuel and Energy are the highest. A possible reason for this difference might pertain to the unit price being highest. The general trend is that sales are about ten times of the surplus. But only in the case of Tobacco and Cotton Textile the sales exceed the ten time general trend.

THE MODEL

Nobel laureates Merton Miller and Franco Modigliani provided proofs for the now famous M & M "irrelevance" propositions. In doing so they provided solid foundation for our current understanding of the determinants of dividend policies. At first glance, the M & M irrelevance propositions may seem the height of ivory-tower irrelevance-especially to those corporate Chief Financial Officers who get paid lots of money to make dividend decisions that "do not matter." But these propositions have considerable practical value by directing our search for factors likely to be important in setting corporate dividend policies. M & M irrelevance propositions and certain other economic basis draw our attention to certain factors that might be considered important; if at all, corporate capital structure decisions affect the values of the companies in fairly predictable ways. On the basis of these we present our model in the following equation:

DPS = (SIZE, SURPLUS, M/BRATIO, EBIT-INC-DEC,)

Where:
Size : Size of the firm
Surplus : Reserves
M/BRatio : Market to book ratio
Quality of firm : Dummy variable for signaling hypothesis

We examine whether company size (measured as total sales) has a systematic effect on dividend yields. Larger companies might be expected to have higher dividend yields as compared to smaller companies.

The surplus variable is expected to influence the dividend declaration negatively. The reason being that if profit goes to surplus and is retained by the firm for other investment and expansion purposes, less is left for dividend. The last year's surplus retained in the investment and if actualized may influence the profit level of the firm positively, as a result of which the chances of dividend payment would increase.

We expect company's dividend choices to be determined in large part by the extent of their investment opportunities. The greater those opportunities (relative to the size of the firm), the lower should company's dividend yield. We used company's market to book ratio as our proxy for its investment opportunity set.

With better information about the value of their firms than outside investors, managers are likely to face circumstances when they would like to communicate this information to the market. Although there are many potentially effective signaling devices, a change in dividend choices is a notable candidate. According to the signaling theory, high quality companies will make higher dividend payments than low quality firms. We classified firms as high quality in any given year in which their earnings before interest and taxes increased in the following year and we designated as low quality all firms whose earnings decreased in the next year.

EVIDENCE ON DETERMINANTS OF DIVIDEND POLICIES

Up to this point we have discussed various factors that may be important in driving corporate dividend choices. We now present some empirical evidence to help identify the relative "explanatory power" of each of these factors. We calculated dividend yield for some 300 companies over the past seven years (1992-98) and in the process we found considerable "cross sectional" and "time series variation".

As described earlier we developed proxies for each of the factors outlined above. Having devised these proxy measures, we then used regression analysis to assess the strength of the correlations between the various proxies and corporate dividend yields.

We encountered selection bias problem in our research. Our sample contained about 70% observations where firms did not pay any dividend. To overcome this Heckman Two Step procedure has been used.

Table 2
RESULTS OF 2nd STAGE REGRESSION

 

Sales

Surplus

Market to Book Ratio

EBIT Increase or Decrease

Unstandarized Coefficient

1.81E-07

-1.07E-05

-5.36E-03

2.35E-02

Standardized Coefficient

0.022

-0.102

-0.205

0.158

T Statistic

0.515

-2.374

-6.139

4.802

Odds Ratio

1

1

1

1

Percentage Increase/Decrease
Significant results in Bold.

2.22%

-9.70%

-19%

17%

The results support our assumptions. We expected size of the firm size to have a positive effect on the dividend payout decisions. The coefficient is in supports our expectation but is not statistically significant.

Current year's surplus also effects dividend payout decisions significantly. Higher the profit retained this year for capital expenditure and operations lower will be the divined payment. Table 2 indicates that increase in surplus by one unit would result in decline in dividend yield by 10%.

Investment opportunities available to the firm (relative to size) also effect the dividend payout decision as expected significantly. If there are good investment opportunities and possibly higher return on investment, the firms intend to increase the shareholder's wealth by retaining and reinvesting the profit. Increase of one unit in Market to Book ratio would result in decline in the dividend yield by 19%.

The results of proxy for signaling theory are also as expected. If the firm is good quality firm then possibility of paying dividend is higher as compared to the firms classified as low quality.

We also examined the effects of Last Year's Net Profit, Last Year's Surplus on the dividend yield of the firms. We could not find any significant relation between these.

CONCLUSION

This study attempts to identify the factors that explain the dividend behavior of the firms registered with Karachi Stock Exchange using firm-wise data for 1992-98. The empirical findings validate the theoretical prediction for positive and significant impact of growth opportunities available to the firm. Moreover, the findings also draw our attention to the signaling theory that states that managers tend to convey information, which they have to the shareholders through dividend choices. We saw from the results that it is also true for Pakistan. We also compared various industries and saw that Chemicals, Cement and Fuel & Energy at times fetched high prices because of higher dividend payments. The results also suggest that higher net profit after tax of these firms does not necessarily ensure higher dividend payment.