EFFECTS OF DIVIDENDS ON COMMON STOCK PRICES: THE NEPALESE EVIDENCE
EFFECTS OF DIVIDENDSON COMMON STOCKPRICES: THE NEPALESEEVIDENCEGroup 5 Sita Bhandari Sonam Singhal Sunny ShresthaSweta Adhikary Uneet Kaur Ranhotra Vishal Beriwal Vivek Tibrewal
Introduction• Dividend policy determines the division of earnings between payments to stockholders and reinvestment in the firm• Due to complex nature of the dividend decision, corporate dividend policy has been a subject of considerable study particularly since the emergence of MM’s classical work (Miller and Modigliani, 1961)• According to MM, given the investment decision of the company, shareholders in a perfect capital market are indifferent whether the company distributes dividend or retains earnings in the business
Purpose To explain share price, dividend, and retained earnings relationships in the context of Nepal
Directed towards resolving the following issuesin the context of Nepal • Whether dividends or retained earnings are more attractive among Nepalese stockholders? • Is there any presence of economies of scale in dividend supply functions? • If the dividends or retained earnings increases by one percentage point, will market price per share increase by less than or more than unity? • Whether the speed of adjustment through dividend payment or retained earnings between desired and actual share prices is slow or fast?
Review Of Related Literature• The relationship between dividend and share price is not yet clear• Those who believe dividend affects the share price based their argument on the hypothesis (Chawla and Shrinivasan,1987): Shareholders prefer current to future income Dividend has information content and its payment indicates company’s good earning capacity
Contd…………• Those who believes in retention of earnings has their hypothesis based on:If earnings are retained, the company has good investment opportunityDividend in the form of capital gains are subject to lower taxes
Optimal dividend policy• Optimal dividend policy implies maximizing shareholders wealth .The argument that the dividend policy may affect the values can be viewed in two categories (Van Horne and McDonald, 1971):1. Investors may have a net preference for dividends relative to capital gains or vice versa owing to uncertainty resolution, transaction and inconvenience costs, and differential tax rates2. Costs associated with the sale of new issues of equity securities may make these issues more costly source of equity financing than retained earning
Required rate of return, dividend and Retained earning• Some argue that required rate of return used by investors to discount dividends expected in future period increases with time (Gordon, 1963)• Required rate of rises with the proportion of earning retained. As a result, investors would value current dividends over capital gains• Another evidence is that the firms with higher dividend yields sell for lower prices (Litzenberger and Ramaswami, 1982)
Corporate Management Attitude• Studies on corporate management attitude towards dividend policy reveal a preference for stable dividend policy (Linter, 1956) and less preoccupation with tax effect (Baker and Edelman, 1986; and Abrutyn and Turner, 1990)• If ordinary income tax rates are lowered and preferential tax treatment of capital gains to be eliminated, increase in corporate dividend pay out can ratios can be noticed (Papaioannou and Savarese, 1994)• More favourable tax rate on capital gains relative to that on dividends creates a powerful bias in favour of retention of earnings (Elton and Gruber, 1970)• If there existed a net preference among investors for capital gains as opposed to dividends, then the firms dividend policy is determined by profitability of its investments
Dividend payout and sale of equity• Raising dividend payout means increasing the likelihood of selling common stock (Easterbrook, 1984)• Public utility management often follows a policy of high dividends combined with frequent external equity financing• We can find regulated electric utility issuing large amounts of new equity while, over the same period, paying proportionately large cash dividends (Hansen, Kumar and Shome, 1994)
Effect of dividend payment• Dividend payment serve as signals to market participants• Recent evidence has increased the uncertainty regarding the information conveyed when a firm drops its dividend (Jensen and Johnson, 1995)• Dividend reduction tends to be followed by a significant increase in firm earnings (DeAngelo and Skinner, 1992;Healy and Palepu, 1988)• At the other end, dividend- drop also marks the end of a firm’s financial decline and the beginning of firm restructuring (Jensen and Johnson, 1995)
Study showing impact of retainedearnings in India• Study conducted by Desai in 1965 showed that the coefficient of retained earnings was insignificant• Study using time series data concluded that the retained earning had no effect on share prices (Sarkar, 1971)• Study using cross section data revealed impact of retained earning in cotton textiles and impact of dividend in electrical and engineering industries (Kumar and Man mohan, 1971)
The Basic Model• As suggested by Friend and Puckett in 1964, this model is the one most commonly applied to cross-section data as follows: MPSit = a + b DPSit + c REit + eitWhere, MPSit = Market price per share DPSit = Dividend per share Reit = Retained earning per shareThe above model assumes the following reasonable a priori hypothesis:DPSit > 0 and Reit > 0• The subscript i denotes the ith company is a sample of n companies selected from a particular industry• All variables are measured in the ith time period• Market price per share is the average of high, low and closing prices of the year
• Higher b than c should be our typical result, which indicates investor preference for dividends• The equation is useful for estimating price behavior within the observed range of dividend payout• If the companies in a sample tend, on the average, – To pay out less than the optimum, b should be greater than c – If they pay out more than the optimum, b should be less than c – If they pay just the optimum, b should equal c• Theory suggests that regardless of the optimum payout for any individual company, at that optimum Rs.1 of dividends would on average have the same effect on stock price as Rs.1 of Retained Earning
• The separate effects on price of all omitted variables should be aggregated to take into account the firm effect.• The firm effects cannot be measured directly as they are both additive and constant over time• Such firm effects include those relevant to investor assessment of both profit prospects and risks, some of which could alternatively be measured directly• The aggregate firm effect for any firm is assumed proportional to its per share earnings• Therefore, under stated assumptions, firm effects can be held constant by introducing into the regression equation a variable (PE)t-1
• The firm effects include the profitability of investment opportunities as assessed by the market• The problem of least squares bias can be handled by specifying a dividend supply function as the following: MPSit = a + b DPSit + c Reit + (PE)t-1 + eit• Short run changes in income evoke relatively small short runchanges in relative price• Here d measures only the influence of past expectations oncurrent expectations• Lagged price also hold constant any firm effects that exist MPSit = a + b DPSit + c Reit + MPSt-1 + eit• This study takes into account a partial adjustment or flexibleaccelerator model• The above models were tested in linear and logarithmic forms
Nature and Sources of Data• Data taken from Financial Statements of Listed Companies, Vol.III published by Nepal Stock Exchange Ltd.• Thus, the above models are estimated by using secondary data for the listed companies of Nepal.
Contd…• Not all the 110 companies listed in Nepal Stock Exchange Ltd. provide scope of their study.• There is absence of data as many of these companies are either new and have just begun their operations or old listed companies which do not submit their financial statements to Nepal Stock Exchange Ltd.
Contd…• The data for the study purpose is based on pooled cross section data of 29 companies from 1994 to 1999• Which resulted in a total of 93 observations.
Empirical Results (Linear relationships)The Basic ModelMPS = 1709.62 + 4.57 DPS - 12.54 RE (2.41) (4.72*) (1.71)R-bar square = 0.43 F = 7.60 SEE = 225.90 DW = 1.29• Coefficient of DPS is significant• Negative coefficient of RE is questionable
Empirical Results (Linear relationships)Including a lagged price earningMPS = 1701.37 + 4.54DPS - 12.43 RE + 0.09 PEt-1 (2.38) (4.71*) (1.69) (0.28)R-bar square = 0.42 F = 5.04 SEE = 229.13 DW = 1.30• Lagged price variable is added to hold form effects constant and reduce the problem of regression weightsSignificant coefficient of DPS but a negative coefficient of RE Dividends are relatively more attractive Negative coefficient of RE - questionable Retained earnings have no effect on share price
Empirical Results (Linear relationships)Including a lagged price variableMPS = 1666.18 - 9.31DPS - 15.57 RE + 0.75 MPSt-1 (3.07) (0.82) (2.78) (8.08)R-bar square = 0.49 F = 30.45 SEE = 77.99 DW = 0.49• Lagged PE ratio is added to allow for slow short run adjustments in prices to current levels of income.• Dividends receive greater relative weight and has a greater effect than retained earnings• Negative coefficient of RE is still surprising which is the major limitation• Large standard error• However, no indication that retained earnings are more important than dividends
Conclusion From Linear Relationships Nepalese markets find dividend more attractive and is yet to recognize the impact of retained earnings
Logarithmic Modelln MPS = 3.62+ 0.60 ln DPS + 0.24 ln RE ……………(7) (0.92) (5.28*) (0.29) R-bar square = 0.63 F = 20.91 SEE = 0.86 DW = 1.44 Here, • Elasticity of dividend with respect to share price is less than unity which shows the absence of economies of scale.
ln MPS = 3.96 +.63 ln DPS - 0.05 ln RE + 0.32 ln PEt-1………(8) (1.10) (5.75*) (0.06) (3.70*)R-bar square = 0.56 F = 13.04 SEE = 0.787 DW = 1.42Here,• Coefficient of RE is negative.• Dividend coefficient is significant.• Elasticity of dividend with respect to share price is less than unity which shows absence of economies of scale
ln MPS = 3.73 - 0.01 ln DPS - 0.51 ln RE + 0.77 MPSt-1……..(9) (1.38) (0.07) (0.91) (8.55*)R-bar square = 0.64 F = 39.84 SEE = 0.591 DW = 2.28Here,• Adjustment coefficient = 0.23• This implies that the speed of adjustment between desired and actual share prices is slow.• Only, 23% of adjustment of actual to desired share price is completed within a year.
Linear Vs Logarithmic Results• Logarithmic relations reduce the problem of regression weights.• Linear regressions can handle satisfactorily very small and negative RE.• Basically, difference between these two regressions is due to the differing degrees of bias in the regression coefficients produced by short run income disturbances.
Summary• Attempt to determine relative importance of dividends and R/E in determining MPS• Using pooled cross section data of 29 companies from 1994 to 1999 (93 observations)• Finding indicates share value is affected by dividend payments• Consistent with existing preference
Conclusion• Higher investor valuation may be placed on dividends than on R/E• Possible to increase stock price by rising dividends• For growing companies, management might increase stock price by greater retention of earnings, but could not be revealed by this study
Further Research• Study at different points in time to ascertain whether importance of R/E has increased over a period of time• Industry wise analysis to study the importance of dividends or R/E in different industries
Limitations• Undertaking of few companies (93 observations)• Regression model explained less than half of the total variance in linear equations and exhibited other empirical shortcomings