Dividend Policy of Indian Corporate Firms: An Analysis of Trends and Determinants Dr. Y.

Subba Reddy1
The present study examines the dividend behavior of Indian corporate firms over the period 1990 2001 and attempts to explain the observed behavior with the help of trade-off theory, and signaling hypothesis. Analysis of dividend trends for a large sample of stocks traded on the NSE and BSE indicate that the percentage of companies paying dividends has declined from 60.5 percent in 1990 to 32.1 percent in 2001 and that only a few firms have consistently paid the same levels of dividends. Further, dividend-paying companies are more profitable, large in size and growth doesnt seem to deter Indian firms from paying higher dividends. Analysis of influence of changes in tax regime on dividend behavior shows that the tradeoff or tax-preference theory does not appear to hold true in the Indian context. Test of signaling hypothesis reinforces the earlier findings that dividend omissions have information content about future earnings. However, analysis of other non-extreme dividend events such as dividend reductions and non-reductions shows that current losses are an important determinant of dividend reductions for firms with established track record and that the incidence of dividend reduction is much more severe in the case of Indian firms compared to that of firms traded on the NYSE. Further, dividend changes appear to signal contemporaneous and lagged earnings performance rather than the future earnings performance.

Asst. Professor, Institute for Financial Management and Research (IFMR), Chennai. The views expressed and the approach suggested are of the authors and not necessarily of NSE.

1. Introduction
From the practitioners viewpoint, dividend policy1 of a firm has implications for investors, managers and lenders and other stakeholders. For investors, dividends whether declared today or accumulated and provided at a later date - are not only a means of regular income2, but also an important input in valuation of a firm3. Similarly, managers flexibility to invest in projects is also dependent on the amount of dividend that they can offer to shareholders as more dividends may mean fewer funds available for investment. Lenders may also have interest in the amount of dividend a firm declares, as more the dividend paid less would be the amount available for servicing and redemption of their claims. However, in a perfect world as Modigliani and Miller (1961) have shown, investors may be indifferent about the amount of dividend as it has no influence on the value of a firm. Any investor can create a home made dividend if required or can invest the proceeds of a dividend payment in additional shares as and when a company makes dividend payment. Similarly, managers may be indifferent as funds would be available or could be raised with out any flotation costs for all positive net present value projects. But in reality, dividends may matter, particularly in the context of differential tax treatment of dividends and capital gains. Very often dividends are taxed at a higher rate compared to capital gains. This implies that dividends may have negative consequences for investors4. Similarly, cost of raising funds is not insignificant and may well lead to lower payout, particularly when positive net present value projects are available. Apart from flotation costs, information asymmetry between managers and outside investors may also have implications for dividend policy. According to Myers and Majluf (1984), in the presence of information asymmetry and flotation costs, investment decisions made by managers are subject to the pecking order of financing choices available. Managers prefer retained earnings to debt and debt to equity flotation to finance the available projects. Information asymmetry between agents (managers) and principals (outside shareholders) may also lead to agency cost (Jensen and Meckling, 1976). One of the mechanisms o reducing expropriation of outside f shareholders by agents is high payout. High payout will result in reduction of free cash flow available to managers and this restricts the empire building efforts of managers. The presence of information asymmetry may a mean that managers need to signal their ability to lso generate higher earnings in future with the help of high dividend payouts (Bhattacharya, 1979, John and Williams 1985, and Miller and Rock, 1985). However, the credibility of signals depends on the cost of signaling the cost being loss of financial flexibility. High payout results in reduction of free cash flow when in fact the firm needs more funds to pursue high growth opportunities. Rozeff (1994) models payout ratios as a function of three factors: flotation costs of external funding, agency cost of outside ownership and financing constraints as a result of higher operating and financial leverage5. To summarize, several theories have been proposed in explaining why companies pay dividends6. While many earlier studies point out the tax-preference theory, more recent studies emphasize signaling and agency cost rationale of dividend payments. However, the dividend puzzle is yet unresolved and the words of

Brealey (1992) poses the dividend policy decision as What is the effect of a change in cash dividends, given the firms capital-budgeting and borrowing decisions? In other words, he looks at dividend policy in isolation and not as a by-product of other corporate financial decisions. 2 Lintner (1956) finds that firms pay regular and predictable dividends to investors, where as the earnings of corporate firms could be erratic. This implies that shareholders prefer smoothened dividend income. 3 Bernstein (1998) observes that given the concocted earnings estimates provided by firms, the low dividend payout induces reinvestment risk and earnings risk for the investors. 4 Black (1976) notes that in the presence of taxes, investors prefer smaller dividends or no dividends at all. 5 According to Kalay (1982), in the absence of restraining covenants, shareholders can transfer wealth from bondholders by paying off dividend to themselves either by selling existing assets or by reducing investment or by using proceeds of a senior debt. 6 Baker, Powell and Veit (2002) survey different streams of research work on dividends.
1

Fischer Black (Black 1976) may well apply in todays context: The harder we look at the dividend picture, the more it seems like a puzzle, with pieces that just dont fit together. One of the striking aspects that have been noticed in recent periods is the lower dividend paid by corporate firms in the US. Fama and French (2001) analyze the issue of lower dividends paid by corporate firms over the period 1973-1999 and the factors responsible for such a decline. They attribute the decline to changing firm characteristics of size, earnings and growth. However, it is to be seen whether the change towards lower dividends is a permanent feature or will there be reversal. A decline in dividends, according to Fama and French, could be due to lower transaction costs, improved corporate governance mechanisms, and the increasing preference towards capital gains.

1.1 Indian Scenario

In the Indian context, a few studies have analyzed the dividend behavior of corporate firms. Mahapatra and Sahu (1993) find cash flow as a major determinant of dividend followed by net earnings. Bhat and Pandey (1994) undertake a survey of managers perceptions of dividend decision and find that managers perceive current earnings as the most significant factor. Narasimhan and Asha (1997) observe that the uniform tax rate of 10 percent on dividend as proposed by the Indian union budget 1997-98, alters the demand of investors in favor of high payouts. Mohanty (1999) finds that firms, which issued bonus shares, have either maintained the pre-bonus level or only decreased it marginally there by increasing the payout to shareholders. Narasimhan and Vijayalakshmi (2002) analyze the influence of ownership structure on dividend payout and find no influence of insider ownership on dividend behavior of firms. However, it is still not clear as to what is the dividend payment pattern of firms in India and why do they initiate and omit dividend payments or reduce or increase dividend payments. Hence it is proposed to analyze the dividend payout of firms in India and analyze the dividend initiations and omissions and other changes in dividends and the signals that these events convey. Following Fama and French (2001), the present study also attempts to analyze the impact of profitability, size and growth on the dividend payout of firms. Similarly, following Healy and Palepu (1988) an attempt is made to analyze the signaling hypothesis, i.e. earnings information conveyed by dividend initiations and omissions. Since, initiations and omissions construe extreme dividend events, changes in dividends i.e., increases and decreases and the information that they convey is also examined following DeAngelo, DeAngelo and Skinner (1992). There have been several changes in the tax regime in the last few years. The union budget 1997-98 made dividends taxable at t e hands of company paying them and not in the hands of investors receiving them. h Similarly there have been changes in the capital gains tax and exemption of dividend income under Section 80 L of the Income Tax Act 1961. All these changes have implications for the dividend policy of corporate firms. According to tax-preference or trade-off theory, favorable dividends tax should lead to higher payouts. Hence it is proposed to analyze the impact of tax regimes on dividend policies of corporate firms.

1.2 Objectives
1. To study the trends in the dividend payment pattern of Indian corporate firms; 2. To analyze the impact of changes in dividend tax on the propensity to pay dividends; 3. To analyze the influence of firm characteristics such as profitability, growth and size on the dividend payment pattern; 4. To analyze the signaling hypothesis, specifically earnings information conveyed by dividend initiations and omissions; and 5. To analyze the influence of loss on dividend reductions.

In other words, the present study focuses on an analysis of dividend trends and attempts to analyze the determinants of these trends with the help of trade-off or tax-preference theory and signaling hypothesis. There are other important determinants of dividend behavior such as transactions costs, which we will not analyze, in the present study. In the next Section, we review the relevant literature, followed by a description of the database employed and methodology adopted in Section 3. Dividend trends are discussed in Section 4, and the analysis of characteristics of dividend payers is presented in Section 5. Sections 6 and 7 deal with the signaling hypothesis: first the case of dividend initiations and omissions and second dividend reductions. Section 8 summarizes the finding of study, points out limitations and concludes with directions for further research.

2. Review of Relevant Literature

DeAngelo, DeAngelo and Skinner (1992) analyses the relationship between dividends and losses and the information conveyed by dividend changes about the earnings performance. They examine the dividend behaviour of 167 NYSE firms with at least one annual loss during 1980-95 and those of 440 firms with no losses during the same period, where all the firms had a consistent track record of ten or more years of positive earnings and dividends. They find that 50.9% of 167 firms with at least one loss during 1980-95 reduced dividends, compared to 1% of 440 firms without losses. Their findings support signaling hypothesis in that dividend changes improve the ability to predict future earnings performance. Glen et al. (1995) study the dividend policy of firms in emerging markets. They find that firms in these markets have a target dividend payout rate, but less concerned with volatility in dividends over time. They also find that shareholders and governments exert a great deal of influence on dividend policy and observe that dividends have little signaling content in these markets. Benartzi, Michaely, Thaler (1997) analyzes the issue of whether dividend changes signal the future or the past. For a sample of 7186 dividend announcements made by NYSE or AMEX firms during the period 1979-91, they find a lagged and contemporaneous relation between dividend changes and earnings. Their analysis also shows that in the two years following dividend increases, earnings changes are unrelated to the sign and magnitude of dividend changes. Bernsterin (1998) expresses concern over the decline in payout over a period of time in the US market. He observes that given the concocted earnings estimates provided by firms, the low dividend payout induces reinvestment risk and earnings risk for the investors. He asserts that try calculating the historical correlation between payout ratios in year t and earnings growth over t + 5. The correlation coefficient is positive and statistically significant 7. Fama and French (2001) analyze the issue of lower dividends paid by corporate firms over the period 1973-1999 and the factors responsible for the decline. In particular they analyze whether the lower dividends were the effect of changing firm characteristics or lower propensity to pay on the part of firms. They observe that proportion of companies paying dividend has dropped from a peak of 66.5 percent in 1978 to 20.8 percent in 1999. They attribute this decline to the changing characteristics of firms: The decline in the incidence of dividend payers is in part due to an increasing tilt of publicly traded firms toward the characteristics small size, low earnings, and high growth of firms that typically have never paid dividends8. Baker, Veit and Powell (2001) study the factors that have a bearing on dividend policy decisions of corporate firms traded on the Nasdaq. The study, based on a sample survey (1999) response of 188 firms out of a total of 630 firms that paid dividends in each quarter of calendar years 1996 and 1997, finds that the following four factors have a significant impact on the dividend decision: pattern of past dividends, stability
7 8

Bernstein (1998), pp. 1. Fama and French (2001), p. 79

of earnings, and the level of current and future expected earnings. The study also finds statistically significant differences in the importance that managers attach to dividend policy in different industries such as financial versus non-financial firms. Ramacharran (2001) analyzes the variation in dividend yield for 21 emerging markets (including India) for the period 1992-99. His macroeconomic approach using country risk data finds evidence for pecking order hypothesis lower dividends are paid when higher growth is expected. The study also finds that political risk factors have no significant impact on dividend payments of firms in emerging markets. Lee and Ryan (2002) analyze the dividend signaling-hypothesis and the issue of direction of causality between earnings and dividends - whether earnings cause dividends or vice versa. For a sample of 133 dividend initiations and 165 dividend omissions, they find that dividend payment is influenced by recent performance of earnings, and free cash flows. They also find evidence of positive (negative) earnings growth preceding dividend initiations (omissions).

2.1 Previous Indian Studies

Kevin (1992) analyzes the dividend distribution pattern of 650 non-financial companies which closed their accounts between September 1983 and August 1984 and net sales income of one crore rupees or more. He finds evidence for a sticky dividend policy and concludes that a change in profitability is of minor importance. Mahapatra and Sahu (1993) analyze the determinants of dividend policy using the models developed by Lintner (1956), Darling (1957) and Brittain (1966) for a sample of 90 companies for the period 1977-78 1988-89. They find that cash flow is a major determinant of dividend followed by net earnings. Further, their analysis shows that past dividend and not past earnings is a significant factor in influencing the dividend decision of firms. Bhat and Pandey (1994) study the managers perceptions of dividend decision for a sample of 425 Indian companies for the period 1986-87 to 1990-91. They find that on an average profit-making Indian companies have distributed about one-third of their net earnings and that the average dividend payout ratio is 43.6 percent. They also find that the average dividend payout ratio is 54 percent for the sample of both profitmaking and loss-making companies and the average dividend rate is in the range of 14.3 percent to 19.2 percent. They also observe variation in dividend policy of different industries. Further, a survey of these 425 companies has been attempted. How ever, only 31 questionnaires have been received and of these they find 28 amenable for further analysis. Their analysis of the respondents shows that managers perceive current earnings as the most significant factor influencing their dividend decision followed by patterns of past dividends. They also find two other variables increasing equity base and expected future earnings to have significant influence. However, they find industry to have the least influence on the dividend, which has been contrary to the expectations. Mishra and Narender (1996) analyze the dividend policies of 39 state-owned enterprises (SoE) in India for the period 1984-85 to 1993-94. The find that earnings per share (EPS) is a major factor in determining the dividend payout of SoEs. Narasimhan and Asha (1997) discuss the impact of dividend tax on dividend policy of firms. They observe that the uniform tax rate of 10 percent on dividend as proposed by the Indian union budget 1997-98, alters the demand of investors in favor of high payouts rather than low payouts as the capital gains are taxed at 20 percent in the said period. Mohanty (1999) analyzes the dividend behavior of more than 200 firms for a period of over 15 years. He finds that in most bonus issue cases firms have either maintained the pre-bonus level or only decreased it marginally there by increasing the payout to shareholders. The study also finds that firms that declared bonus during 1982-1991 showed higher returns to their shareholders compared to firms which did not issue bonus shares but maintained a steady dividend growth. He finds evidence for a reversal of this trend in the 1992-

1996 period. He attributes such a reversal in trend to the changed strategy of multi-national corporations (MNCs) and their reluctance to issue bonus shares. Narasimhan and Vijayalakshmi (2002) analyze the influence of ownership structure on dividend payout of 186 manufacturing firms. Regression analysis shows that promoters holding as of September 2001 has no influence on average dividend payout for the period 1997-2001.

3. Database and Methodology 3.1 Database

Dividend payment pattern of all companies that are listed for trading on one of the two major exchanges namely National Stock Exchange (NSE) and Bombay Stock Exchange (BSE) during the period 1989-1990 to 2000-2001 (we refer each year henceforth with the end year i.e., for 2000-2001 to 2001) are employed for analysis. The data has been sourced from Prowess database of the Centre for Monitoring Indian Economy (CMIE). For the purpose of this study, only final cash dividends are considered and stock repurchases and stock dividends are not considered. Unlike the firms in developed countries that pay quarterly dividends, Indian companies typically pay only one dividend during a year. A few firms do pay interim dividends, however, data regarding these are not readily accessible and it is extremely difficult to get such data for a reasonable number of years. Further, stock repurchases have been permitted only recently and only about a hundred companies have bought back their stocks so far. Hence, in the present study stock repurchases are not considered for analysis. Stock price data for the prior year of dividend announcement are also taken from the Prowess database.

3.2 Methodology for Analysis of Trends

To analyze the trends in dividend payment pattern, number of companies paying dividend as percentage of total firms, average dividend paid, dividend per share, payout ratio, and dividend yield are computed for the period 1990 to 2001. Dividend per share (DPS) is calculated as
DPS j ,t = Dividend j ,t EQCap j ,t

Where, DPSj,t refers to dividend per share for company j in year t; Dividend j,t refers to amount of dividend paid by company j in year t; and EQCap j,t refers to paid -up equity capital for firm j in year t. Equity capital is employed instead of the usual number of outstanding shares in the denominator as it facilitates comparison of rupee dividend paid per share by removing the impact of different face or par values. Dividend payout ratio (PR) is computed as

PR jt =

Dividend j , t PAT j ,t

Where, PR j,t is dividend payout ratio, Dividend j,t refers to amount of dividend paid by company j in year t; and PATj,t refers to net profit or profit after tax for firm j in year t. Dividend Yield (DY) is computed as

DY jt =

DPS j ,t Price j ,t 1

Where, DYjt refers to dividend yield for firm j in year t, DPSjt refers to dividend per share for firm j in year , and Pricej,t-1 is closing price of previous year for firm j.

Further, the entire sample is categorized into payers and non-payers to examine the trends in dividends across different subgroups. Payers are those firms that have paid dividend in the current year, where as nonpayers have not paid dividend in the current year. Payers are further classified into regular payers, initiators and current payers. Regular payers are those firms that have paid dividend regularly without ever skipping the payments. Initiators on the other hand refers to those firms with a maiden dividend, where as current payers are those firms who are neither regular payers nor initiators. Non-payers are further categorized into never paid, former payers and current non-payers. Never paid firms are those that have never paid even a single dividend, where as former payers are those firms which at some previous point had paid dividends. Current non-payers are those firms which are recently listed and that they are neither former-payers nor are in the never paid category in any of the previous years.

3.3 Influence of Tax Regime Change: Test of Trade-off Theory

Paired samples t-test has been employed to analyze the influence of changes in dividend tax during 199798 on the dividend propensity of Indian corporate firms. According to the tradeoff theory, corporate firms pay more dividends when the dividend tax is low compared to that of capital gains tax. The tax regime ushered in during 1997-98, whereby dividends are taxed at source at a uniform rate of 10%, has tilted the balance in favor of dividends. Changes in dividends are captured with the help of two measures dividend per share and dividend payout percentage. For this purpose total dividend per share and average dividend payout percentage during the previous tax regime, i.e., the incidence of dividend tax is on the investors are compared with that of changed tax regime where dividend taxes are payable by corporate firms at a flat rate of 10%. The period 1994-95 to 1996-97 constitutes the first sub-period and the period 1998-99 to 2000-01 constitutes the second period. The following hypotheses are tested using paired samples t-test: (i) Null hypothesis of no differences between the total dividend per share between the two periods; and (ii) Null hypothesis of no difference between the average percentage payout between the two periods. Further, changes in the propensity of regular payers and changes in the payment pattern between 1996-97 and 1998-99 as a result of change in tax regime are also tested.

3.4 Characteristics of Payers and Non-Payers

Consistent with Fama and French, logit regression coefficients are estimated to analyze the influence of firm characteristics on the dividend payment pattern, for each year t during 1990-2001. The dependent variable assumes a value of 0 when the firm pays no dividend and assumes a value of 1 when pays a dividend. The explanatory variables are: Et/At is profitability measured as the ratio of aggregate earnings before interest to aggregate assets; dAt/At, is growth rate of assets; Vt/At is market-to-book ratio i.e., the ratio of the aggregate market value to the aggregate book value of assets; and the NSEPt is the percent of firms with the same or lower market capitalization. Coefficients are computed for each of the year

and the aggregate coefficients and associated t values are analyzed to infer the influence o profitability, f growth and size.

3.5 Test of Signalling Hypothesis: Case of Dividend Initiations and Omissions

For this part of the analysis, a firm is classified as initiator if it has paid dividend in the current year but has not paid dividends for the preceding 3 years. Similarly a firm is categorized as omission firm, if the firm has not currently paid dividend but has paid dividend in the preceding three years. To analyze signaling hypothesis, consistent with Healey and Palepu, earnings patterns of firms initiating and omitting dividend for 3 years before the year of event and 3 years after event are examined. To aggregate results across firms, earnings changes in these years are expressed as a percentage of the previous years closing stock price, PJ. The standardized change in earnings for firm j in year t, is defines as
E j ,t = E j ,t E j ,t 1 Pj

Where Ej,t are earnings per share before extraordinary items and discontinued operations9 for firm j in year t. The null hypotheses of average earnings changes are zero is tested with the help of Dunnetts C (Post Hoc) test. Analysis pertaining to initiations and omissions only cover a particular sample of extreme events and excludes firms not having a dividend track record of less than 3 years. In order to cover other dividend events like dividend reductions and increases in the following we arrive at yet another sample.

3.6 Test of Signaling Hypothesis: Case of Dividend Reductions

To analyze the relationship between dividends and losses a sample is drawn with firms having consistent profitability and dividend track records during 1990 1995 and who have earnings and dividend information for the period 1996 2001. The importance of annual losses on dividend reductions and annual dividend omissions has been analyzed with the help of logit analysis. The dependent variable equals zero if a firm has maintained or increased its dividend per share and is equal to one if the firm announced a reduction in dividend per share. The loss dummy assumes a value of one if the firm reports a loss for the year under study and zero otherwise. The level of net income and changes in net income are standardized with the previous years net worth for each firm. For firms in loss sample, the initial loss year constitutes the event year where as for non-loss firms, the initial year of earnings decline constitutes the event year. Similarly to examine the influence of past and future levels of earnings logit analysis has been employed on the subset for event years 1997 and 1998. The dependent variable equals zero if a firm has maintained or increased its dividend per share and is equal to one if the firm announced a reduction in dividend per share. The explanatory variables are earnings in 1 year before the event (t-1), 2 years preceding the event (t-2), current earnings (t), earnings in the year following the event year (t+1), earnings in 2 years following the event (t+2). Similarly, mean difference in earnings over t 2 through t+2 years is also examined with the help of Dunnetts C test. This analysis would be useful in determining whether dividend changes are impacted by contemporaneous or lagged or expected earnings performance.

In the Indian context an approximate value for this is derived from other income.

4. Trends in Dividends and Influence of Changes in Tax Regime

Average profit after tax (PAT) has increased from Rs. 4.68 crore in 1990 to Rs. 6.11 crore in 2000 and Rs. 9.36 crore in 2001 (Table 4.1). However, there have been several fluctuations in average PAT reflecting the changes in Indian economy. In the early phases of economic reform, many firms had to restructure as the economy was opened up and structural adjustments were undertaken resulting in a reduction in PAT. The subsequent pick up in the mid -90s has seen an increase in average PAT. The late 1990s, which marked a significant decline in economic activity, have had their impact on PAT of firms.

4.1 Average Dividend Paid

Despite fluctuations in PAT, the average aggregate dividend payments have steadily increased from Rs. 0.99 crore in 1990 to Rs. 2.93 crore in 2000 and Rs. 4.19 crore in 2001. Further, compared to PAT the dividend payments have exhibited a smooth trend implying that dividend smoothening is occurring in the Indian context (Figure 4.1).
Table 4.1
Trend in Dividends and PAT During 1990-2001
Year Number of Firms 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Common Firms 1707 2184 2505 3097 4020 5115 5600 5855 5980 6248 6225 4766 871 Average Dividend Rs. Crore 0.99 0.98 1.11 1.11 1.27 1.56 1.85 2.05 2.26 2.39 2.93 4.19 SD of Average SD of Dividend PAT PAT Rs. Crore Rs. Crore Rs. Crore 3.92 4.68 48.45 3.79 4.05 37.88 4.54 4.19 40.45 4.85 3.06 46.76 6.19 4.15 51.41 8.42 6.96 57.55 10.80 7.19 62.92 13.91 6.38 65.65 17.18 5.69 103.52 22.14 5.09 88.19 26.46 6.11 103.54 44.71 9.36 134.39

Number of firms paid dividend during the study period have shown an up trend till 1995 and have fallen subsequently (Appendix Figure 4.1), where as the percentage of companies paying dividends has declined from 60.5 percent in 1990 to 32.1 percent in 2001 (Table 4.2 and Figure 4.2). This is consistent with the trend observed in the US market (Fama and French 2001). The fact that percentage of companies paying dividends have declined whereas the average dividend paid has increased implies tha t companies which have been paying dividend have paid higher amounts in recent years. Total non-payers have steadily increased from 1990 to 2000 before declining slightly in 2001 (Appendix Table A4.1 and Figures A4.2 and A4.3). Firms, which have never paid dividend, constituted a significant proportion through out the sample period constituting more than 50% from 1991 to 2001 continuously. The number of firms, which at some previous time paid dividend, have increased overtime and reached almost 50% of non-payers in 2001.
Figure 4.1

Trend in Average Dividends, and PAT During 1990-2001

Average Dividend Average PAT

10 9 8 7 6 5 4 3 2 1 0 1990 1992 1994 1996 1998 2000

Rs. Crores

Year

Table 4.2
Trend in Dividend Payments During 1990-2001
Year Paid Dividend No. 1033 1272 1533 1823 2333 2775 2723 2386 2101 2007 1988 1531 % 60.50 58.20 61.20 58.90 58.00 54.30 48.60 40.80 35.10 32.10 31.90 32.10 Not Paid Dividend No. 674 912 972 1274 1687 2340 2877 3469 3879 4241 4237 3235 % 39.50 41.80 38.80 41.10 42.00 45.70 51.40 59.20 64.90 67.90 68.10 67.90 Total Number of Firms 1707 2184 2505 3097 4020 5115 5600 5855 5980 6248 6225 4766

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Total number of firms paying dividend has increased up to 1995 and has registered sustained decline there after (Table 4.2, Appendix Figures A4.4 and A4.5). Mirroring these trends firms, which have paid dividends regularly, peaked in 1995 and recorded declines thereafter. Initiators have shown a steady decline from 1991 and have fallen to 5% in 2001. Average dividend paid by payers has increased steadily from Rs. 1.69 crore in 1991 to Rs. 9.16 crore in 2000 and Rs. 13.05 crore in 2001 (Figure 4.3, Appendix Table A4.2). Regular payers are more in number and have paid higher average dividend compared to that of current payers and initiators (Appendix Figures A4.6 and A4.7). Current payers have paid higher dividend compared to initiators except in the year 2001. The number of initiators have increased up to the year 1995 and have shown a decline thereafter, where as current payers have steadily increased in number up to 2000.

10

Figure 4.2
Dividend Behaviour of Indian Corporate Firms During 1990 - 2001 (in %)
80% 70% 60% % Non-Payers % Payers

% of Firms

50% 40% 30% 20% 10% 0% 1990 1992 1994 1996 1998 2000

Year

Figure 4.3
Comparision of Average Dividend Paid During 1991 2001 by Payer Group
Initiator Current Payers Regular Payers Total Payers

20 15 10 5 0

Rs. Crore

1991

1993

1995

1997

1999

2001

Year

A comparison of index and non-index firms shows that the former group of companies on average has paid more dividend than the latter group (Table A4.3 and A4.4). Similarly, it is observed that companies, which constitute popular market indices such as Sensex and Nifty paid more dividends compared to companies in the broad market indices such as BSE 100, CNX Mid-Cap, BSE 200, CNX 500, and BSE 500. These observations are on the expected lines as higher dividend payment is one of the important criteria for inclusion of stocks into indices. A study of number of companies paying dividend also reveals that a significantly larger proportion of index firms have paid dividend compared to non-index firms. 29 out of 30 Sensex firms and 49 out of 50 Nifty firms have paid dividend in 2001, the exception being Tata Engineering and Locomotive Company Ltd. (TELCO). Analysis of industry-wise average dividend paid shows that in the early 1990s, firms in the diversified industry have paid more dividends followed by mining firms and electricity firms (Table 4.3). However, by the end of 2000 and 2001 firms in the electricity industry have paid more dividend followed by mining and diversified companies. It has also been observed that textile companies have continued to pay low amounts on an average throughout the sample period where as firms in the financial services industry have improved their average dividend payments over the sample period. The recent h growth firms in the computer igh

11

hardware and software segments, which are part of the machinery industry, have generally shown lower dividend payments. In sum, the number of firms paying dividend during the study period have shown an up trend till 1995 and have fallen subsequently. Further, compared to PAT the dividend payments have exhibited a smooth trend implying that dividend smoothening is occurring in the Indian context. Regular payers are more in number and have paid higher average dividend compared to that of current payers and initiators. Of the nonpayers, former payers are growing in numbers. Index firms appear to pay higher dividends compared to that of non-index firms. Further, smaller indices appear to have higher average dividend compared t that of o larger indices. Industry trends indicate that firms in the electricity, mining and diversified industries have paid more dividend where as textile companies have paid less dividends. Firms in the machinery industry which includes computer hardware and software segments have shown lower dividends.
Table 4.3
INDUSTRY Average Dividend Paid During 1990-2001 Industry-wise (in Rs. Crore) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
1.09 3.56 1.28 .67 .88 .70 .80 2.57 .39 .50 1.02 .48 1.25 .96 3.88 1.14 1.39 .97 .65 .90 2.79 .51 .62 .76 .47 1.17 1.05 4.24 1.19 1.47 .98 .72 1.37 2.97 .72 .70 .86 .47 1.20 .97 5.11 2.26 1.38 .89 .73 1.36 3.57 .62 .64 .92 .53 1.06 1.08 6.14 5.85 1.49 .94 .83 1.72 2.87 .73 .63 1.01 .72 1.39 1.38 1.57 1.69 1.92 7.72 10.13 10.99 12.86 9.54 13.08 18.31 17.37 2.10 2.46 2.72 3.16 1.02 .80 .90 1.12 .99 1.11 1.13 1.20 2.20 2.39 2.14 1.80 2.94 8.87 17.44 22.23 .70 .75 .57 .35 .85 1.18 1.00 .86 1.07 1.18 1.23 1.34 .86 .82 .58 .51 2.02 2.83 3.58 3.18 1.68 17.17 26.33 3.20 1.13 1.34 1.40 21.99 .56 .90 1.34 .48 2.95 2.41 22.76 27.24 4.25 1.34 1.58 1.72 26.31 .58 1.12 1.42 .56 3.44

2001 Firms
2.46 29.55 48.67 5.29 1.89 2.11 3.08 35.36 1.05 1.51 4.07 .56 3.03 1138 184 58 1097 745 1065 555 81 324 296 1264 750 225

Chemicals and Plastics Diversified Electricity Financial Services Food and Beverages Machinery Metals and Metal Product Mining Misc. Manufacturing Non-Metallic Mineral Pro Other Services Textiles Transport Equipment

4.2 Dividend Per Share

Average dividend per share (DPS) has increased from 14 paisa in 1990 to 26 paisa in 2000 and 15 paisa in 2001 (Table 4.4, Figure 4.4). An analysis of distribution of firms shows that 39 percent have paid nil DPS in 1990 and the percentage has increased to 67.7 in 2001 (Table 4.5). Percentage of firms in the average class i.e., DPS in the range of Rs. 0 to Rs. 0.25 have declined from a high of 45.9 in 1990 to 18.5 in 2001. This implies that the increased average DPS over the latter period has mainly been due to a few firms paying larger DPS. Firms in chemicals and plastics industry have steadily improved their DPS from 14 paisa in 1990 to 27 paisa in 2000 and 25 paisa in 2001 (Table 4.6). Where as textiles firms have shown a decline in DPS from 13 paisa in 1990 to 6 paisa in 2001. Machinery firms have paid a steady 12 to 14 paisa except for the years 1996 and 1997 when they paid marginally more. An analysis of index and non-index firms DPS shows that index firms on an average paid more DPS than non-index firms (Table A4.14). Similarly, narrow indices have high average DPS than broad indices.

12

Table 4.4
Average Dividend Per Share (DPS) During 1990-2001
(in Rs.) Year Number Minimum Maximum of Firms DPS DPS 1990 1694 0 12.71 1991 2153 0 10.58 1992 2468 0 15.58 1993 3028 0 51.2 1994 3953 0 57.5 1995 5032 0 135.33 1996 5536 0 174.67 1997 5801 0 222 1998 5911 0 350.33 1999 6176 0 249.75 2000 6167 0 266.38 2001 4734 0 61.5 Common 866 Firms10 Average DPS 0.1406 0.1385 0.1427 0.1514 0.1582 0.1803 0.2158 0.198 0.2337 0.2544 0.2571 0.1538 Std. Deviation 0.3455 0.3009 0.3568 1.0025 1.2983 2.3543 3.3243 3.4834 5.8833 4.8938 4.4156 1.2899

Average DPS (1% trimmed) by all payers have increased from 21 paisa in 1991 to 31 paisa in 2000 and 29 paisa in 2001 (Figure 4.5). Of the payers, regular payers have consistently paid more dividend per share compared to other payers. Similarly initiators have always paid lower dividend per share compared to current payers.
Figure 4.4
Average Dividend Per Share (DPS) During 1990-2001 Average DPS (in Rs.)
Average DPS

0.30 0.25 0.20 0.15 0.10 0.05 0.00 1990 1992 1994 1996

1998

2000

Year

An analysis of recurrence of dividend per share group shows that two firms have consistently paid dividend in the range of 25 to 50 paisa per share for all the 12 years, where as 18 firms have paid up to 25 paisa (Appendix Table A4.6 and A4.7). An analysis of dividend reductions by firms shows that only five companies namely Mahindra Sintered Products Ltd, Otis Elevator Co. (India), Bharat Electronics, Amritlal Chemaux, and Carborundum Universal have consistently paid higher dividend per share out of a 330 firms that paid dividends in all years of the sample period (Appendix Table A4.5). 43 firms registered a single instance of dividend per share reduction, where as 68 firms lowered twice, 82 firms lowered thrice etc. On the whole average DPS has shown a steady growth except in the year 2001. Regular payers have consistently paid more dividend per share compared to other payers, where as initiators have always paid
5 common firms are lost on account of missing information on number of outstanding stocks and hence there is difference in the number of common firms from that of Table 4.1.
10

13

lower dividend per share. Analysis also shows that only a few firms have consistently paid same levels of dividend. Index firms on an average paid more DPS than non-index firms. Similarly, narrow indices have high average DPS than broad indices (Appendix table A4.8). Firms in chemicals and plastics industry have steadily improved their DPS, where as textiles firms have shown a decline in the study period. Machinery firms have paid a steady DPS.
Figure 4.5
1% Trimmed Dividend Per Share by Payer Type
Current Payers Initiators Regular Payers Total

0.35 0.3

DPS (in Rs.)

0.25 0.2 0.15 0.1 0.05 1991 1993 1995 1997 1999 2001

Year

Table 4.5
Distribution of Firms in terms of Dividend Per Share During 1990 2001
DPS Rs. 0 Rs. 0 0.25 Rs. 0.25 0.50 Rs. 0.50 0.75 Rs. 0.75 1 Rs. 1 2 Rs. 2 5 > Rs. 5 Percentage of Companies in Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 39 41 37.9 39.9 41.1 44.9 50.8 58.9 64.5 67.5 67.8 67.7 45.9 43.1 46.2 46.9 45 42.3 35.8 27.5 22.2 19.5 18.6 18.5 13.5 13.7 13.7 11.2 12.1 10.6 10.4 9.8 8.7 7.6 7.4 7.8 0.9 1.3 1.4 0.9 0.7 1.1 1.5 2.3 2.8 2.5 2.6 2.7 0.4 0.5 0.4 0.7 0.8 0.4 0.6 0.6 0.6 1.1 1.2 1.3 0.2 0.3 0.3 0.2 0.2 0.3 0.4 0.6 1 1.1 1.4 1.4 0.1 0.1 0 0.1 0.1 0.2 0.2 0.1 0.2 0.3 0.6 0.4 0.1 0 0 0.2 0.1 0.1 0.2 0.2 0.2 0.3 0.4 0.3

Table 4.6
Industry-wise Dividend Per Share (DPS) During 1990-2001 (in Rs.)
INDUSTRY Chemicals and Plastics Diversified Electricity Financial Services Food and Beverages Machineray Metals and Metal Product Mining Misc. Manufacturing Non-Metallic Mineral Pro Other Services Textiles Transport Equipment 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 FIRMS .14 .15 .14 .12 .17 .15 .12 .17 .17 .18 .27 .25 1138 .19 .21 .26 .20 .20 .19 .21 .22 .21 .22 .27 .21 184 .13 .10 .11 .11 .11 .10 .12 .09 .10 .10 .13 .10 58 .08 .11 .13 .34 .24 .21 .28 .12 .15 .14 .19 .18 1097 .20 .20 .18 .23 .31 .47 .49 .58 .85 .21 .16 .13 745 .12 .13 .14 .14 .13 .13 .17 .19 .12 .14 .14 .14 1065 .13 .11 .11 .09 .10 .10 .12 .09 .07 .06 .07 .07 555 .05 .07 .06 .07 .09 .06 .07 .08 .13 .10 .11 .09 81 .12 .12 .14 .10 .11 .10 .10 .15 .06 .16 .21 .30 324 .10 .11 .11 .09 .09 .09 .10 .08 .08 .07 .09 .09 296 .17 .15 .17 .15 .13 .24 .38 .28 .42 .88 .73 .12 1264 .13 .14 .13 .11 .12 .09 .08 .06 .06 .05 .07 .06 750 .12 .12 .12 .12 .13 .13 .15 .18 .16 .15 .21 .17 225

14

4.3 Dividend Payout Ratio

An analysis of average percentage dividend payout (PR) during 1990 2001 shows a volatile trend (Table 4.7 and Figure 4.6). Percentage PR increased from 27.39 in 1990 to 32.95 in 1997 and then showed a declining trend till 2000 before reaching the peak average percentage PR of 40.53 in 2001. However, 1% trimmed average percentage PR showed a more stable pattern of around 24 percent PR up to 1997 and then has shown a declining trend before finally reaching 16.81 percent in 2001 (Appendix Table A4.9).
Table 4.7
Average Percentage Payout During 1990 2001
Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 No. of Average Std. Firms % Payout Deviation 1382 1714 2022 2533 3156 3770 4042 4258 4335 4503 4383 3387 27.39 25.19 27.54 27.98 28.19 25.88 27.44 32.95 31.39 22.82 21.6 40.53 37.77 41.04 48.31 37.83 61.96 38.06 88.12 139.85 453.37 120.19 67.49 1196.96 1% Trimmed Average % Payout 24.98 23.11 24.25 25.72 24.92 23.84 23.99 23.91 18.64 16.98 17.47 16.81 1% Trimmed No. of Firms 1369 1697 2002 2508 3125 3733 4002 4216 4292 4458 4340 3354

An analysis of distribution of firms by dividend payout percentage shows that as high as 26 percent of firms in 1990 and 56.6 percent in 2001 have paid out nothing (Table 4.8 and Appendix, Figure A4.6). However, more than 10 percent firms have paid dividend in excess of 75 percent of their net profits. An analysis of dividend payout recurrence shows that very few firms have maintained the same payout for a longer period of time (Appendix Table A4.10 and A4.11). For instance, only one firm Hindustan Lever Limited has paid out a dividend in the range of 50 to 75% of its net profit for entire sample period. Similarly another firm Maharashtra Scooters Limited - maintained a dividend payout in the range of 10 to 20% for 11 of the 12-year sample period. Similarly, Kinetic Engineering Ltd., Lakshmi Machine Works Ltd., and Dalmia Cement (Bharat) Ltd. have paid out in the range of 10 20% for 10 of the 12-year sample period.
Figure 4.6
Average % Payout During 1990-2001
Average % Payout 50 40 30 20 10 0 1990 1992 1994 1996 1998 2000 1% Trimmed Average % Payout

Average Payout %

Year

15

An analysis of industry-wise DPO shows a declining trend across all industries during the sample period (Table 4.9). Diversified firms, which have a DPO in excess of 25 percent in 1990, have less than 14 percent in 2001. Firms in metals and metal products industry have registered a high degree fall in DPO from 22.84 percent in 1990 to 8.74 percent in 2001.
Table 4.8
Distribution of Firms Payout Percentage During 1990 - 2001
Dividend Payout % 0 0 - 10 10 - 20 20 - 30 30 - 40 40 - 50 50 - 75 75 - 100 100 - 200 > 200 Firms % of Firms 1990 1991 26 6.9 14.5 16.5 12.6 8.2 10.1 3.5 1.2 0.4 1382 1992 1993 28.9 7.2 11.9 13.5 12.3 9.5 10.5 4.6 1.3 0.4 2533 1994 26.6 8 14.3 15 12.4 7.7 10.2 4.5 0.9 0.3 3156 1995 26.7 6.6 15.6 16.7 12.5 8.7 8.6 3.4 0.9 0.3 3770 1996 33.3 5.5 13.6 13.7 10.8 7.3 8.6 5.4 1.4 0.4 4042 1997 1998 1999 2000 2001 45.4 52.8 57 55.8 56.6 3.1 3.4 3.4 3.8 3.8 7.9 7.6 6.7 6.6 7.6 10.9 9.8 8.2 8.9 7.9 8.5 7.5 6.9 6.7 6.9 6.4 5.4 5.2 5.4 4.8 9.1 7.8 6.7 6.5 7.1 5.2 3.2 3.9 4.2 3.2 2.1 1.6 1.3 1.5 1.5 1.3 1 0.7 0.7 0.7 4258 4335 4503 4383 3387

26.5 25.3 9.3 9.2 14.1 13.9 17.2 16.1 12.6 13.3 7.1 8.8 9 8.9 2.9 2.7 0.9 1.4 0.2 0.4 1714 2022

Table 4.9
Industry-wise Dividend Payout During 1990 2001 (in %)
INDUSTRY Chemicals and Plastics Diversified Electricity Financial Services Food and Beverages Machineray Metals and Metal Product Mining Misc. Manufacturing Non-Metallic Mineral Pro Other Services Textiles Transport Equipment 1990 23.92 25.28 17.98 23.28 24.47 23.93 22.84 10.28 18.10 19.71 20.01 16.83 19.31 1991 20.38 20.95 16.21 27.01 23.15 20.36 21.47 7.29 18.08 17.75 21.15 15.98 19.96 1992 21.51 22.78 14.15 28.50 24.19 22.87 19.86 12.28 15.69 16.95 19.25 17.26 21.61 1993 23.38 25.48 13.37 32.11 22.14 23.42 20.65 9.56 17.18 16.27 19.84 20.98 21.29 1994 20.14 22.74 12.48 29.87 20.40 23.67 20.92 14.04 17.87 14.78 21.15 20.54 23.26 1995 21.88 23.23 16.98 27.25 17.01 22.07 19.76 12.10 18.91 14.92 19.60 19.20 20.99 1996 20.53 21.61 12.70 31.74 17.23 20.83 18.82 16.58 17.81 13.87 19.34 17.30 19.69 1997 18.37 23.27 16.32 29.19 16.14 19.45 16.78 14.65 15.55 13.62 17.43 13.84 22.46 1998 14.76 19.34 10.42 16.12 12.73 16.28 12.56 11.50 9.84 10.78 14.00 11.29 20.96 1999 13.84 17.41 9.35 14.82 12.67 15.36 9.37 9.87 12.18 9.66 12.27 7.99 18.74 2000 14.18 17.52 12.68 16.21 12.80 15.24 9.16 11.98 12.59 8.93 12.85 9.04 20.18 2001 13.71 13.59 13.08 14.30 10.22 15.15 8.74 11.76 15.09 11.29 12.54 8.02 17.29

Total payers have registered an increase in payout from 31.25% in 1991 to a peak of 43.02% in 1997 and finally paid out 37.64% in 2001 (Figure 4.7 and Appendix Table 4.12). Of the payers, regular payers have consistently paid higher payout compared to that of current payers. Further, initiators have shown higher fluctuations in their payout compared to that of regular payers. In sum, average percentage PR showed a more stable pattern up to 1997 and then has shown a declining trend. Analysis of dividend payout recurrence shows that very few firms have maintained the same payout for a longer period of time. Industry-wise DPO shows a declining trend across all industries during the sample period. Of the payers, regular payers have consistently paid higher payout compared to that of current payers. Further, initiators have shown higher fluctuations in their payout compared to that of regular payers.

16

Figure 4.7
1% Trimmed Dividend Payout % by Payer Type
Current Payers Regular Payers 50

Initiators Total Payers

% Payout

45 40 35 30 25 20 1991 1993 1995 1997 1999 2001

Year

4.4 Dividend Yield

Average dividend yield for all companies during the period 1991 to 2001 has declined from 1.73% in 1991 to .55 in 1993 before finally recovering to 1.61 in 1998 and again falling marginally to 1.24% in 2001 (Table 4.10 and Figure 4.8). On the whole the dividend yield is range bound in the region of 0.5% to 1.73%. The reason for the fall in 1993 could be due to high increases in market capitalizations of a number of stocks in the face or irregularities in the stock market in 1992. Analysis of dividend yield by type of payer shows that initiators have always paid higher levels of dividend yield compared to that of current payers and regular payers (Figure 4.9, and Appendix Table A4.23). Similarly current payers have paid higher dividend yield compared to that of regular payers. Dividend yields of initiators have declined from 6% in 1991 to 1.51% in 1993 before recovering and reaching an all time high of 10% in 1998. Compared to this current payers yielded about 5% in 1992 before falling to 1.81 in 1993 and have subsequently recovered and reached all time high of 8.12% in 2000. On the other hand regular payers started with a yield of close to 5% but have fallen to a low of 1.5 in 1993 before reaching an all time high of 7.76% in 2000.
Table 4.10
1% Upper Trimmed Dividend Yield (%)During 1991 - 2001
Year Mean Median SD Firms 1991 1.73 .0 2.74 1452 1992 1.66 .0 2.57 1603 1993 0.55 .0 0.94 1989 1994 1.68 .0 3.02 2559 1995 1.44 .0 2.85 3481 1996 1.01 .0 1.88 4214 1997 1.46 .0 2.99 4864 1998 1.61 .0 3.80 5049 1999 1.44 .0 3.86 5235 2000 1.43 .0 3.96 5182 2001 1.24 .0 3.15 4097 Note: Median values are considered only up to 1 decimal. However, there are non-zero values.

On the whole dividend yield of aggregate payers shows a significant increase from 1991 to 2001.

17

Average dividend yield has differed from industry to industry (Table 4.11). Diversified firms, followed by firms in electricity, food and beverages and textiles industries paid higher dividend yields in 1991 while financial services and mining firms paid the lowest. By 2001 diversified firms and electricity continue to pay higher dividend yields where firms in transport industry have improved their dividend yields by 2001. However, food and beverages and textile firms recorded lowered their dividend yield by 2001, where as firms in financial services, and mining have improved their dividend yields.
Figure 4.8
1% Upper Trimmed Dividend Yield During 1991 2001
2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1991 1993 1995 1997 1999 2001

Average (%)

Year

Figure 4.9
1% Upper Trimmed Dividend Yield by Payer Type
Current Payer 12 Initiator Regular Payer Total

Average (%)

10 8 6 4 2 0 1991 1993 1995 1997 1999 2001

Year

On the whole the dividend yield is range bound during the study period. Analysis of dividend yield by type of payer shows that initiators have always paid higher levels of dividend yield compared to that of current payers and regular payers. Diversified firms and firms in the electricity industry have paid higher dividend yields during the study period.

4.5 Summary of Analysis of Dividend Trends

The number of firms paying dividend during the study period has shown an up trend till 1995 and has fallen subsequently. Average DPS on the other hand has shown a steady growth e xcept for year 2001. Average percentage PR showed a more stable pattern up to 1997 and then has shown a declining trend. Dividend yield measure is range bound.

18

Analysis also shows that only a few firms have consistently paid same levels of dividend. Analysis of dividend payout recurrence shows that very few firms have maintained the same payout for a longer period of time. Of the payers, regular payers have consistently paid higher payout as well as higher average dividend compared to that of current payers. Iinitiators have always paid higher levels of dividend yield compared to that of current payers and regular payers. Further, narrower indices appear to have higher dividends compared to that of broader indices. Industry trends indicate that firms in the electricity, mining and diversified industries have paid higher dividends where as textile companies have paid less dividends. Firms in the machinery industry which includes computer hardware and software segments have shown lower dividends.
Table 4.11
Average Dividend Yield (%) Industry-Wise During 1991 - 2001
Industry Chemicals and Plastics Diversified Electricity Financial Services Food and Beverages Machinery Metals and Metal Product Mining Misc. Manufacturing Non-Metallic Mineral Products Other Services Textiles Transport Equipment 1991 1.79 2.97 2.27 0.2 2.18 1.66 1.76 0.11 1.41 1.74 1.18 2.06 1.53 Average 1% Upper Trimmed Dividend Yield in Year 1992 1993 1994 1995 1996 1997 1998 1999 1.92 0.55 1.68 1.39 0.99 1.55 1.91 1.82 2.49 0.8 2.64 1.56 1.3 2.16 2.44 2.12 1.31 0.69 1.49 1.04 1.14 1.07 0.93 0.85 0.9 0.41 2.28 1.98 1.45 1.87 1.29 1.05 2.06 0.58 1.4 0.92 0.7 1.21 1.63 1.38 1.55 0.61 1.8 1.57 1.07 1.54 1.87 1.7 1.81 0.53 1.62 1.71 1.15 1.43 1.33 1.22 0.05 0.01 0.02 0.21 0.52 0.45 0.56 1.12 0.98 0.33 1.51 1.32 0.89 1.18 1.35 1.74 1.55 0.49 1.15 1.02 0.86 1.08 1.36 1.46 1.37 0.5 1.33 1.3 0.81 1.23 1.33 0.97 1.8 0.62 2.08 1.72 1 1.41 1.74 1.48 1.48 0.55 1.61 1.36 1.22 1.97 2.42 2.24 2000 1.66 2.99 1.47 1.33 1.12 1.32 1.29 0.58 1.34 1.66 1.05 1.65 2.76 2001 1.35 2.11 1.99 1.03 1.06 1.01 1.2 0.81 1.29 1.43 0.98 1.6 2.04

4.6 Changes in Tax Regime and Dividend Propensity

Analysis of influence of change in tax regime on dividend propensity shows that total dividend per share has come down from an average of Rs. 0.84 to Rs. 0.71, where as average payout percentage has increased from 33.33% to 51.05% (Table 4.12). Mimicking the trends for total firms, regular payers have registered lower DPS and higher payout percentage. As opposed to these changes over sub-periods of 3 years before and after the change in tax regime, one year changes show that DPS has more or less remained at the same level, where as payout percentage has come down from 1997 to 1999. However, paired samples t-test shows that these differences are not statistically significant, except in the case of payout percentage from 1997 to 1999 (Table 4.13). In sum, it can be inferred from the present study that tax regime changes have not really influenced the dividend behavior of Indian corporate firms and that the tradeoff theory does not hold true in the Indian context.

19

Average Dividends Before and After the Tax Regime Change

Variable Total DPS (in Total Firms Rs) Regular Payers Total DPS (in Rs.) Immediate DPS (in Rs.) Years Average Total Firms Payout % Average Regular Payers Payout % Immediate Payout % Years Sample After Before After Before 1999 1997 After Before After Before 1999 1997 Mean .71 .84 1.55 1.72 .22 .22 51.05 33.33 60.53 38.07 27.78 35.87 N 2597 2597 765 765 4848 4848 1217 1217 1000 1000 2987 2987 SE Correlation .17 .519 .24 .27 .241 .71 .06 .426 .05 19.19 .015 1.43 23.35 .008 1.68 2.65 .072 2.87 Sig. .000 .000 .000 .610 .795 .000

Table 4.12

Influence of Change in Tax Regime on Dividend Propensity: Paired Samples T-test

Difference SE After - Before Total Firms -.13 .21 Total DPS Regular Payers -.17 .70 (in Rs.) Immediate Years .01 .06 Total Firms 17.72 19.23 Average 22.46 23.39 Payout % Regular Payers Immediate Years -8.09 3.76 t -.62 -.24 .11 .92 .96 -2.15 df 2596 764 4847 1216 999 2986 Sig. .536 .810 .909 .357 .337 .032

Table 4.13

5. Characteristics of Dividend Payers and Non-Payers 5.1 Profitability

Payers on an average have more than twice the payoff on assets compared to that of non-payers (Table 5.1). This finding is consistent with Fama and French (2001). Of the payers Initiators appear to have on an average higher payoff on assets compared to current payers and regular payers, though their payoffs on assets have shown considerable fluctuations. Current payers and regular payers have similar levels of payoff on assets. Of the non-payers, former payers appear to have higher payoff on assets compared to firms, which never paid dividends. Never paid in turn appears to higher payoff on assets compared to current non-payers. An analysis of EPS of payers and non-payers shows that the former have on an average higher EPS compared to the latter. The difference in magnitude is also quite substantial compared to that of payoff on assets. Of the payers, regular payers have consistently higher EPS compared to that of the other two groups of payers. EPS of current payers and initiators has shown considerable fluctuations over the sample period. Initiators have higher average EPS in the early part of 1990s and last few years of 1990s, where as in the intervening years their EPS has shown a decline. Current payers on the other hand shown an opposite trend compared to that of initiators. All the non-payer groups have shown considerable fluctuations in EPS during the sample period and on average registered a decline in EPS from 1990 to 2001. An analysis of common stock earnings to book equity

20

shows that on an average payers have dominated non-payers as the former firms registered 24% in 1991 and 15% in 2001 to 4% and 6% by the latter in the corresponding years. Of the payers, initiators have higher common stock earnings to book equity compared to that of regular payers and current payers. Regular payers and current payers have similar equity earnings to book equity. However there is a gradual decline in earnings to book equity from 1991 to 2001. Of the non-payer firms, never paid firms appear to have higher equity earnings to book equity compared to current non-payers and former payers. The difference between payers and non-payers is larger in terms of stock earnings to book equity compared to payoff on firms assets. These findings are consistent with Fama and French. To sum up it can be concluded that profitability has positive influence on the dividend payment of a corporate firm. Dividend payers are more profitable compared to non-payers. Further, corporate firms in general and non-dividend payers in particular have become less profitable.

5.2 Growth or Investment Opportunities

An analysis of growth of assets shows that payers on an average have higher growth compared to that of non-payers. Payers have grown at percentages of 29.03 in 1991, 23.69 in 2000 and 10.82 in 2001 compared to 18.65, 4.12 and 1.86 in the corresponding years for non-payers. Of the payers initiators appear to have higher growth percentage compared to that of regular payers. Initiators have grown at percentages of 29.87 in 1991, 49.13 in 2000 and 57.54 in 2001 compared to 28.92, 23.59 and 6.78 in the corresponding years for regular payers. Regular payers in turn appear to have higher growth compared to that of current payers. Of the non-payers, never paid have on an average lower growth in assets compared to former payers and current payers. These findings are not consistent with Fama and French where they find never paid firms to have higher growth in assets compared to that of other non-payer and payer groups. Similar trends are observed with regard to growth opportunities as measured by R&D investment to total assets. Payers appear to have higher growth opportunities compared to non-payers. Of the payers, regular payers have higher growth opportunities compared to initiators and current payers. Of the non-payers, never paid appears to have lower growth opportunities compared to current non-payers. However the percentage growth opportunities for payers as well as for non-payers are considerably low as the payers on an average have 0.02% in 1991 and 0.27% in 2001 compared to 0.003% and 0.0447% in the corresponding years for non-payers. An analysis of aggregate market value to book value of assets shows that payers and non-payers do not differ significantly. However, there are differences with in the payer and non-payer groups. For instance, initiators appear to have higher market value to book value compared to regular and current payers, where as in non-payer group, former payers appear to be dominated by both never paid and current non-payers. On the whole in the Indian context higher growth and growth opportunities have not resulted in lower dividend payments by corporate firms. This finding contradicts the findings of Fama and French, whereby they contend that growth opportunities are an important reason for reduced dividend payments by firms.

5.3 Size
Dividend payers appear to be much larger in size compared to that of non-payers. This observation is consistent with Fama and French (2001). Average size as measured by assets of payers averaged Rs. 104.4 crore in 1991 and Rs. 1413.43 in 2001 compared to that of Rs. 56.92 and Rs. 181.20 in the corresponding years for non-payers.

21

Of the payers, regular payers have higher assets compared to that of current payers. Current payers in turn have higher assets compared to initiators. Similarly, regular payers have grown an average asset base of Rs. 112 crore in 1991 to Rs. 1711 crore in 2001 compared to Rs. 54.71 crore and Rs. 581.48 core for initiators and Rs. 47.11 crore in 1992 and Rs. 654.9 crore for current payers. Of the non-payers, former payers appear to have higher assets compared to current never paid who in turn have higher asset base compared to current non-payers. Asset base of former payers has grown from Rs. 90.14 crore in 1991 to Rs. 239.2 crore in 2001 while in the corresponding period never paid have grown from Rs. 51.69 crore to Rs. 80.57 crore. However, current non-payers have registered a decline in their asset base from Rs. 43.5 crore to Rs. 18.73 crore during the same period. An analysis of indebtedness of firms s hows that non-payers appear to have higher levels of long-term borrowings to assets compared to that of payers. Of the non-payers, never paid appears to have higher longterm borrowings to assets compared to former payers, who in turn appear to have higher levels compared to current non-payers. Of the payers, regular payers appear to have higher long-term borrowings to assets compared to current payers. Current payers in turn have higher levels compared to initiators. On the whole, the size of assets of firms have gone up during the period 1990 2001 and that increased assets seems to have been financed through long-term borrowing implying pecking order of preference for funds.
Table 5.1
Characteristics of Dividend Payers and Non-Payers
Year 1991 1992 1993 Average % Payoff on Assets Current Payers 11.20 12.23 Initiators 9.79 15.15 12.57 Regular Payers 11.69 12.03 12.00 Total Payers 11.44 12.32 12.07 Current Non-Payers 6.58 5.16 3.69 Former Payers 10.24 7.41 6.23 Never Paid 4.44 6.71 5.29 Total Non-Payers 5.49 6.68 5.29 Average 1% Trimmed EPS Current Payers 3.20 4.83 Initiators 7.05 7.47 5.49 Regular Payers 14.11 12.79 9.07 Total Payers 13.20 11.97 8.46 Current Non-Payers -1.61 -1.18 -0.49 Former Payers 0.71 -2.72 -3.45 Never Paid 0.07 1.41 -0.88 Total Non-Payers 0.04 0.49 -1.41 Average Common Stock Earnings to Book Equity % Current Payers 21 18 Initiators 29 39 27 Regular Payers 22 20 19 Total Payers 24 24 21 Current Non-Payers -15 -7 -41 Former Payers 8 -27 58 Never Paid 14 23 47 Total Non-Payers 4 13 23 Average % Growth (Assets) Current Payers 46.25 27.29 Initiators 29.87 92.24 66.77 Regular Payers 28.92 62.44 32.20 Total Payers 29.03 63.66 33.40 Current Non-Payers 16.13 2.34 26.55 1994 12.67 15.19 12.24 12.58 3.16 5.37 4.91 4.79 7.30 4.53 9.37 8.67 -0.35 -1.64 -0.62 -0.81 23 32 21 24 13 72 14 21 27.95 50.41 36.31 36.17 46.48 1995 13.99 13.66 12.21 12.56 1.99 5.94 5.73 5.41 6.95 3.98 8.90 8.15 0.28 0.51 0.59 0.54 20 26 22 23 4 -65 10 -3 1996 12.27 11.25 12.02 11.99 3.67 9.06 3.89 5.61 6.81 4.01 8.58 8.02 0.19 1.57 0.12 0.60 20 25 19 20 -6 -26 -7 -11 1997 11.38 10.86 11.82 11.71 2.36 4.81 3.19 3.88 5.15 3.88 8.52 7.82 0.36 -0.73 -0.41 -0.54 15 22 16 16 -4 -36 -6 -15 38.62 51.10 54.26 51.63 1998 11.44 2.56 11.38 11.16 1.71 1.89 2.51 2.18 4.98 7.10 9.15 8.29 1.32 -2.58 -0.92 -1.77 14 25 16 16 50 -8 2 1 16.95 41.09 21.24 20.89 1999 9.98 17.02 11.31 11.18 6.30 0.05 0.63 0.31 6.17 6.19 9.57 8.78 0.83 -4.34 -1.10 -2.99 14 25 15 16 16 31 4 16 15.75 57.68 26.15 24.70 2000 10.02 14.95 11.17 11.02 -5.81 -1.52 -0.17 -0.97 6.99 4.76 11.69 10.30 -3.42 -4.38 -1.08 -3.02 15 21 17 17 15 -46 -2 -20 20.92 49.13 23.59 23.69 2001 10.39 14.20 11.56 11.38 -3.63 -0.04 2.65 0.94 7.04 4.07 11.78 10.33 -0.22 -4.31 -0.42 -2.87 15 23 14 15 50 -17 5 -6 17.47 57.54 6.78 10.82

55.23 39.16 93.31 2908.81 61.54 19.70 62.42 145.56 0.00

22

Table 5.1
Characteristics of Dividend Payers and Non-Payers
Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Former Payers 29.50 36.26 6.65 12.65 26.38 11.89 32.97 5.30 0.52 0.55 1.01 Never Paid 16.37 39.77 9.59 14.74 32.09 20.13 281.29 10.98 5.06 9.80 3.47 Total Non-Payers 18.65 38.05 9.01 14.46 30.35 16.91 164.07 7.95 2.33 4.12 1.86 Average % Aggregate Market Value to Book Value of Assets Current Payers 151.43 74.39 113.22 122.15 82.86 56.38 58.75 95.41 198.87 61.78 Initiators 113.22 301.63 101.57 227.90 193.18 200.89 79.02 125.80 323.85 688.46 212.80 Regular Payers 117.01 219.69 111.52 153.81 147.09 118.01 78.10 81.34 148.16 165.82 87.37 Total Payers 116.51 226.08 108.24 157.58 149.38 119.84 74.68 78.12 141.18 191.89 86.00 Current Non-Payers 119.83 280.04 125.33 113.40 138.36 171.23 74.53 63.71 171.59 27.76 Former Payers 68.11 121.58 65.79 78.82 84.93 68.29 45.55 41.31 49.49 64.09 47.27 Never Paid 146.19 247.66 105.09 130.99 134.11 271.69 79.16 60.47 68.57 105.56 664.76 Total Non-Payers 132.19 230.35 98.84 116.90 123.39 202.44 64.61 50.55 57.12 81.14 272.24 Average % Growth Opportunities (R&D to Assets) Current Payers 0.01 0.17 0.09 0.16 0.18 0.13 0.14 0.11 0.08 0.22 Initiators 0 0.00 0.01 0.03 0.05 0.03 0.07 0.01 0.00 0.05 0.10 Regular Payers 0.03 0.05 0.13 0.16 0.19 0.22 0.22 0.22 0.27 0.18 0.29 Total Payers 0.02 0.04 0.11 0.12 0.16 0.19 0.19 0.19 0.20 0.14 0.27 Current Non-Payers 0.00320 0.00020 0.00010 0.00000 0.00290 0.00000 0.00030 0.00000 0.00000 0.00020 0.00000 Former Payers 0.01020 0.05210 0.06370 0.06920 0.06860 0.16400 0.07520 0.09040 0.07580 0.05700 0.07510 Never Paid 0.00240 0.00610 0.01540 0.02910 0.02920 0.04740 0.03270 0.04830 0.02200 0.02320 0.01600 Total Non-Payers 0.0031 0.0089 0.0172 0.0243 0.0259 0.0685 0.0434 0.0615 0.0422 0.0363 0.0447 Average Long-Term Borrowings to Assets Current Payers 24 24 21 20 23 20 21 18 17 18 Initiators 33 25 23 15 14 18 15 32 5 5 13 Regular Payers 31 25 24 24 23 28 22 20 18 16 17 Total Payers 32 25 24 22 21 26 21 21 17 16 17 Current Non-Payers 22 21 18 13 6 5 2 1 0 1 1 Former Payers 33 28 26 24 21 22 21 23 25 25 25 Never Paid 57 49 45 41 34 113 27 25 23 26 44 Total Non-Payers 48 41 35 31 25 78 24 23 23 25 33 Average Assets Current Payers 47.41 70.66 82.36 167.45 166.17 390.66 503.89 599.65 590.23 654.90 Initiators 54.71 47.88 1314.34 45.60 47.06 33.98 75.89 65.06 435.68 107.72 581.48 Regular Payers 112.00 156.35 233.57 391.44 377.02 397.83 630.04 942.39 1233.47 1550.57 1711.13 Total Payers 104.40 142.09 326.28 326.38 318.96 341.94 571.88 835.89 1083.68 1265.88 1413.43 Current Non-Payers 43.50 77.72 34.30 38.05 13.35 24.37 13.02 7.55 24.26 65.09 18.73 Former Payers 90.14 106.20 98.56 106.24 81.26 55.59 70.47 122.97 172.29 166.22 239.20 Never Paid 51.69 52.68 60.66 53.63 47.08 42.87 48.50 64.99 82.18 85.62 80.57 Total Non-Payers 56.92 63.53 65.74 63.50 51.47 46.85 57.60 94.89 134.95 133.16 181.20

5.4 Logit Analysis of Size, Profitability and Investment opportunities

An analysis of the effects of size, profitability and investment opportunities on the likelihood that a firm pays dividend shows that larger firms are more likely to pay dividends compared to small firms as measured by market capitalization percentile, and profitable firms are more likely to pay higher dividends compared to less profitable firms or loss making firms (Table 5.2). These results are consistent with Fama and French (2001). However, the likelihood that firms with more investment opportunities pay less or more dividend is not statistically significant, though the results show that firms with more asset growth are more likely to pay dividends than firms with less asset growth. This result contradicts the findings of Fama and French where their study finds that firms with more investment opportunities pay lower dividends.

23

Table 5.2
Logit Analysis: Characteristics of Dividend Payers and Non-Payers
Market Market Value Intercept Cap to Book Value Percentile of Assets -1.368 0.024 -0.001 1991 0.266 0.004 0.002 -0.828 0.021 -0.002 1992 0.244 0.004 0.001 -1.872 0.028 -0.003 1993 0.218 0.003 0.001 -1.984 0.032 -0.002 1994 0.189 0.003 0.001 -1.836 0.035 -0.002 1995 0.155 0.002 0.001 -3.006 0.046 0.000 1996 0.143 0.002 0.001 -3.647 0.052 -0.004 1997 0.134 0.002 0.001 -3.912 0.053 -0.001 1998 0.143 0.002 0.001 -5.012 0.057 -0.001 1999 0.197 0.003 0.000 -3.912 0.047 0.000 2000 0.159 0.002 0.000 -4.356 0.057 -0.002 2001 0.188 0.003 0.000 -2.885* 0.041* -0.002 Avg 0.185 0.003 0.001 T -15.586 15.067 -2.000 *Significant at .05 level Pseudo R2 Growth Payoff on Rate of Firms Cox & Snell Nagelkerke Firms Assets Assets 0.004 0.143 18% 26% 728 0.003 0.020 0.003 0.118 13% 20% 824 0.001 0.017 0.020 0.135 26% 38% 1033 0.003 0.015 0.007 0.133 27% 38% 1259 0.002 0.013 0.004 0.084 25% 35% 1712 0.001 0.010 0.001 0.083 33% 45% 2478 0.000 0.008 0.000 0.100 38% 51% 3186 0.000 0.008 0.003 0.061 35% 47% 3116 0.001 0.006 0.011 0.157 43% 58% 2789 0.002 0.010 0.001 0.096 35% 48% 2697 0.001 0.007 0.002 0.073 37% 51% 2332 0.001 0.007 0.005* 0.108* 0.001 0.011 3.733 9.777

6. Signaling Hypothesis: Case of Dividend Initiations and Omissions

On an average slightly more than 50 companies initiated dividends in each year during the sample period, while omissions have increased from 29 in 1994 to a maximum of 299 in 1998 before dropping to 138 in 2001 (Table 6.1, and Figure 6.1). Industry-wise analysis of dividend initiations show that chemicals and plastics industry firms have registered the highest number of initiations followed by firms in machinery industry (Table 6.2 and Figure 6.2 and 6.3). Electricity firms registered the lowest initiations. Similar industry trends were observed in the case of dividend omissions by firms.
Dividend Initiations and Omissions During 19932001
Number of Firms and Dividend Year Initiations Omissions Paid Not Paid 1993 34 47 1823 1274 1994 70 29 2333 1687 1995 54 48 2775 2340 1996 62 141 2723 2877 1997 55 227 2386 3469 1998 50 299 2101 3879 1999 60 218 2007 4241 2000 81 133 1988 4237 2001 53 138 1531 3235 Total 3097 4020 5115 5600 5855 5980 6248 6225 4766

Table 6.1

24

Table 6.2
Industry-wise Dividend Initiations and Omissions During 1993-2001
Dividend Initiations Industry Chemicals and Plastics Diversified Electricity Financial Services Food and Beverages Machinery Metals and Metal Product Mining Misc. Manufacturing Non-Metallic Mineral Pro Other Services Texti les Transport Equipment 1993 1994 1995 1996 1997 1998 1999 2000 2001 7 15 9 12 7 7 9 13 11 2 3 1 2 2 2 1 2 2 2 2 2 4 3 7 2 6 13 23 6 3 8 7 2 4 2 5 7 3 6 9 7 7 9 10 12 18 8 3 1 3 . 5 2 2 4 6 3 5 7 5 6 3 2 6 2 6 7 4 2 3 8 4 5 3 2 2 2 9 4 7 1 1 3 2 5 8 3 1 1 1 1 7 4 2 5 1 3 7 2 1 3 4 5 1 10 Dividend Omissions Tota l 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total 90 5 9 5 22 31 48 34 20 23 197 15 2 3 6 3 4 9 5 3 9 44 6 2 1 1 1 5 66 1 1 10 44 65 33 14 16 184 41 3 1 2 21 11 14 15 16 12 95 86 5 7 5 14 26 36 31 22 26 172 33 13 26 17 62 33 31 11 3 4 3 7 3 2 3 2 2 4 1 3 7 13 1 16 1 3 1 13 36 1 27 2 13 11 22 29 5 31 3 12 11 28 28 13 20 9 4 26 30 10 10 1 4 7 23 6 7 6 2 3 2 19 10 9 127 9 48 45 144 161 49

Figure 6.1
Dividend Initiations and Omissions During 1993 - 2001
Initiations 350 Omissions

Number of Firms

300 250 200 150 100 50 0

1993 1995 1997 1999 2001

Year

25

Figure 6.2
Industry-wise Dividend Initiations During 1993-2001

Transport Equipment Textiles Other Services Non-Metallic Mineral Pro Misc. Manufacturing Mining Metals and Metal Product Machineray Food and Beverages Financial Services Electricity Diversified Chemicals and Plastics 0 20 40 60 80 100

Number of Firms

Figure 6.3
Industry-Wise Dividend Omissions 1993-2001
Transport Equipment Textiles Other Services Non-Metallic Mineral Pro Misc. Manufacturing Mining Metals and Metal Product Machineray Food and Beverages Financial Services Electricity Diversified Chemicals and Plastics 0 50 100 150 200 250

Number of Firms

An analysis of past and future levels of standardized earnings of dividend omitting and initiating firms show that the former have lower and negative average earnings compared to that of latter (Table 6.3). Firms, which omitted dividends in the current year, have average negative earnings in the current and all 3 future years. This implies that the decision to omit dividends is based on the current earnings and expected future earnings. Firms, which initiate dividends on the other hand, have positive and increasing past earnings and the current earnings have reached the highest level. Initiators have positive future earnings for the next period but on average have negative earnings 2 years after the dividend initiation. Average earnings of dividend omitting firms have shown significant difference over the past 3 and next 3 years, where as initiating firms have exhibited a contrasting trend (Table 6.4). This implies that Firms omitting dividends have experienced a change in their earnings pattern, where as the initiating firms have similar past, current and future earnings.

26

Past and Future Levels of Standardized Earnings for Initiating and Omitting Firms
Omissions Initiations Year Mean SD Firms Mean SD Firms -3 6.08 16.60 730 -5.24 42.72 246 -2 4.64 14.89 851 1.79 30.00 263 -1 1.80 7.85 918 2.28 38.70 270 0 -7.63 26.35 921 7.22 40.80 285 1 -7.34 20.76 719 4.47 9.14 253 2 -7.96 28.85 606 -7.75 163.55 206 3 -7.82 33.67 460 1.20 15.55 181 Total -2.05 22.37 5205 .94 64.78 1704

Table 6.3

Dunnetts C (Post Hoc) test a pair-wise comparison test based on Studentized range when the variances are unequal also shows similar results (Table 6.5). Current earnings are statistically significant from t-3 year, t-2, and t-1 year earnings, where as they are not s ignificantly different from next 1, 2 and 3 year earnings. This reinforces the finding that dividend omission decision is based on the perception that earnings trend has reversed. Dunnetts C test shows that current earnings for initiating firms are not significantly different from that of past 1 and 2 year earnings and next 1, 2 and 3 year earnings, where as they are significantly different from that of t-3 years. To sum up, firms omit dividend at the sign of first trouble where as firms take a while before they initiate dividend payments.
Table 6.4
F 66.3

One-Way ANOVA of Past and Future Earnings and Dividend Initiations and Omissions
Sum of Squares 185090 2418420 2603510 Omissions df Mean Square 6 30848.4 5198 465.3 5204 Sig. .000 Sum of Squares 40040 7106691 7146732 Initiations df Mean F Square 6 6673.4 1.594 1697 4187.8 1703 Sig. .145

Between Groups Within Groups Total

27

Dunnetts C Post Hoc Test for Analysis of Influence of Past and Future Earnings on Omissions
Omissions Initiations Associated Mean SE Mean SE Year Year Difference Difference -3 -2 1.4 .8 -7.0 3.3 -1 4.3* .7 -7.5 3.6 0 13.7* 1.1 -12.5* 3.6 1 13.4* 1.0 -9.7* 2.8 2 14.0* 1.3 2.5 11.7 3 13.9* 1.7 -6.4 3.0 -2 -3 -1.4 .8 7.0 3.3 -1 2.8* .6 -.5 3.0 0 12.3* 1.0 -5.4 3.0 1 12.0* .9 -2.7 1.9 2 12.6* 1.3 9.5 11.5 3 12.5* 1.7 .6 2.2 -1 -3 -4.3* .7 7.5 3.6 -2 -2.8* .6 .5 3.0 0 9.4* .9 -4.9 3.4 1 9.1* .8 -2.2 2.4 2 9.8* 1.2 10.0 11.6 3 9.6* 1.6 1.1 2.6 0 -3 -13.7* 1.1 12.5* 3.6 -2 -12.3* 1.0 5.4 3.0 -1 -9.4* .9 4.9 3.4 1 -.3 1.2 2.8 2.5 2 .3 1.5 15.0 11.6 3 .2 1.8 6.0 2.7 1 -3 -13.4* 1.0 9.7* 2.8 -2 -12.0* .9 2.7 1.9 -1 -9.1* .8 2.2 2.4 0 .3 1.2 -2.8 2.5 2 .6 1.4 12.2 11.4 3 .5 1.8 3.3 1.3 2 -3 -14.0* 1.3 -2.5 11.7 -2 -12.6* 1.3 -9.5 11.5 -1 -9.8* 1.2 -10.0 11.6 0 -.3 1.5 -15.0 11.6 1 -.6 1.4 -12.2 11.4 3 -.1 2.0 -8.9 11.5 3 -3 -13.9* 1.7 6.4 3.0 -2 -12.5* 1.7 -.6 2.2 -1 -9.6* 1.6 -1.1 2.6 0 -.2 1.8 -6.0 2.7 1 -.5 1.8 -3.3 1.3 2 .1 2.0 8.9 11.5 * The mean difference is significant at the .05 level.

Table 6.5

7. Signaling Hypothesis: Case of Dividend Reductions

To analyze the relationship between dividends and losses a sample is drawn with firms having consistent profitability and dividend track records during 1990 1995 and who have earnings and dividend information for the period 1996 2001. This process yielded a sample of 599 firms out of which 363 have no losses during the period of 1996 2001, where as the remaining 236 firms had at least one negative earnings (Table 7.1). Among the loss sample, one-fourth of them have a single year loss where as 45 firms have 2 loss years, while 5 firms have losses during all 6 years. However, there is an increasing trend in the proportion of firms

28

reported loss in the sample over the period 1996 2001. Percentage of loss firms has increased from 2.7% in 1996 to 27.2% in 2001.
Table 7.1

Distribution of Firms In Terms of Earnings Performance During 1996 - 2001

No. of Loss Years No. of Firms with Loss No Loss Total None 363 363 1 99 99 2 45 45 3 40 40 4 30 30 5 17 17 All 5 5 Total 236 363 599 Earnings Performance During 1996 - 2001 1996 1997 1998 1999 2000 2001 No Loss 583 556 520 485 470 436 % 97.3 92.8 86.8 81.0 78.5 72.8 Loss 16 43 79 114 129 163 % 2.7 7.2 13.2 19.0 21.5 27.2

Out of 599 firms in the sample, only 15 firms have recorded consistent dividend payment record where as the remaining firms have reduced dividend at least for one year (Table 7.2). Out of the total 584 dividend reducing firms, 11% have reduced dividend per share once, 19.9% of firms twice, 22% of firms thrice where as 11.4% of firms in the sample reduced dividend per share in all the 6 years of sample period. The incidence of annual dividend omission is not so severe compared to reductions during the sample period, as 61.3% of firms in the sample have not skipped even once their dividends, where as 8.7% skipped once, 5.2% skipped twice, where as around 25% of firms have skipped three or more years.
Table 7.2

Distribution of Firms In Terms of Dividend Reductions and Omissions During 1996 - 2001
No. of Years None 1 2 3 4 5 6 Total Reduction % No Reduction % Skipped % Not Skipped % 1996 249 41.6 350 58.4 48 8.0 551 92.0 Reduced Firms 15 66 119 132 115 84 68 599 1997 258 43.1 341 56.9 77 12.9 522 87.1 % 2.5 11.0 19.9 22.0 19.2 14.0 11.4 100 Year 1998 1999 365 339 60.9 56.6 234 260 39.1 43.4 114 151 19.0 25.2 485 448 81.0 74.8 Skipped Firms 367 52 31 41 43 32 33 599 2000 339 56.6 260 43.4 170 28.4 429 71.6 % 61.3 8.7 5.2 6.8 7.2 5.3 5.5 100 2001 438 73.1 161 26.9 207 34.6 392 65.4

29

A preliminary analysis of the impact of losses on dividend reduction shows that firms with no losses for the entire sample period have also reported dividend reductions albeit to a low extent compared to that of firms with more recurring losses (Table 7.3). The 15 firms, which have shown no reduction in dividend per share in all years of the sample period, have also recorded no negative earnings performance during the sample period. Out of a total of 363 firms which recorded positive earnings during the entire sample period 17.6% firms have reduced dividends for 1 year, 28.9%, 25.6%, 15.2%, 6.1% and 2.5% firms have reduced for 2, 3, 4, 5 and in all years respectively. Compared to this firms with more recurring losses have shown more frequent reductions as all the 99 firms with 1 loss year have reduced dividend for at least one year, where as 45 firms with 2 loss years, 40 firms with 3 loss years have reduced dividend at least for two years. Five firms who have recorded losses in all the 6 years have also reduced divided through out the sample period. Further, null hypothesis of no association between losses and dividend reductions is rejected at the 5% significance level indicating the significance of losses for dividend reductions. The Kendalls tau-b value of 0.538 indicates that the association is positive and degree of association is significant.
Table 7.3

Earnings Performance and Dividend Reductions

No. of Loss Years None No. of Years Reduced None 1 2 3 4 5 All Total Firms 15 64 105 93 55 22 9 363 Row % 4.1 17.6 28.9 25.6 15.2 6.1 2.5 100 Column 100 97.0 88.2 70.5 47.8 26.2 13.2 60.6 % Firms 2 12 26 34 18 7 99 Row % 2.0 12.1 26.3 34.3 18.2 7.1 100 Column 3.0 10.1 19.7 29.6 21.4 10.3 16.5 % Firms 1 12 11 13 8 45 Row % 2.2 26.7 24.4 28.9 17.8 100 Column .8 9.1 9.6 15.5 11.8 7.5 % Firms 1 1 10 13 15 40 Row % 2.5 2.5 25.0 32.5 37.5 100 Column .8 .8 8.7 15.5 22.1 6.7 % Firms 5 12 13 30 Row % 16.7 40.0 43.3 100 Column 4.3 14.3 19.1 5.0 % Firms 6 11 17 Row % 35.3 64.7 100 Column 7.1 16.2 2.8 % Firms 5 5 Row % 100 100 Column 7.4 .8 % Firms 15 66 119 132 115 84 68 599 Row % 2.5 11.0 19.9 22.0 19.2 14.0 11.4 100 Column 100 100 100 100 100 100 100 100 % Kendall's tau-b Asymp. Std. Error Approx. T Approx. Sig. Firms .022 19.579 .000 599

All

Total

Value .538

An analysis of the association between earnings performance and annual dividend omission shows that the incidence of annual omission is severe in the case of firms with more recurring losses (Table 7.4). Out of

30

a total of 363 firms with positive earnings through out the sample period, 90% of the firms have not skipped dividends even for a single year. Compared to this only 33.3% of firms with 1 year loss, 13.3 % with 2 year losses, 2.5% with 3 year losses have not skipped dividend during the entire sample period, where as none of the firms with 4 or more years of losses have paid regular dividends during the sample period. Further, the null hypothesis of no association between earnings performance as measured by no of loss years and dividend omissions as measured by number of years dividend skipped is rejected at the 5% significance level. The Kendalls tau-b value of 0.703 indicates that the association between losses and dividend omissions is positive and degree of influence is considerable.
Table 7.4

Earnings Performance and Annual Dividend Omissions

No. of No. of Years Skipped Loss Years None 1 2 3 4 5 All Total None Firms 327 8 4 5 8 4 7 363 Row % 90.1 2.2 1.1 1.4 2.2 1.1 1.9 100 Column 89.1 15.4 12.9 12.2 18.6 12.5 21.2 60.6 % 1 Firms 33 39 11 7 3 2 4 99 Row % 33.3 39.4 11.1 7.1 3.0 2.0 4.0 100 Column 9.0 75.0 35.5 17.1 7.0 6.3 12.1 16.5 % 2 Firms 6 5 10 9 9 5 1 45 Row % 13.3 11.1 22.2 20.0 20.0 11.1 2.2 100 Column 1.6 9.6 32.3 22.0 20.9 15.6 3.0 7.5 % 3 Firms 1 3 12 13 2 9 40 Row % 2.5 7.5 30.0 32.5 5.0 22.5 100 Column .3 9.7 29.3 30.2 6.3 27.3 6.7 % 4 Firms 1 7 10 8 4 30 Row % 3.3 23.3 33.3 26.7 13.3 100 Column 3.2 17.1 23.3 25.0 12.1 5.0 % 5 Firms 2 1 11 3 17 Row % 11.8 5.9 64.7 17.6 100 Column 6.5 2.4 34.4 9.1 2.8 % All Firms 5 5 Row % 100 100 Column 15.2 .8 % Total Firms 367 52 31 41 43 32 33 599 Row % 61.3 8.7 5.2 6.8 7.2 5.3 5.5 100 Column 100 100 100 100 100 100 100 100 % Kendall's tau-b Value Asymp. Std. Error Approx. T Approx. Sig. Firms .703 .025 20.579 .000 599

The importance of annual losses on dividend reductions and annual dividend omissions has been analyzed with the help of logit analysis (Table 7.6). For this analysis a pooled sample of 576 firms is used with 236 loss firms and 340 no loss firms but with an earnings decline in the event year. The dependent variable equals zero if a firm has maintained or increased its dividend per share and is equal to one if the firm announced a reduction in dividend per share. The loss dummy assumes a value of one if the firm reports a loss for the year under study and zero otherwise. The level of net income and changes in net income are standardized with the previous years net worth for each firm. For firms in loss sample, the initial loss year constitutes the event year where as for non-loss firms, the initial year of earnings decline constitutes the event year. This process has resulted in a loss of 23 firms from the 599 firms as these 23 firms have not reported any earnings decline during the 1996 2001 period. The deletion of 23 firms and

31

focus on declining income years for non-loss sample is justified because firms seem unlikely to cut dividends when earnings are positive and increasing. This is consistent with DeAngelo, DeAngelo and Skinner (1992).
Table 7.5

Distribution of Firms by Change in Net Income

% Change < -100 -100 - -50 -50 - 0 No Change 0 - 50 50 - 100 > 100 Total -ve Change +ve Change Total 1996 1997 1998 1999 2000 2001 Firms % Firms % Firm %Firms %Firms % Firms s 2 .3 5 .8 6 1 5 .8 10 4 .7 2 .3 5 .8 3 .5 14 2.3 12 206 34.4 307 51.3 274 45.7 319 53.3 242 40.4 298 1 .2 2 .3 4 .7 2 .3 2 382 63.8 282 47.1 311 51.9 261 43.6 324 54.1 260 4 .7 1 .2 3 .5 2 .3 3 .5 9 2 .3 3 .5 1 .2 4 .7 9 1.5 8 599 100 599 100 599 100 599 100 599 100 599 211 35.2 313 52.3 284 47.4 332 55.4 263 43.9 322 388 64.8 286 47.7 315 52.6 267 44.6 336 56.1 277 599 100 599 100 599 100 599 100 599 100 599 % 1.7 2 49.7 .3 43.4 1.5 1.3 100 53.8 46.2 100

The loss dummy is positive and significant in all the models where individually and collectively it explains a reduction in dividend. The positive sign of the coefficient indicates that in firms that report losses there is a higher probability of dividend reduction. This is consistent with DeAngelo, DeAngelo and Skinner who find negative influence of losses on dividends11. Level of net income has significantly negative impact on dividend reduction implying that firms with higher levels of net income have lower dividend reductions and vice versa. Further, change in net income has also negatively impacted dividend reductions implying that positive (negative) changes in net income have resulted in lower (higher) dividend reductions. When compared with other simple models of dividend reductions, loss dummy alone has greater explanatory power compared to level of net income and change in net income individually. Addition of level of net income or change in net income to the model improves the explanatory power marginally. However, a model with both level of and change in net income has comparatively less explanatory power compared to loss dummy alone.
Logit Regression: Dividend Reductions and Earnings Performance
Parameters Constant Loss Dummy Level of Net Income Change in Net Income 1 0.094 3.256*
(75.08) (0.75)

Table 7.6
3 0.219 2.37*
(2.61) (4.09)

2 0.493*
(5.11) (25.22) (4.46)

2.51*

(0.731) (22.96)

4 -0.128 2.856*
(50.95) (.814)

5 1.053*
(25.37)

(111.83)

6 1.471*

7 -0.089
(0.38)

-0.026*

-0.021 -0.028*

-0.07* -0.038*
(6.44) (51.39)

-0.08*
(84.65)

-0.032*
(5.65)

-0.098*
(48.23)

Pseudo R 2 Cox & Snell 24% 25% 25% Nagelkerke 34% 35% 36% Note: * significant at .05 level Figures in the parentheses are wald statistic values

25% 36%

22% 31%

21% 29%

13% 19%

The difference in the sign is due to the difference in the assignment of codes to the dependent variable. In the present study the dependent variable assumes a value of 1 if there is a reduction in dividend per share where as DeAngelo, DeAngelo and Skinner assume a value of 0 for a dividend reduction.
11

32

Logit regression analysis of the determinants of annual dividend omissions shows that loss dummy is significant and has positive impact on dividend omissions (Table 7.7). However, the level of net income has more impact on dividend omissions compared to loss dummy and change in net income as it had more explanatory power compared to others on an individual basis. Further, level of net income is inversely related to dividend omissions implying that firms with higher net income have lower dividend omissions and vice versa. From the previous logit analysis it is clear that current losses are an important determinant of dividend reductions for firms with established track record. However, there are 8 firms out of a total of 236 firms with losses that did not reduce dividends (Table 7.8). The incidence of dividend reduction is much more severe in the case of Indian firms compared to that of NYSE as analyzed by DeAngelo, DeAngelo and Skinner.
Table 7.7
3 -0.859*
(3.86)

Logit Regression: Annual Dividend Omissions and Earnings Performance

Parameters Constant Loss Dummy Level of Net Income Change in Net Income 1 -3.225*
(130.04)

2 -1.496*
(1744)

4 -3.61*
(134.02)

5 -0.466*
(4.30)

6 -0.982*
(42.76)

7 -2.144*
(156.13)

4.171*
(172.24)

0.862
(2.669)

0.564
(1.12)

3.69*
(125.59)

-0.151*
(37.76)

-0.191*
(37.46)

-0.217*
(103.69)

-0.184*
(134.41)

0.043*
(6.59)

-0.044*
(15.04)

-0.049*
(8.59)

-0.10*
(95.47)

Pseudo R 2 Cox & Snell 44% 50% 50% Nagelkerke 61% 70% 70% Note: * significant at .05 level Figures in the parentheses are wald statistic values

46% 64%

50% 70%

50% 70%

24% 33%

Association between Loss and Dividend Reduction

Dividend No Reduction Reduction No Loss 162 178 Loss 8 228 Total 170 406 Total 340 236 576

Table 7.8

An analysis of mean differences in earnings past and future for dividend reducing and non-reducing firms shows statistically different earnings over different periods for dividend reducing firms, where as for non-reducing firms no difference in earnings levels are observed (Table 7.9). Further, Dunnetts C test of mean differences in earnings performance shows that for dividend reducing firms earnings in t 1, t, t+1, and t+2 are statistically different from each other and that the earnings are declining from t-1 through t+2 (Table 7.10). However, earnings of firms that have not reduced dividends appear to be at the same level from the period t 2 through t+2. On the whole from the analysis of mean differences in earnings, it can be inferred that consistent earning levels have resulted in stable or positive dividend payments, where as consistent and significant reduction in past or expected earnings have negatively impacted the dividends.

33

Test of Mean Difference in Earnings between Lagged and Future Years

Sum of Squares Reductions Between Groups Within Groups Total Non-Reductions Between Groups Within Groups Total 73082.0 4390586.9 4463668.9 4720.7 3075645.2 3080365.9 df 4 2757 2761 4 2026 2030 Mean Square F Sig. .000

Table 7.9

18270.5 11.5 1592.5 1180.2 .777 1518.1

.540

Logit analysis of the impact of lagged, current and future earnings performance on annual dividend changes shows that current earnings explain a relatively higher 9% change in Dividends compared to that of past (5%) and future earnings (6.8% for t+1 and 6.2% for t+2) (Table 7.11). Earnings in t-1, t+1, and t+2 years have statistically significant negative influence on the likelihood of dividend reductions when considered individually. Addition of t-1 earnings to current earnings, improves the explanatory ability of the model to 16.2%. Addition of earnings in t+1 and t+2 only improve the explanatory ability of the models marginally. Removal of t-1 earnings and inclusion of t+1 earnings improve the models explanatory ability only to 9%. From the above analysis it may be concluded that dividend changes are impacted more by contemporaneous and lagged earnings performance rather than future earnings performance. These results are consistent with the findings of Benartzi, Michaely, and Thaler (1997).
Table 7.10

Dunnetts C Test of Mean Differences in Earnings Performance for Dividend Reductions and Increases
Dividend Reducing Firms Non-reducing Firms Earnings i Year i Year j SE Earnings i - j SE j t-2 t-1 2.8 3.0 3.7 3.1 t 2.6 .6 3.2 11.1* * t+1 2.7 2.3 3.2 11.7 t+2 2.7 3.6 3.2 12.4* t-1 t-2 -2.8 3.0 -3.7 3.1 t 2.3 -3.1 2.4 8.3* * t+1 2.4 -1.5 2.3 8.9 t+2 2.4 -.1 2.4 9.5* t t-2 2.6 -.6 3.2 -11.1* t-1 2.3 3.1 2.4 -8.3* t+1 .6 1.9 1.7 2.5 t+2 1.3 1.9 3.0 2.5 t+1 t-2 2.7 -2.3 3.2 -11.7* t-1 -8.9* 2.4 1.5 2.3 t -.6 1.9 -1.7 2.5 t+2 .7 2.0 1.4 2.4 t+2 t-2 2.7 -3.6 3.2 -12.4* t-1 2.4 .1 2.4 -9.5* t -1.3 1.9 -3.0 2.5 t+1 -.7 2.0 -1.4 2.4 *Significant at 0.5 significance level Earnings In

34

Logit Regression: Annual Dividend Reductions and Earnings Performance

Parameters Constant Earnings (t-2) Earnings (t-1) Earnings (t) Earnings (t+1) Earnings (t+2) -.035*
(65.1)

Table 7.11
5 .52*
(48.2)

1 .658*
(51.8)

2 .376*
(28.2)

3 .538*
(50.5)

4 .37*
(27.8)

Model 6 .58*
(47.0)

7 .648*
(62.5)

8 .588*
(46.5)

9 .58
(43.3)

10 .599*
(48.0)

-.002
(3.0)

-.003 .055*
(45.6) (81.5) (.74) .061* (42.5) -.095* (82.9)

-.005
(3.11)

-.004*
(3.94)

.056*
(40.9) (46.7)

.067*
(46.9) -.08* (44.2)

-.091* -.025*
(39.5)

-.028*
(21.4)

-.081* -.015*
(5.4)

-.008
(3.6)

-.013
(3.7)

-.024*
(37.9)

-.011
(2.42)

Pseudo R 2 Cox & Snell 9% 5% 6.8% 4% 6.2% Nagelkerke 12% 6% 9.1% 5% 8.4% Note: * significant at .05 level Figures in the parentheses are wald statistic values

16.2% 21.8%

9.3% 12.5%

16.7% 22.5%

17.3% 23.3%

18.3% 24.5%

8. Summary and Conclusion

The present study examines the dividend behavior of Indian corporate firms over the period 1990 2001 and attempts to explain the observed behavior with the help of trade-off theory, and signaling hypothesis. Trends indicate that the number of firms paying dividend during the study period has shown an up trend till 1995 and has fallen subsequently. Average DPS on the other hand has shown a steady growth except for year 2001. Average percentage PR showed a more stable pattern up to 1997 and then has shown a declining trend. Analysis also shows that only a few firms have consistently paid same levels of dividend. Of the payers, regular payers have consistently paid higher payout as well as higher average dividend compared to that of current payers. Initiators have always paid higher levels of dividend yield compared to that of other payers. Further, smaller indices appear to have higher dividends compared to that of larger indices. Industry trends indicate that firms in the electricity, mining and diversified industries have paid higher dividends where as textile companies have paid less dividends. Analysis of influence of tax regime changes shows that the tradeoff theory does not hold true in the Indian context, as Indian corporate firms on average do not appear to have increased dividend payments despite a tilt in tax regime in favor of more dividends. Analysis of characteristics of payers and non-payers shows that dividend-paying companies are more profitable and large in size. However, growth doesnt seem to deter Indian firms from paying higher dividends. Further, firms appear to prefer the pecking order of funds in building their larger asset base. An analysis of signaling hypothesis shows that average earnings of dividend omitting firms have shown significant difference over the past 3 and next 3 years, where as initiating firms have exhibited a contrasting trend. This reinforces the finding of Benartzi, Michaley and Thaler that dividend omission decision is based on the perception that earnings trend has reversed. This analysis implies that dividend omissions have

35

information content in that these firms expect lower earnings for the future. However, this is not the case with regard to dividend initiations. An analysis of other non-extreme dividend events such as dividend reductions and non-reductions shows that current losses are an important determinant of dividend reductions for firms with established track record. The incidence of dividend reduction is much more severe in the case of Indian firms compared to that of NYSE as analyzed by DeAngelo, DeAngelo and Skinner. Further analysis also shows that dividend changes are impacted more by contemporaneous and lagged earnings performance rather than by future earnings performance. The present study has considered only cash dividends and not share repurchases. Share repurchases or buyback has been permitted in the Indian context only recently and this may well have influenced the dividend behavior of Indian companies, as some firms would have substituted share repurchases for cash dividends. Similarly, in the present study only final cash dividends are considered and the stock dividends by firms are not considered which may limit generalizations of the findings. Further, the present study has not considered the stock market reactions to dividend events and has not examined at great depth the interrelations between dividend and other corporate finance decisions. Future studies may examine the market reaction to dividend announcements, other possible determinants of dividend behavior such as flotation costs, and the relationships between dividend decision and financing and investment decisions.

36

References
1. Baker, H.K., E.T. Veit and G.E. Powell (2001), Factors Influencing Dividend Policy Decisions of Nasdaq Firms, The Financial Review, V. 36, No. 3, pp. 19-38. 2. Baker, H.K., G.E. Powell, and E.T. Veit (2002), Revising the Dividend Puzzle Do all the Pieces Now Fit?, Review of Financial Economics, In Press. 3. Benartizi, S., R. Michaely, and R. Thaler (1997), Do Changes in Dividends Signal the Future or the Past?, Journal of Finance, Vol. 52, No. 3, July, pp. 1007-1034. 4. Bernstein, P.L. (1998), The Hidden Risks in Low Payouts, The Journal of Portfolio Management, Vol. 25, No.1, Fall, p. 1. 5. Bhat, R. and I.M. Pandey (1994), Dividend Decision: A Study of Managers Perceptions, Decision, Vol. 21, No.s 1 & 2, January-June 1994. 6. Bhattacharya, S. (1979), Imperfect Information, Dividend Policy, and the bird in the hand Fallacy, Bell Journal of Economics, Vol. 10,No.1, Spring, pp. 259-270. 7. Black, F. (1976), The Dividend Puzzle, Journal of Portfolio Management, Vol. 2, No. 2, Winter, pp. 58. 8. Brealey, R.A. (1994), Does Dividend Policy Matter? in Stern, J.M. and D.H. Chew (eds.), Revolution in Corporate Finance, 2nd edition, Blackwell Publishers Inc., Cambridge, Massachusettes. 9. DeAngelo, H. L. DeAngelo and D.J. Skinner (1992), Dividends and Losses, Journal of Finance, Vol. 47, No. 5, December, pp. 18371863. 10. Fama, E.F. and K.R. French (2001), Disappearing Dividends: Changing Firm Characteristics or Lower Propensity to Pay?, Journal of Applied Corporate Finance, Vol. 14, No. 1, Spring, pp. 67-79. 11. Glen, J.D., Y. Karmokolias, R.R. Miller, and S. Shah, Dividend Policy and Behavior in Emerging Markets, Discussion Paper No. 26, International Finance Corporation, 1995. 12. Healey, P.M. and K.G. Palepu (1988), Earnings Information Conveyed by Dividend Initiations and Omissions, Journal of Finanscial Economics, Vol. 21, No. 2, September, pp. 149-175. 13. Jensen, M.C. and W.H. Meckling (1976), Theory of the firm: Managerial behavior, agency costs and ownership structure, Journal of Financial Economics, Vol. 3, No. 4 October, pp. 305-360. 14. John, K. and J. Williams (1985), Dividends, Dilution, and Taxes: A Signaling Equilibrium, Journal of Finance, Vol. 40, No. 4, pp. 1053-1070. 15. Kalay, A. (1982), Stockholder-Bondholder Conflict and Dividend Constraints, Journal of Financial Economics, Vol. 10, No. 2, pp. 211-233. 16. Kevin, S. (1992), Dividend Policy: an Analysis of Some Determinants, Finance India, Vol. VI, No. 2, June, pp. 253-259. 17. Lee, H.W. and P.A. Ryan (2002), Dividends and Earnings Revisited: Cause or Effect?, American Business Review, January, Vol. 20, No.1, January, pp. 117 122. 18. Lintner, J. (1956), Distribution of Incomes Corporations Among Dividends, Retained Earnings and Taxes, American Economic Review, Vol. 46, No.2, May, pp. 97-113. 19. Mahapatra, R.P. and P.K. Sahu (1993), A Note on Determinants of Corporate Dividend Behaviour in India An Econometric Analysis, Decision, Vol. 20, No. 1, January-March, pp. 122. 20. Mishra, C. and V. Narender (1996), Dividend Policies of SoEs in India An Analysis, Finance India, Vol. X, No. 3, September, pp. 633-645. 21. Miller, M.H. and F. Modigliani (1961), Dividend Policy, Growth and the Valuation of Shares, Journal of Business, Vol. 34, No. 4, October, pp. 411-433.

37

22. Miller, M. and K. Rock (1985), Dividend Policy under Asymmetric Information, Journal of Finance, Vol. 40, No. 4, pp. 1031-1051. 23. Mohanty, P. (1999), Dividend and Bonus Policies of the Indian Companies, Vikalpa, Vol.24, No. 4, October-December, pp. 35-42. 24. Myers, S.C., N. Majluf (1984), Corporate financing and investment decisions when firms have information investors do not have, Journal of Financial Economics, Vol. 13, No.2, pp. 187-221. 25. Narasimhan, M.S. and C. Asha (1997), Implications of Dividend Tax on Corporate Financial Policies, The ICFAI Journal of Applied Finance, Vol. 3, No.2, July, pp. 11-28. 26. Narasimhan, M.S. and S. Vijayalakshmi (2002), Impact of Agency Cost on Leverage and Dividend Policies, The ICFAI Journal of Applied Finance, Vol. 8, No. 2, March, pp. 16-25. 27. Ramcharran, H. (2001), An Empirical Model of Dividend Policy in Emerging Equity Markets, Emerging Markets Quarterly, Spring, pp. 39-49. 28. Rozeff, M. (1994), How Companies set their Dividend-Payout Ratios, in Stern, J.M. and D.H. Chew (eds.), Revolution in Corporate Finance, 2nd edition, Blackwell Publishers Inc., Cambridge, Massachusetts.

38

Appendix
Table A4.1
Distribution of Dividend Payers and Non-Payers: Number of Firms and Percentages
Payers / Non-Payers Non-Payer
%

1990 1991 674 912

40 42

1992 1993 1994 1995 Non-Payer Group 972 1274 1687 2340
39 41 42 46

Year 1996 2877

51

1997 3469
59

1998 3879
65

1999 4241
68

2000 2001 4237 3235

68 68

Current Non-Payers
%

324
35.5

201
20.7

369
29

563
33.4 53.6

763
32.6 53.4

417
14.5

354
10.2

276
7.1

379
8.9

232
5.5

21
0.6

Never Paid
%

535
58.7

686
70.6

737
57.8

904 1250 220

13

1776
61.7

1979
57

2107
54.3

2123
50.1

2205 1640
52 50.7

Former Payers
%

53
5.8

85
8.7

168
13.2

327
14

684
23.8

1136
32.7

1496
38.6

1739
41

1800 1574
42.5 48.7

Payer
%

Payers-Group 1033 1272 1533 1823 2333 2775

61 58 61 59 58 54

2723
49

2386
41

2101
35

2007
32

1988 1531
32 32

Current Payer
%

35
2.3

94
5.2

188
8.1

276
9.9

339
12.4

383
16.1

424
20.2

454
22.6

527 422
26.5 27.6

Initiators
%

331
26 74

322
21 76.7

419
23 71.9

569
24.4 67.6

594
21.4 68.6

398
14.6

220
9.2

181
8.6

225
11.2

177
8.9

74
4.8

Regular Payer
%

941 1176 1310 1576 1905

1986
72.9

1783
74.7

1496
71.2

1328
66.2

1284 1035
64.6 67.6

Figure A4.1
Dividend Behaviour of Indian Corporate Firms During 1990 - 2001
Non-Payer Payer 5000 4000 3000 2000 1000 0 1990

Companies

1992

1994

1996

1998

2000

Year

Table A4.2
Average Dividend Paid by Payers (in Rs. Crore)
Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Initiator 0.75 0.77 0.72 0.64 0.78 1.69 1.97 1.13 2.45 2.71 7.03 Current Payers 1.30 1.01 1.26 1.94 2.32 2.33 2.85 2.87 3.34 4.03 Regular Payers 2.01 2.11 2.32 2.85 3.66 4.48 5.98 8.08 9.87 12.44 17.17 Total Payers 1.69 1.81 1.89 2.19 2.87 3.80 5.02 6.43 7.46 9.16 13.05

39

Figure A4.2
Behaviour of Non-Payers During 1990 - 2001 (% of Firms)
Current Non-Payer % 80% Never Paid % Former Payer %

Percentage

60% 40% 20% 0% 1990 1992 1994 1996 1998 2000

Year

Figure A4.3
Behaviour of Non-payers During 1990 - 2001 (No. of Firms)
Total Non-Payers Current Non-Payers Former Payers

No. of Companies

5000 4000 3000 2000 1000 0 1990

Never Paid

1992

1994

1996

1998

2000

Year

Figure A4.4
Behaviour of Payers During 1990 - 2001
Payer Current Payer Initiator Regular Payer

No. of Companies

3000 2000 1000 0 1990 1992 1994 1996 1998 2000

Year

Figure A4.5
Behaviour of Payers During 1990 - 2001 (in %)
Current Payers % 100% 80% 60% 40% 20% 0% 1990 1992 1994 1996 1998 2000 Initiators % Regular Payers %

% of Firms

Year

40

Table A4.3
Comparison of Index and Non-index Firms - Average Dividend Paid (Rs. Crore)
CATEGORY 1990 1991 1992 1993 4.54 .90 5.05 .44 5.71 1.04 20.30 .82 27.33 .87 14.79 .71 10.21 .58 5.73 .43 1994 6.31 1.03 7.06 .52 8.37 1.18 27.75 .96 37.47 1.01 20.72 .82 13.93 .69 7.58 .53 1995 9.27 1.26 10.48 .63 12.88 1.45 38.35 1.19 52.62 1.27 31.35 .99 20.73 .83 10.92 .65 1996 10.81 1.53 13.04 .79 15.44 1.73 52.76 1.39 71.75 1.49 41.84 1.15 27.22 .97 13.81 .77 1997 12.71 1.69 15.13 .88 19.07 1.91 64.63 1.51 87.25 1.62 50.36 1.24 32.66 1.03 16.49 .82 1998 14.02 1.87 16.57 .99 21.64 2.09 75.42 1.64 102.17 1.77 58.08 1.33 38.01 1.09 18.34 .89 1999 13.03 2.05 17.57 1.11 24.95 2.21 84.24 1.73 118.44 1.85 68.10 1.35 46.06 1.01 21.04 .84 2000 15.21 2.52 21.47 1.33 29.61 2.71 110.11 2.06 146.40 2.23 84.79 1.60 55.10 1.21 25.45 .99 2001 15.68 3.71 26.97 1.65 59.70 3.61 124.27 2.92 171.07 3.14 111.67 1.89 73.17 1.20 32.70 .96 Firms 200 7582 500 7282 51 7731 50 7732 30 7752 100 7682 200 7582 500 7282 CNXMcap 3.07 3.50 4.06 Non-CNXMCap .81 .79 .89 CNX500 3.37 3.64 4.35 Non-CNX500 .39 .40 .45 Nifty Junior 4.07 4.65 5.74 Non-Nifty Junior .93 .91 1.02 Nifty 13.09 14.33 16.68 Non-Nifty .72 .73 .83 Sensex 17.89 19.77 22.31 Non-Sensex .76 .77 .88 BSE100 8.96 9.39 12.34 Non-BSE100 .64 .65 .71 BSE200 6.65 6.89 8.79 Non-BSE200 .53 .55 .59 BSE 500 4.03 4.20 4.97 Non-BSE500 .36 .38 .44 Note: Index compositions as of March 31, 2002

Table A4.4
Number of Index Firms Paying Dividend During 1990 - 2001
CATEGORY 1990 1991 1992 1993 41 1782 26 1797 40 1783 79 1744 155 1668 144 1679 379 1444 329 1494 1994 42 2291 26 2307 44 2289 83 2250 159 2174 146 2187 399 1934 349 1984 1995 46 2729 28 2747 46 2729 90 2685 169 2606 168 2607 430 2345 385 2390 1996 47 2676 28 2695 48 2675 92 2631 173 2550 170 2553 439 2284 393 2330 1997 49 2337 28 2358 47 2339 94 2292 174 2212 174 2212 436 1950 395 1991 1998 49 2052 28 2073 48 2053 95 2006 164 1937 171 1930 416 1685 386 1715 1999 49 1958 28 1979 49 1958 95 1912 168 1839 178 1829 407 1600 394 1613 2000 49 1939 29 1959 50 1938 96 1892 176 1812 182 1806 404 1584 410 1578 2001 Total 49 1482 29 1502 49 1482 98 1433 166 1365 185 1346 378 1153 396 1135 50 7732 30 7752 51 7731 100 7682 200 7582 200 7582 500 7282 500 7282 Nifty 35 39 41 Non-Nifty 998 1233 1492 Sensex 22 24 25 Non-Sensex 1011 1248 1508 Nifty Junior 30 35 39 Non-Nifty Junior 1003 1237 1494 BSE100 62 72 77 Non-BSE100 971 1200 1456 CNXMcap 117 130 147 Non-CNXMCap 916 1142 1386 BSE200 107 122 136 Non-BSE200 926 1150 1397 CNX500 282 313 346 Non-CNX500 751 959 1187 BSE 500 237 272 301 Non-BSE500 796 1000 1232 Note: Index compositions as of March 31, 2002

DPS Reductions by Firms Paid Dividend Continuously from 1990 2001

No. of Reductions 0 1 2 3 4 5 6 7 Total Firms 5 43 68 82 74 41 10 7 330 % 1.5 13.0 20.6 24.8 22.4 12.4 3.0 2.1 100.0

Table A4.5

41

Table A4.6
Recurring Dividend Per Share of Firms During 1990 - 2001
DPS Nil Rs. 0 - Rs. 0.25 Rs. 0.25 - Rs. 0.50 Rs. 0.50 - Rs. 0.75 Rs. 0.75 - Rs. 1 Rs. 1 - Rs. 2 Rs. 2 - Rs. 5 > Rs. 5 0 1421 3580 6221 7271 7549 7565 7710 7740 No. of Firms with Recurrences of 1 2 3 4 5 6 7 8 9 10 11 12 630 1035 909 874 798 702 521 307 213 159 128 85 812 802 597 503 439 372 261 160 108 76 54 18 442 297 232 153 158 124 57 37 33 14 12 2 252 126 72 33 13 13 2 140 48 23 13 6 1 2 123 57 20 9 5 3 42 17 7 3 2 1 16 13 3 3 1 3 1 1 1

Transition Probabilities for DPS Groups Based on Changes from 1990 to 2001
DPS (in Rs.) 1990 0 0 - 0.25 0.25 - 0.50 0.50 - 0.75 0.75 - 1 1-2 >5 0 0.70 0.46 0.32 0.15 0.60 0.50 0 - 0.25 0.18 0.23 0.21 0.23 0.25 - 0.50 0.07 0.20 0.21 0.40 2001 0.50 - 0.75 0.75 - 1 0.02 0.01 0.06 0.02 0.08 0.08 0.23 0.23 0.50 1-2 0.01 0.02 0.07 0.15 2-5 0.01 0.01 0.01 >5 0.01

Table A4.7

1.00

Table A4.8
Average Dividend Per Share for Index and Non-Index Firms During 1990-2001 (in Rs.)
CATEGORY non-nifty nifty Non-sensex Sensex Non-cnxmcap cnxmcap Non-BSE100 BSE100 Non-BSE200 BSE200 Non-cnx500 cnx500 Non-BSE500 BSE500 1990 .14 .28 .14 .28 .13 .22 .14 .24 .13 .22 .12 .20 .13 .21 1991 1992 .14 .14 .32 .30 .14 .14 .33 .29 .13 .14 .23 .23 .13 .14 .26 .25 .13 .14 .22 .23 .12 .13 .21 .21 .12 .13 .21 .21 1993 .15 .41 .15 .32 .15 .21 .15 .30 .15 .25 .14 .22 .14 .22 1994 1995 .16 .18 .34 .38 .16 .18 .35 .39 .15 .18 .25 .28 .16 .18 .29 .34 .15 .18 .26 .29 .15 .17 .24 .26 .15 .17 .24 .25 1996 .21 .45 .21 .51 .21 .31 .21 .39 .21 .33 .21 .29 .21 .27 1997 1998 .20 .23 .51 .56 .20 .23 .54 .58 .19 .23 .33 .37 .19 .23 .44 .50 .19 .23 .36 .41 .19 .23 .30 .32 .19 .23 .30 .32 1999 2000 2001 FIRMS .25 .25 .15 7732 .62 .82 .80 50 .25 .25 .15 7752 .65 .73 .85 30 .25 .25 .14 7582 .39 .49 .46 200 .25 .25 .14 7682 .58 .77 .78 100 .25 .25 .13 7582 .48 .60 .60 200 .25 .24 .13 7282 .34 .41 .39 500 .25 .24 .13 7282 .35 .43 .41 500

42

Table A4.9
Comparison of 1% Trimmed Dividend Per Share (in Rs.) by Payer Type During 1991 - 2001
Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Current Payers Initiators Regular Payers Total Payers Mean Median SD Firms Mean Median SD Firms Mean Median SD Firms Mean Median SD Firms 0.18 0.15 0.11 325 0.22 0.20 0.11 924 0.21 0.20 0.11 1249 0.18 0.17 0.09 35 0.15 0.12 0.09 318 0.22 0.20 0.11 1155 0.21 0.20 0.11 1508 0.18 0.15 0.11 93 0.13 0.12 0.09 406 0.21 0.20 0.11 1292 0.19 0.17 0.11 1791 0.18 0.15 0.10 186 0.13 0.11 0.09 566 0.21 0.20 0.12 1539 0.19 0.17 0.11 2291 0.18 0.15 0.11 274 0.12 0.10 0.10 579 0.21 0.19 0.12 1869 0.19 0.15 0.12 2722 0.19 0.16 0.12 332 0.12 0.10 0.12 379 0.22 0.20 0.14 1948 0.20 0.17 0.14 2659 0.20 0.16 0.14 376 0.15 0.11 0.14 212 0.23 0.20 0.16 1737 0.22 0.18 0.16 2325 0.21 0.16 0.18 413 0.16 0.11 0.17 170 0.26 0.22 0.20 1458 0.24 0.19 0.19 2041 0.25 0.17 0.24 439 0.20 0.12 0.23 214 0.28 0.22 0.23 1292 0.27 0.20 0.23 1945 0.27 0.17 0.30 512 0.20 0.12 0.23 172 0.33 0.23 0.31 1243 0.31 0.22 0.30 1927 0.26 0.16 0.25 409 0.15 0.11 0.15 74 0.31 0.22 0.26 1001 0.29 0.22 0.26 1484

Figure A4.6
Distribution of Firms' Payout % During 1990 - 2001
0 0 - 10 75 - 100 10 - 20 100 - 200 20 - 30 > 200 30 - 40 40 - 50

120 100

50 - 75

% of Firms

80 60 40 20 0
1990 1992 1994 1996 1998 2000

Year

Table A4.10
Payout Recurrence of Firms During 1990 - 2001
Payout % No. of Firms Paid out Dividend for Years 1 2 3 4 5 6 7 8 9 0 1336 1167 829 688 440 318 181 82 67 0-10 635 231 95 50 24 18 15 11 4 10-20 967 482 226 120 75 45 24 17 5 20-30 1079 538 282 168 95 55 25 7 6 30-40 1154 514 235 116 38 24 13 2 40-50 1032 374 148 59 18 11 3 2 50-75 1102 451 167 98 39 24 5 1 75-100 740 219 62 32 10 5 1 1 100-200 409 57 6 1 > 200 206 19 3 Total 8660 4052 2053 1332 739 500 267 123 82 * Hindustan Lever Limited paid out 50-75% for all the 12 years Total 10 39 2 3 11 29 2 1 12 12 5188 1087 1965 2255 2096 1647 1888 1070 473 228 17897

1*

44

32

13

43

Transition Probabilities for Pay Out Groups Based on Changes from 1990 to 2001
Pay out 1990 Up to 0 0 - 10 Up to 0 0.71 0.01 0 - 10 0.27 0.07 20-Oct 0.35 0.04 20 - 30 0.41 0.03 30 - 40 0.40 0.02 40 - 50 0.32 0.02 50 - 75 0.47 0.02 75 - 100 0.44 0.06 100 - 200 0.67 > 200 0.40 20-Oct 20 - 30 0.03 0.09 0.15 0.20 0.17 0.11 0.07 0.15 0.08 0.09 0.04 0.14 0.06 0.10 0.09 0.09 0.11 0.40 2001 30 - 40 40 - 50 50 - 75 75 - 100 100 - 200 > 200 0.06 0.02 0.05 0.02 0.01 0.01 0.14 0.02 0.07 0.02 0.03 0.03 0.11 0.10 0.07 0.04 0.01 0.01 0.08 0.10 0.12 0.03 0.03 0.01 0.09 0.09 0.13 0.07 0.02 0.01 0.16 0.07 0.14 0.05 0.04 0.02 0.04 0.09 0.14 0.04 0.03 0.15 0.09 0.06 0.03 0.22 0.20

Table A4.11

Table A4.12
Comparison of Average 1% Trimmed Dividend Payout by Payer Type During 1991 - 2001 (in %)
Year Current Payers Initiators Regular Payers Total Payers MeanMedian SD FirmsMean Median SD Firms Mean Median SD Firms Mean Median SD Firms 1991 29.51 24.82 22.69 329 31.87 27.89 20.02 926 31.25 27.08 20.77 1255 1992 33.77 32.87 20.98 33 26.77 22.56 21.58 318 33.49 29.20 22.45 1162 32.08 27.84 22.39 1513 1993 33.62 32.32 22.98 93 37.59 32.00 26.27 411 35.50 31.25 22.86 1294 35.88 31.50 23.70 1798 1994 31.49 29.00 20.69 185 32.19 25.93 25.38 563 34.70 30.07 21.70 1554 33.83 29.25 22.60 2302 1995 32.05 28.34 20.29 274 34.39 30.00 23.72 587 31.99 27.66 19.44 1877 32.51 28.02 20.53 2738 1996 32.90 28.11 23.18 334 41.75 34.34 28.04 393 35.08 29.63 23.14 1956 35.79 29.85 24.05 2683 1997 37.86 29.81 27.99 377 46.13 34.85 40.40 218 43.74 35.45 33.96 1749 43.02 34.70 33.81 2344 1998 38.86 29.69 32.23 407 37.83 30.99 30.52 179 39.01 33.03 28.06 1472 38.88 31.96 29.14 2058 1999 38.42 31.55 28.74 447 34.54 27.65 28.14 221 39.33 33.15 27.02 1294 38.58 32.51 27.57 1962 2000 40.37 34.45 28.47 512 39.93 36.36 29.60 171 38.29 31.58 26.74 1262 38.98 32.41 27.47 1945 2001 37.71 32.42 28.18 410 32.56 27.87 25.33 74 37.98 31.03 28.08 1014 37.64 31.26 27.98 1498

Table A4.13
1% Upper Trimmed Dividend Yield (%) by Payer During 1991 - 2001
Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Current Payer Initiator Regular Payer Total Mean Median SD Firms Mean Median SD Firms Mean Median SD Firms Mean Median SD Firms 6.06 6.21 2.95 61 4.45 4.00 2.47 480 4.63 4.09 2.57 541 4.91 4.13 2.03 17 5.69 5.48 2.97 36 4.11 3.50 2.38 578 4.22 3.67 2.44 631 1.81 1.60 1.17 43 1.51 1.31 1.00 30 1.50 1.23 0.98 642 1.52 1.24 1.00 715 5.91 5.35 3.51 94 6.82 6.93 3.08 45 4.70 3.91 3.19 733 4.94 4.17 3.27 872 5.09 3.85 3.79 147 6.14 5.78 3.75 54 4.18 3.15 3.31 941 4.39 3.33 3.43 1142 3.34 2.79 2.13 187 3.45 3.17 1.91 54 3.12 2.68 2.06 1100 3.16 2.71 2.07 1341 5.40 4.55 3.42 225 5.88 5.48 3.92 38 5.00 3.99 3.59 1132 5.09 4.22 3.58 1395 8.11 6.77 5.41 227 7.38 5.08 5.47 21 6.64 5.45 4.87 924 6.94 5.73 5.02 1172 8.07 6.43 6.12 199 10.01 8.82 7.31 20 7.35 6.09 5.51 777 7.55 6.11 5.69 996 8.12 6.34 6.29 219 8.51 6.50 7.72 24 7.76 6.74 5.77 703 7.86 6.57 5.94 946 6.17 5.02 4.76 210 5.44 3.66 5.09 19 5.77 4.83 4.37 635 5.86 4.85 4.48 864

44

Figure A5.1
Average Percentage Payoff of Payers
Current Payers Initiators Regular Payers Total

20 18 16 14 12 10 8 6 4 2 0 1991 1993 1995 1997 1999 2001

% Payoff

Year

Figure A5.2
Average Percentage Payoff of Non-Payers
Current Non-Payers Former Payers Total Non-Payers Never Paid 15 10 5 0 -5 1991 -10

Average % Payoff

1993

1995

1997

1999

2001

Year

Figure A5.3
Average 1% Trimmed EPS by Payer Type
Current Payers Regular Payers 16 14 12 10 8 6 4 2 0 1991 1993 1995 1997 1999 2001 Initiators Total Payers

Average Trimmed EPS

Year

Figure A5.4

45

Average 1% Trimmed EPS by Non-Payer Type

Current Non-Payers 2.0 1.0 Never Paid Former Payers Total Non-Payers

Average EPS

0.0 -1.0 -2.0 -3.0 -4.0 -5.0 1991 1993 1995 1997 1999 2001

Year

Figure A5.5
Growth (Assets) by Payer Type
Current Payers 160 140 120 Initiators Regular Payers Total Payers

Average

100 80 60 40 20 0 1991 1993 1995 1997 1999 2001

Year

Figure A5.6
Growth (Assets) by Payer Type
Current Non-Payers Never Paid 300 250 200 Former Payers Total Non-Payers

Average

150 100 50 0 -50 1991 1993 1995 1997 1999 2001

Year

46

Figure A5.7
Assets by Payer Type
1800 1600 1400 1200 Current Payers Initiators Regular Payers Total Payers

Average

1000 800 600 400 200 0 -200 1991 1993 1995 1997 1999 2001

Year

Figure A5.8
Assets by Non-Payer Type
300 Current Non-Payers 250 200 150 100 50 0 1991 1993 1995 1997 1999 2001 Never Paid Former Payers Total Non-Payers

Average

Year

47