CHAPTER 5 DATA ANALYSIS OF LINTNER MODEL

Transcription

1 CHAPTER 5 DATA ANALYSIS OF LINTNER MODEL In this chapter the important determinants of dividend payout as suggested by John Lintner in 1956 have been analysed. Lintner model is a basic model that incorporates the dominant determinants of corporate dividend decisions. Lintner surveyed corporate Chief Executive Officers and Chief Financial Officers and found that dividend policy is an active variable because managers believe that stable dividends lessen negative investors reactions. The active determination of dividend policy implies that the levels of retained earnings and savings are dividend decision byproducts. According to him shareholders prefer smoothened dividend income. Firms are primarily concerned with stability of dividends. Managers appear to believe that market puts a premium on firms with stable dividend policy. Lintner observed that earnings were most important determinants of any change in dividends. Lintner also reported that majority of managers develop long-term payout ratio targets and use periodical partial adjustments to reach target levels. The enormity of the percentage change in payout levels is more important than the absolute solar amount. Based on interview of 28 management teams, Lintner reported a target payout ratio of 50%. In addition, most of the firms had a standard with respect to speed at which they would move toward their payout targets; these adjustments ranged from one-sixth to onehalf. For example, let s assume that target payout was 50 % and that the speed of adjustment factor was 25%. Using these parameters, if EPS increased from $ 2 to$ 3, the first year dividend increase would be ($1)(0.50)(.25) or $ This calculation represents the earnings increase times the target payout ratio times the adjustment factor. The full $.50 dividend increase (EPS change times target payout ratio) might take several years to achieve assuming the new EPS level remains same. If there was any sudden unexpected increase in earnings firms adjusted their dividends slowly. Firms were reluctant to cut dividend. The firms set dividend policy first and other policies were then set, taking dividend policy as given. Lintner developed a model, which was able to explain 85% of dividend changes in his sample of companies. 179

2 D* it = α i E it (5.1) D it D i(t-1) =a i + C i {D* it -D i(t-1) }+ u it (5.2) Where, D* it = desired dividend payment during period t D it = Actual dividend payment during period t α i = Target payout ratio E it = Earnings of firm during period t a i = a constant related to dividend growth C i = partial adjustment factor u it = error term The constant a i will be zero for some companies but will generally be positive to reflect the greater reluctance to reduce than to raise dividends which was commonly observed as well as the influence of the specific desire for a gradual growth in dividend payments found in about a third of the companies visited. The variable u it represents the discrepancy between the observed change in D it and that expected on the basis of other terms in equation. This model can further be simplified in the form of a multiple regression equation D it D i(t-1) =a i + C i {D* it -D i(t-1) }+ u it (5.3) D it D i(t-1) = a i + C i { α i E it -D i(t-1) }+ u it (5.4) D it= a i + α i C i E it + (1-C i ) D i(t-1) +u it (5.5) Lintner model gave three important results: Stabilize dividends with gradual, sustainable increase whenever possible Establish an appropriate target payout ratio Avoid dividend cuts, if at all possible. 180

3 This model remains the best description of dividend setting process available. Thus, even after 5 decades, Lintner s research remains the definitive study of management behavior finance Classic. Nevertheless, Lintner s model failed to explain why companies pay dividends. Professor Lintner s main finding is that low values of r can be attributed to smaller companies emphasizing on internal financing of investment projects either as a matter of policy or because of the difficulty of access to the investment market, will choose low target payout ratios; that is widely differing target payout ratios can be justified in terms of the particular internal- external financing policy pursued by the company. The reaction coefficient c is mostly based on the variance of net income, managerial attitudes towards expectations and the degree of emphasis laid on stable dividend rates. Companies enjoying relatively stable net income are likely to choose high reaction coefficients and respond quickly to changes in net income. The size of c in case of companies subject to large fluctuations in their net income depends on their ability and willingness to maintain stable dividend rates. Companies interested in reducing fluctuations in dividend rates have to choose very low reaction coefficients; whereas those companies which are not interested in dividend stability ; a luxury they cannot afford will have high reaction coefficients, allowing dividends to fall in line with their net income. One might expect that companies belonging to the same industry are likely to follow a similar financial policy derived from common investment requirements; this in fact is not the case. 5.1 STATISTICAL MODEL DEVELOPED TO TEST LINTER DIVIDEND POLICY MODEL The regression equation developed to test the model: Y=α + β 1 X 1 + β 2 X 2 (5.6) Where, Y= dependent variable (equity dividend during period t) X 1 = independent variable (PAT) α = Constant 181

4 β 1 = regression coefficient of PAT (target payout ratio *adjustment factor) X 2 = Equity dividend during period t-1 β 2= regression coefficient of dividend paid during period t-1 i.e. (1-c) and c is the adjustment factor. Target payout ratio* adjustment factor = β 1 αi* Ci = β 1 αi*(1-β 2 ) = β 1 This implies α i = target payout ratio = β 1 /(1-β 2 ) Regression coefficient of PAT (β 1 ) will highlight the target payout ratio of Information Technology, FMCG and service sector and with the help of regression coefficient of lagged dividend (β 2 ), we can obtain the partial adjustment factor as discussed in the Lintner model. If α =0 and C i =1, the actual changes in dividends coincide with the desired changes. Conversely, if C i =0, no changes in dividends towards the desired level are undertaken. The hypothesis that firms gradually adjust dividends in response to changes in earnings implies that the speed of adjustment coefficient C i is with in the range 0<C i >1.Furthermore a positive α represents the management reluctance to decrease dividends. 182

5 5.2 IT SECTOR TABLE 5.1: Regression results of pooled data, Fixed effect one way and Fixed effect two model in IT sector Pooled data Model Fixed effect firm model Fixed effect firm and time model Regressors Regression Coefficient P value Regression Coefficient P value Regression Coefficient P value PAT (Rs ** ** ** Crore) Div (t-1) ** ** T T T T T T T T Constant Adjusted R F statistics ** ** ** Hausman - - χ2 = ** χ2 = ** Statistics LM test - - χ2 = χ2 = F Test Durbin- Watson Target payout ratio.352/(1.169) = =30.11% /( )= =32.53% /( )= =30.90% Adjustment factor ( ) =1.169 ( ) = ( )= Note: * *indicates values significant at10% and 5% level of significance 183

6 5.2.1 DATA INTERPRETATION Table 5.1 shows the results of pooled data, One-way and two way Fixed effect model. The pooled data results depict regression coefficient PAT during the current year is significant at 5% level of significance while regression coefficient of dividend paid during the previous year is significant at 10%.The F statistics tests the validity of the Lintner model in the IT sector. The Durbin Watson test has been applied to examine the existence of autocorrelation. The DW statistics is The test results indicate that there is no problem of serial autocorrelation in the data. The adjusted R square for pooled data analysis is 83.6%. Table 5.1 also presents the regression results of one-way Fixed effect model showing that regression coefficient of dividend paid during previous year and PAT is significant at 5% level of significance. The Adjusted R square is 80%.F statistics is significant at 5% level of significance demonstrating overall validity of the model. The results highlight that there is Low dividend smoothing in this sector as it is characterized by low target payout ratio and high speed of adjustment coefficient. The results fixed effect firm and time model shows that both the independent variables PAT and dividend paid during previous year are statistically significant at 5% level of significance. The value of adjusted R square is 81.44% None of the time dummy is found to be significant at 5% level of significance portraying that there are no time effects. LM test results reported above are statistically insignificant. This implies that classical linear regression model should be preferred to Fixed effects and random effect models. Therefore, the results of pooled data should be used for interpretation. Hausman statistics is significant at 5% level in both the cases indicating that fixed effect models can be used over panel data models. F test results are insignificant at 10% depicting that both firm and time effects are absent. 184

7 5.2.2 CONCLUSION The above analysis shows that firms in IT sector are not averse to cut dividends. Profits seem to be a noteworthy determinant of dividend payments during entire time period under study. PAT is significant at 5% level depicting for all the IT firms profitability is an indispensable factor to be considered while determining dividend policy. But it should always be kept in mind there are other factors also apart from profitability that affect dividend payments. However the regression coefficient of Div (t-1) is insignificant showing that dividend paid during current year is not governed by dividend paid during previous year.the constant is also insignificant at 5% level of significance. This also proves the fact that only one of the variables given by Lintner have an impact on dividend declared by the firms in IT sector. The target payout ratio under all three models hovers around 30%. This payout ratio is lower than target payout ratio of 50 % suggested by Lintner.The target payout ratio of a firm depend on factors like growth and earning prospects of a particular company; the average cyclical movement of investment opportunities; working capital requirements; and internal fund flows judged by past experience; the relative importance attached by management to long-term capital gain as compared with current dividend income for its stock holders; its access to the capital markets on favorable terms, and company policies with respect to use of outside debt and new equity issues; and management s confidence in the soundness of earnings figures as reported by its accounting department, and its confidence in its budgets and projections of future sales, profits; The normal payouts and speeds of adjustment of competitive companies or those whose securities were close substitutes investment wise; the financial strength of the company, its access to the capital market on favorable terms, and company policies with respect to the use of outside debt and new equity issues and so on. As regards the speed of adjustment coefficient the value lies between 1.16 to 1.51 which is much higher compared to what was suggested by Lintner.This high speed of adjustment coefficient denotes that actual changes in the dividend may be much higher than desired changes. This low target payout ratio coupled with high speed of adjustment factor shows the absence of dividend smoothing and dividend signaling. The IT firms dividend payout fluctuates with changes in the earnings. This suggests that higher dividend payout is witnessed in IT 185

8 industry only in the case of increased profitability of the companies. Any variation in the earnings is quickly reflected in dividend payment. Through the results it may be stated that the principal determinant of dividend policy is profitability. Thus, IT firms are using dividends to signal their surge in profit margins over the years. It is further seen that IT companies do not attempt to sustain stable dividends. The dividend payments may fluctuate with the fluctuations in earnings. The regularity of dividend payment and constancy of its rate is not an important objective of the dividend policy followed by IT industry in India. These findings are validated by the fact that IT sector is characterized by low target payout ratio and high speed of adjustment factor. IT sector demonstrates a pattern, which is seen in emerging economies like Tunisia, Zimbabwe and Turkey. These sectors are characterized with low payout ratio and high speed of adjustment coefficient. In nut shell, it can be stated that there is low smoothing and instability of dividend policy in IT sector. Through the results no indication has been obtained that IT firms place a higher importance on stable dividends to signal their future profitability or to minimize agency costs. Low payout ratio and high adjustment factor indicate that IT firms in India frequently change their dividend payments with change sin earnings and dividend smoothing is of low order. 186

9 5.3 FMCG SECTOR Table 5.2: Regression results of pooled data, Fixed effect one way and Fixed effect two model in FMCG sector Pooled data Model Fixed effect firm model Fixed effect firm and time model VARIABLES Regression coefficient P value Regression coefficient P value Regression coefficient P value PAT (Rs ** ** ** Crore) Div (t-1) ** ** ** Constant T T T T T T T T Adjusted R F statistics ** ** Hausman - - χ2= ** χ2= ** Statistics 34 LM test - - χ2= χ2= F Test ** ** Durbin Watson Target payout ratio /0.116= =128.45% /( ) = =71.64% /( ) = =69.67% Adjustment factor (1-.884)=0.116 ( ) = ( )= Note: * *indicates values significant at10% and 5% level of significance 34 Though the values of Hausman statistics and F test have been reported.but these tests loose their significance in interpretation since LM test results are insignificant. 187

10 5.3.1 DATA INTERPRETATION Table 5.2 presents the pooled data, Fixed effect one way and two way regression model results. The pooled data results show that PAT and dividend paid during previous year are significant at 5% respectively. The Durbin Watson statistics is The F statistics value is denoting the validity of Lintner model in the FMCG sector. The adjusted R square has assumed a value of 95.3% which denotes that Lintner Model explains 95% of variations of dividend behavior in FMCG sector. The regression results of Fixed effect firm(one way) model shows that PAT and dividend paid during previous year are significant at 5% level of significance. The value of Adjusted R square is 96.49%. The F statistics are also significant at 95% confidence interval showing the overall validity of the model in the FMCG sector. The fixed effect firm and time estimations highlight that both the independent variables PAT and dividend paid during previous year are statistically significant at 5%The value of adjusted R square 96.39% however none of time dummy variable is significant at 5% level of significance indicating complete absence of time effects. LM test results reported above are statistically insignificant; the results of pooled data should be preferred for interpretation. Hausman statistic values are and in Fixed effect one way and two way model respectively which is significant at 5% level in both the cases indicating that fixed effect models can be used over random effect models. F test results are significant at 5% depicting that both firm and time effects are present CONCLUSION An analysis of the data shows that earnings of the present year and equity dividend of the preceding year have significance on the dividend to be distributed during the current year. This fact is validated by the findings of the study undertaken since the regression coefficients of both independent variables are found to be significant at 5% level of significance. The intercept term (constant) in the present study does not gain any significant coefficient during the study. Perhaps; this might be attributed to the small sample size and the nature of the firms under consideration. This also proves the fact that Lintner model, which is a finance classic, holds good in the Indian FMCG 188

11 industry. This result is in contrast to the results of Mahapatra and Sahu (1993) but corroborates the result of Mishra and Narender (1996)[103][115]. The target payout ratio derived from β 1 hovers with in the range of 69 to 128%. The partial adjustment factor is found to be.115 to.52 The partial adjustment factor obtained from the analysis lies between one sixth and one half as highlighted by Lintner in his findings. Due to strong bias against dividend cuts, increase in earnings is translated into increase in dividends only gradually to avoid future downward revision. This lag in adjustment of current dividends to increase in earnings is a kind of safety device designed to make dividends a function of permanent earnings rather than transitory earnings that cannot be sustained. From the study, it may be stated that the principal determinant of dividend policy is profitability. Thus, FMCG firms are using dividends to signal their surge in profit margins over the years. It is further seen that FMCG companies attempt to sustain stable dividends. The regularity of dividend payment and constancy of its rate is an important objective of the dividend policy followed by FMCG industry in India. This can be attributed to the fact that regression coefficient of lagged dividend is significant at 5% level. These findings are in conformity with Mookerjee Rajen (1992) who found that Lintner model is successful in explaining dividend payout behavior of private corporate sector. However, he found that explanatory power of the model was significantly enhanced by the inclusion of external finance as an explanatory variable. This in turn suggests that firms in India base dividend decisions on the availability of external finance. The amount of dividend distributed during the previous year is a significant factor governing the dividend distribution of the current year. The results signify that the dividend signaling and smoothing effects are present in the FMCG sector. These firms follow stable dividend payments year on year basis, even though earnings might change dramatically. The findings in the FMCG sector are in alignment with Brave et.al that mangers are very reluctant to cut dividends once they are initiated. This reluctance leads to dividends that are sticky, smoothed from year to year and tied to long run profitability of the firm. The results in this sector is in conformity with the pattern seen in developed economies like USA and Germany.Fama and Babiak (1968) in USA found an average speed of adjustment of approximately 0.37, slightly higher 189

12 than Lintner finding of 0.30, and a target payout ratio of 50%, almost equal to one found by Lintner.Behm and Zimmermann (1993) tested the partial adjustment model for a sample of 32 German listed firms. They found the speed of adjustment between 0.13 and.58, and a target payout ratio of between 25and 58%.Varouj Aivazian, Laurence Booth, Sean Cleary (2003) also found Lintner model works remarkably well fro US markets[116] 5.4 SERVICE SECTOR TABLE 5.3: Regression results of pooled data, Fixed effect one way and Fixed effect two model in Service sector Pooled data Model Fixed effect firm model Fixed effect firm and time model Variables Regression P value Regression P value Regression P value coefficient coefficient coefficient PAT (Rs ** ** ** Crore) Div (t-1) ** ** ** T T T T T T T T Constant Adjusted R F statistics ** ** Hausman - - χ2 = ** χ2 = ** Statistics LM test - - χ2 = χ2 = F Test Durbin Watson Target payout ratio.130/.369 = =35.23% ( )/( ) = =31.585% ( )/( ) = =33.26% Adjustment factor (1-.631)= = ( )= Note: * *indicates values significant at10% and 5% level of significance 190

13 5.4.1 DATA INTERPRETATION Table 5.3 presents the pooled data, Fixed effect one way and two way regression model results. The pooled data results demonstrate that PAT and dividend paid during previous year are significant at 5% respectively. The Durbin Watson statistics is1.959.the F statistics value is which is significant at 5% denoting the overall validity of regression model developed in the Service sector. The adjusted R square has assumed a value of 76.7%. The regression results of Fixed effect firm model shows Dividend paid during the previous year and PAT both are significant at 5% level of significance. The value of adjusted R Square is 76.18% The overall validity of the model has been found out with the help of F statistics which is significant at 5% level of significance The fixed effect firm and time estimations highlight that both the independent variables PAT and dividend paid during previous year are statistically significant at 5%The value of adjusted R squared 75.90%. None of time dummy variable is found to be significant at 5% level of significance indicating absence of time effects. LM test results reported above are statistically insignificant; the results of Classical Linear regression should be preferred over fixed effect and random effect models..hausman statistics value is and which is significant at 5% level respectively in both the cases indicating that fixed effect models can be used over random effect models.f test results are insignificant at 5% depicting the absence of firm and time effects CONCLUSION An analysis of the pooled data shows that earnings of the present year and equity dividend of the preceding year are significant factors affecting the dividend to be distributed during the current year. This fact is validated by the findings since the regression coefficients of both independent variables are found to be significant at 5% level of significance. It also reflects the companies ensure that the current dividend signals their performance and the desire of management to maintain a stable dividend. 191

14 The target payout ratio derived from β 1 lies between the broad range of 31 to 35%. This is relatively a low target payout ratio as compared to what was suggested by Lintner. The partial adjustment factor is found to be with in the range of to The partial adjustment factor obtained lies between one sixth and one half as highlighted by Lintner in his findings. The choice of a particular speed of adjustment factor depends upon possible variations in net earnings after tax and stability of dividends required. Stable net earnings after tax would induce a management to choose a higher adjustment coefficient.but if net earnings are subject to wide fluctuations, a desire to have stable dividend would lead to choosing lower adjustment coefficient. It may be stated that the principal determinant of dividend policy is profitability. The results signify that Service sector companies score high on dividend stability.. Firms in service sector use cash dividends to signal their prosperity to the shareholders. The results are in agreement with the previous studies on banking industry, which state that banks use their dividend history to set their dividend. These results were established by Dickens N.Ross and Newman.A.Joseph in their study Bank Dividend policy: explanatory factors and Pal Karam and Goyal Puja Leading determinants of Dividend policy: A case study of the Indian Banking Industry. Their study displayed that stable dividend policy is followed by Indian banking industry as lagged dividend emerged as the most significant determinant of dividend payout[116]. Another study by Bodla B.S.,Pal Karam,Sura S Jasvir Examining Application of Lintner s Dividend Model in Indian Banking Industry highlighted that major determinants of current dividends are lagged dividends and current earnings.[117] The evidence found through the study is almost similar to the markets in developed countries such as US and is in support to the argument of information content of dividend in context of dividend proceeds.hence,dividend policy can be used as a signaling device by the management of banks. Thus, through the analysis and the discussions presented an endeavor has been made to empirically test the Lintner model in three major sectors of India. Thus, the study unearths the applicability of smoothing and signaling approaches and relevance of information asymmetry models in IT, FMCG and Service sectors in India. 192