Conservatism in Tehran Stock Exchange Finance Essay

The aim of this study is to investigate the relationship between accounting conservatism and dividend policy in Tehran Stock Exchange. To do so, a sample of 90 firms is selected to be studied during the period of 2004 to 2011. Basu (1997) model is extended for measuring accounting conservatism. The results show the there is not significant relationship between accounting dividend policy and conservatism in TSE.

Key words: accounting conservatism, dividend policy and Tehran Stock Exchange

1. INTRODUCTION

Accounting conservatism has been long a subject of scrutiny by both standard setters and accounting researchers. Recently more debates have been on accounting conservatism as a result of standard setter’s decision to omit it from conceptual framework. Some argues that accounting conservatism has information content while others believe that it does not benefit for stockholders (Shorevarzi et al., 2009).

Accounting conservatism is traditionally defined by the adage "anticipate no profit, but anticipate all losses" (Watts, 2003). Givoly and Hayn (2000) define accounting conservatism as "the choice of an accounting approach that in uncertainties leads to recognize the least assets and revenue and have the lowest positive impact on equity". According to Basu (1997), accounting conservatism is "asymmetric timeliness of loss and gains recognition". Gao (2011) defines conservatism as "more supportive evidence for good news than bad news". However, managers face many incentives that affect their tendency for good or bad news recognition. They may recognize bad news timely for tax purpose and saving their reputation or delay it for avoiding credit rate reduction which increases their cost of capital (Zhen Qi, 2011).

On the other hand, dividend policy is one of the most important areas in finance literature. Dividends can play a useful role in the world of significant agency problems between corporate insiders and outsiders. Insiders return corporate earnings to investors by paying dividends and hence are no longer capable of using these earnings to benefit themselves. Payment of dividends makes companies to need capital gathering, and hence gives outside investors an opportunity to exercise some control over the insiders at that time (Easterbrook, 1984). In this regard, Booth and Chang (2011) find that since the mid-1980s the difference in information asymmetry between dividend- and non-dividend-paying firms has increased sharply, before the mid-1980s the market did not differentiate strongly between them, but subsequently the market has reacted less negatively to announcements by dividend payers.

The linkage between accounting conservatism and dividend policy is important because both can be served as a tool for reducing agency cost. More conservatism is expected to reduce agency cost while reduce earning management. In addition, more payout policy can also reduce agency cost through reduction of renounces under control of mangers. Managers may not have a willing to distribute cash to maintain their control in firm which in turn may cause that they use these resources to invest in unfavorable investments. In addition, anticipating all future losses cause the reduction of income and consequently reduces dividend payout (Watts, 1993).

However, it is expected that more dividend payout results in more conservatism in case they serve as a tool to reduce agency cost. The main aim of the study is to investigate this relationship.

2. LITRATURE REVIWE

Mashayekhi et al. (2009) investigated the relationship between conservatism and dividend level and earnings persistency. They found that increasing conservatism leads to decreasing dividend. Moreover, they indicated that decreasing earnings persistency does not result in increasing conservatism. Ahmed et al. (2002) investigated the role of accounting conservatism in mitigating bondholder-shareholder conflicts over dividend policy and in reducing debt costs. They find that firms facing more severe conflicts over dividend policy tend to use more conservative accounting. Furthermore, they document that accounting conservatism is associated with a lower cost of debt after controlling for other determinants of firms’ debt costs. Frankel et al. (2008) find that the asymmetric timeliness of cash flows is significantly related to the asymmetric sensitivity of shareholder payouts. On the basis of their other measures of conservatism (earnings skewness and accumulated nonoperating accruals) they show that these measures are also not significantly related to the sensitivity of shareholder payouts given bad news. Their results suggest that accounting policies do not significantly constrain shareholder distributions conditional on news that does not affect cash flows. Dewenter and Warther (1998) find that information asymmetries or agency conflicts affect dividend policy. Li and Zhao (2008) find that firms that are more subject to information asymmetry are less likely to pay, initiate, or increase dividends, and disburse smaller amounts. Iyengar and Zampelli (2010) investigated whether or not there is a link between conservative accounting practices and the sensitivity of executive pay to accounting performance. They find that the sensitivity of executive pay to accounting performance is higher for firms that report conservative accounting earnings.

3. HYPOTHESES DEVELOPMENT

H1.There is a relationship between accounting dividend payout and conservatism in firms listed in TSE.

To investigate above hypothesis in detail, we break accounting conservatism into timeliness of bad news recognition and timeliness of good news recognition.

It is expected that recognition of bad news in a firm affects dividend payout. Firms facing bad conditions in business have less tendency to distribute dividend and vice versa. However, following hypothesis is posited:

H1a. The timeliness of bad news recognition is negatively related to dividend payout.

In addition, it is expected that recognition of good news in a firm affects dividend payout. Firms facing bad conditions in business have more tendency to distribute dividend and vice versa. However, following hypothesis is posited:

H1b. The timeliness of good news recognition is positively related to dividend payout.

4. METHODOLOGY AND DATA COLLECTION

Present study is applied research regarding classification based on goal. The aim of the applied research is to develop applying knowledge in the given subject. In addition, since the study tries to find a relationship between dividend payout and accounting conservatism, the method of study is empirical correlation. The aim of this sort of study is to determine the relationship between the research variables. Using regression and Pierson correlation the relationship between dividend payout and accounting conservatism is investigated. To do so, firstly Basu (1997) is extended to include dividend. After testing first hypothesis using model 1, research sample is separated into two groups namely the firms with negative return to capture asymmetric timeliness of bad news and the firms with positive return to capture asymmetric timeliness of bad news.

The population of the study includes all listed firms in TSE. However, to have a reasonable homogenous population, the following conditions are considered in data gathering:

1. Information must be available for the past 8 years.

2. Fiscal year must be ended at the end of year.

3. Transaction intervals must not be more than 3 month.

4. Data must be available for testing hypotheses.

5. Sample firms must not be investment, bank or financing firm.

As a result of these conditions a sample of 70 firms was obtained and studied during the period between 2004-2012. Literature and conceptual framework were gathered by documental method. Financial statement and notes issued by TSE were used as a research tool. In addition, Rahavarde Novin software was applied to extract the research data.

Models and variables definition

To test the main hypothesis, following Zhen Qi (2011) methodology Basu (1997) model is extend to capture dividend policy effect on conservatism as following:

(1)

Where:

EARN is firms’ net income before extraordinary items, RET is annual buy-and-hold returns, D is an indicator variable equal to 1 if returns (RET) are negative, and 0 if returns are positive, DIV is firms dividend policy captured by dividing cash dividend by net income. The coefficient of the three-way interaction between DIV, returns, and the negative return dummy captures the difference in the degree of asymmetric timeliness of earnings between high dividend payers firms and low dividend payers firms. We predict that the coefficient of the interaction between RET, D, and DIV (α7) is negative and significant, indicating that high dividend payers firms report earnings in a less conservative manner.

First and second sub-hypotheses are tested through following regression for bad news and good news separately.

(2)

Where all variables are defined as previous model, negative and significant interaction between DIVÃ-RET proves sub-hypothesis 1, and positive and significant interaction between DIVÃ-RET proves sub-hypothesis 2.

5. EMPRICAL RESULTS

5.1. Descriptive statistic

Descriptive statistic only reports an image of data distribution of research and does not report the relationships between variables. Research descriptive statistic is shown in Table 1.

Table 1. Descriptive statistic

DIV

D

EARN

RET

Mean

1.049526

0.354167

0.170649

34.03626

Median

0.670000

0.000000

0.149539

12.80000

Maximum

100.0000

1.000000

6.762778

820.1600

Minimum

9.432275

0.000000

-2.695823

-79.64000

Std. Dev.

4.197662

0.478592

0.363669

82.89660

Skewness

19.38364

0.609850

9.517949

3.534478

Kurtosis

440.3771

1.371917

176.0665

23.83058

Coefficient of variation

3.999579

1.351317

2.131094

2.435538

Notes: DIV is dividend, D is dummy variable, EARN is net income before extraordinary items, RET annual buy-and-hold returns.

5.2. Correlation

Correlation between research variables is presented in Table 2.

Table 2. Correlation Matrix

DIV

D

EARN

RET

DIV

1.000000

D

0.013203

1.000000

EARN

-0.028274

-0.008812

1.000000

RET

-0.037736

-0.507693

-0.043648

1.000000

As it is shown in Table 2, the correlation between variables is negative except for the relationship between dividend and dummy variable. The low correlation between variables also shows that there is not collineary problem.

5.3. Hypotheses test

H1.There is a relationship between dividend payout and accounting conservatism in TSE.

To test this hypothesis model 1 regressed. The results of model regression are presented in Table 3.

Table 3. Regression of the relationship between conservatism and dividend payout

Variable

Coefficient

Std. Error

t-Statistic

Prob.

C

0.193784

0.020928

9.259409

0.0000

RET

-0.000277

0.000194

-1.430914

0.1529

D

-0.087339

0.036886

-2.367784

0.0182

RETD

0.001552

0.000255

6.078255

0.0000

DIV

-0.001661

0.004058

-0.409375

0.6824

DIVRET

-1.59E-05

7.94E-05

-0.200868

0.8409

DIVD

0.010483

0.020668

0.507205

0.6122

DIVRETD

0.000177

0.000294

0.601002

0.5480

R-squared

0.053801

Adjusted R-squared

0.044432

Durbin-Watson

2.140430

F-statistic

5.742835

Prob(F-statistic)

0.000002

The results of hypothesis test show that there is not significant relationship between dividend policy and accounting conservatism. This result indicates that dividend policy of TSE firms does not impact on their conservatism. However, our hypothesis is rejected. The value of Durbin-Watson stat (2.14) shows that there is not auto-correlation problem in models residual.

H1a. The timeliness of bad news recognition is negatively related to dividend payout.

To test this hypothesis model 2 regressed for only negative return sample. The results of model regression are presented in Table 4.

Table 4. Regression of timeliness of bad news recognition and dividend payout

Variable

Coefficient

Std. Error

t-Statistic

Prob.

C

0.082625

0.057761

1.430449

0.1538

RET

-0.003748

0.001867

-2.008257

0.0457

DIV

0.006796

0.028555

0.237990

0.8121

DIVRET

0.000251

0.000413

0.609650

0.5426

R-squared

0.016612

Adjusted R-squared

0.004858

F-statistic

1.413342

Durbin-Watson stat

2.106888

Prob(F-statistic)

0.239367

The results of hypothesis test show that there is not significant relationship between bad news recognition and dividend payout. However, the hypothesis is rejected. The value of Durbin-Watson stat (2.10) shows that there is not auto-correlation problem in models residual.

H1b. The timeliness of good news recognition is positively related to dividend payout.

To test this hypothesis model 2 regressed for only positive return sample. The results of model regression are presented in Table 5.

Table 5. Regression of timeliness of good news recognition and dividend payout

Variable

Coefficient

Std. Error

t-Statistic

Prob.

C

0.189575

0.015672

12.09624

0.0000

RET

-0.000210

0.000146

-1.436260

0.1516

DIV

-0.001440

0.003029

-0.475559

0.6346

DIVRET

-2.30E-05

5.93E-05

-0.387659

0.6984

R-squared

0.007079

Adjusted R-squared

0.000546

F-statistic

1.083629

Durbin-Watson stat 1.803134

Prob(F-statistic)

0.355669

The results of hypothesis test show that there is not significant relationship between good news recognition and dividend payout. However, the hypothesis is rejected. The value of Durbin-Watson stat (1.80) shows that there is not auto-correlation problem in models residual.

6. DISCUSSION AND CONCLUSION

The aim of this study was to investigate the relationship between accounting conservatism and dividend policy in Tehran Stock Exchange. To do so, a sample of 90 firms is selected to be studied during the period between 2004 to 2011. Basu (1997) model was extended for measuring accounting conservatism. The results show the there is not significant relationship between accounting dividend policy and conservatism in TSE. These results suggest that mangers of TSE firms do not use dividend and conservatism as a tool for reduction of agency cost at the same time. In addition, the results in detail show that not only timeliness of good news does not affect dividend policy nor timeliness of good news does. This result may be from the fact that TSE firms smooth dividend payout regardless good or bad news. On the whole, in can be suggested from the results that inferred that TSE firms do not use dividend and conservatism as a tool for reduction of agency cost at the same time, to do so.