Overconfidence, managerial accounting competence and dividend smoothing: evidence from emerging markets Open Access

The traditional method regarding dividends is based on the Lintner (1956) research, which show that firm managers change the payment of dividends based on long-term goals and current earnings. Based on manager’s perspective, the dividend is reduced only if there is no substitute procedure and increase in the dividend happen when there is certainty about the future cash flows stability. Capital market practitioners place a higher value on companies with uniform dividend process, and the market considers the reduction of dividends by firms as a communication of unfavorable information (Guttman et al., 2010). For this purpose, the managers first determine the dividend and then based on that, they adjust the decisions related to liquidity up to the determined level. In this regard, Brav et al. (2005) concluded that the of dividend smoothing plays a fundamental role in listed firms.

Investors consider the reduction of dividends as bad news, which is caused by information asymmetry between firm’s manager and capital market practitioner. Informed managers classify the earnings between dividends and capital expenditures and increase the value of stock in the long term. According to their predictability, dividends are a function of the two hypotheses of marking and smoothing. Based on signaling hypothesis, dividends can predict the behavior of future earnings and share price, and the dividend smoothing hypothesis states that the process of dividend is depend on current and past profits (Goddard et al., 2006). Managers do not want to deviate dividends from the standard state, and for this reason, they use dividends as an instrument to communicate data and information about future changes in earnings (Kao and Wu, 1994). The level of dividends is function of managers' future expectations of earnings, and changes in it leads to fluctuate in share prices, while in efficient markets, the firm value is not related to the method of dividend distribution (Miller and Modigliani, 1961). As a result, manager’s ability and investment decisions are the factors that change in dividends approach.

Overconfident managers borrow most of the resources they need because they overestimate the probability of the firm’s future success. Specifically, this group of managers assume that they have important information that the market is not aware of it (Tan, 2017). In this regard, Hribar and Yang (2016) concluded that overconfident managers predict earnings with respect to self-articulate information and the estimated earnings is presented very optimistically. The lack of similarity between managers' and investors' information leads to mispricing of bonds (Myers and Majluf, 1984). Improper pricing of bonds leads to higher costs for shareholders than creditors (Patel et al., 2009) and managers use dividend smoothing in order to keep the risk at a low level in order to control the expectations of shareholders.

In case of dividend smoothing, fluctuations in profits should be absorbed by other factors. Assuming the stickiness of managers' behavior, the accounting competence of firm managers should absorb shocks to profits. In other words, it is expected that the sustainability of dividends will also increase with the increase in the accounting competence of managers. Accounting qualification can be considered the previous experiences of managers who worked as managers or audit partners. Accounting competence of manager show that the executive team have high knowledge and experience about the quality of accounting report and efficiency of internal controls (Albrecht et al., 2018). The inability of companies to adapt to environmental conditions as a result of fluctuations has led to an increase in the risk of incorrect selection for investors, and so, the cost of capital and changes in investors' expectations are formed. By identifying the trend and stability of dividends based on the ability of managers in companies, the created risk can be partially covered (Rashidi, 2020). In emerging economies, the banks have the main role in providing financial resources needed by companies, and in these markets, the level of efficiency is weak and they are at the beginning of the development process (Al-Najjar and Clark, 2017). In this situation, capital market activities are subject to changes, corporate governance reforms and improvement of internal controls guidelines that have been formed based on the implications of relevant regulations and institutions (Rashidi, 2022). In this regard, in 2006, Iran capital market was faced with the provision of corporate disclosure instructions, internal control instructions, audit committee and internal audit instructions. In this situation, the link between the behavioral characteristics and the level of managers’ knowledge with the dividends smoothing is unknown. In our review, the effects of managers' overconfidence and accounting competence on the dividends smoothing have been investigated, and we believe that the results of this research will fill the literature gap related to the behavioral characteristics of managers, and the smoothing of dividends in emerging markets.

The knowledge enhancement of the current research can be expressed in different fields. First, the previous researches emphasize on limiting agency costs by creating limits for the optimistic actions of managers, while framing and standardizing a major part of management behaviors is not possible. In this research, the manager’s executive ability has been examined in the form of accounting competence at the same time as excessive self-confidence, in order to control part of the inherent limitations caused by the managers' behavior. Second, the financial reporting approach is derived from the ability and competence of managers. In other words, the accounting competence of managers is effective on considerations related to distortion in financial statements and smoothing of dividends. Third, the high ability of management by creating opportunities and motivation to achieve personal benefits cause to a negative change in the financial reporting quality, but in our research, the positive aspect of managers' ability in the form of accounting competence is considered. Finally, the relationship between overconfidence and accounting competence of managers with firm characteristics (high information quality firms and those with low information quality) has been review.

The structure of our article is presented as follows. In the first part, we examine the theoretical foundations and prepare the main hypotheses of the research. In the next section, we discuss the sample and methodology used to conduct the research. Then we analyze the experimental results, and finally, we present the conclusions and implications of the research.

To prepare financial reports, earning is considered as an indicator of the firm’s and manager performance. Historical earnings and default dividend payment patterns play a basic role in changing the dividend policy conditions (Skinner and Soltes, 2009). Earnings patterns and dividends have the firm’s value characteristics that are valued by the market. Therefore, policies influence management decision-making (Barth et al., 1999). Managers try to control and increasing dividends regardless of unexpected profits is based on dividend smoothing (Mikhail et al., 2003; Lintner, 1956; DeAngelo et al., 2004, 2006). Because of the fact that managers tend to reduce ongoing obligations in the form of regular profit distributions, a decrease in dividends is a sign of a decrease in the expected profitability of the firm (DeAngelo et al., 1992; Koch and Sun, 2004; Brav et al., 2005). For this reason, in order to reduce information asymmetry and control bad news, manager smooth dividends (Joose and Plesko, 2005) and this action highlight the dividends role in describing future earnings (DeAngelo et al., 1992; Charitou, 2000). In fact, dividend policy regardless of income patterns and dividend distribution, has informational content in describing future incomes, and it shows that changes in dividend pattern improve the ability of current earnings to show the future performance according to historical earning trends. and profits (Allen and Michaely, 2003; DeAngelo et al., 2004; Brav et al., 2005). In this regard, Miller and Modigliani (1961) stated that if an organization adopts a long-term dividend pattern, market practitioners are preferring to interpret the dividend rate as a change in manager’s beliefs about the future performance of firm.

The dividend smoothing literature show that the cause of this behaviors associated with shareholders and managers’ information asymmetry (Guttman et al., 2010). So, in order to control agency costs caused by cash flows (Lambrecht and Myers, 2012) and the financing costs (Aivazian et al., 2006), dividend smoothing is suggested. According to Lintner (1956) and Fama and Babiaik (1968), many companies have a long-term and smooth plan for paying dividends. Firm managers have found that shareholders prefer cash dividend distribution to have a smooth trend over time because the smoothing trend of cash dividend distribution carries the message that the firm has a stable financial position, less business risk, and therefore its stock market value will increase in the near future.

Overconfidence is an individual behavior that could be defined as optimistic and unrealistic (positive) beliefs related to any aspect of an event under situation of uncertainty (Malmendier and Tate, 2005). Most managers with this behavior are so positive into their financial decisions and outcomes, mainly in the investment context (Cooper et al., 1988). Because of this approach, overconfidence managers may predict projects’ cash flows very favorable and hence, estimate projects’ result higher than their actual value. In other ways, they believe that the firm capacity is more than market estimation, and this market misevaluation cause to increase financing cost. For that, overconfident managers may be more inclined to overinvest if the firm has internal resources, but underinvest may take place in condition that project financing requires more external resources (Malmendier and Tate, 2005). Success in the market requires the desire to pursue unrecognized opportunities (March, 1991), but sometimes managers do not have the desire to follow and point out these opportunities. Nevertheless, option plans can be a factor to convince managers to accept risks (Armstrong and Vashishtha, 2012), however the desire to do basic decisions is often restricted by inherent and internal motivations, and so, financial incentives lose their importance (March and Shapira, 1987). The potential trait of overconfidence is the main source of internal motivations and show that managers with overconfidence are more willing to do unique plans that have a potential earnings (Hirshleifer et al., 2012). However, in order to achieve success, only implementing bold investment projects is not enough because the probability of achieving the expected results in these plans is low (Broihanne et al., 2014).

Overconfident managers have a high tendency to implement high-risk projects and postpone unfavorable news and results, which leads to an increase in the scope of risk imposed on investors. According to the agency framework, overconfident managers do not desire procedural stability and fixed budgeting. firms tend to maintain financial strength and reduce the capital markets accessing cost (DeAngelo et al., 2006), and because of this perspective, big firms avoid reducing dividends due to the costs it entails for them (including reduced credit in the capital market) avoid (Daniel et al., 2008). On the other hand, the variable budget approach with regard to contractual obligations and investment in fixed assets that do not change in the short term, causes the need to change the process of dividends to control the risk in the stable threshold.

Companies may be inclined to finance and increase the level of debt in order to use its benefits, but the benefits of cash payments to shareholders are higher than the costs of external financing because dividends reduce information asymmetry between investors. and participation (Acharya et al., 2007). In this regard, Hackbarth (2008) concluded that overconfident managers have a higher tendency to increase the level of debt. Also, Malmendier and Tate (2008) showed that overconfidence cause to do plans that are more likely to reduce the firm value. Deshmukh et al. (2013) concluded that managers’ overconfidence considers external financing costly and therefore tend to reduce dividends.

Managers' competence is usually considered synonymous with their specialized knowledge (Blue Ribbon Committee, 1999). However, the accounting qualification can be considered the previous experiences of managers who worked as managers or audit partners. An accounting qualification provides the management team with high ability and knowledge about the quality of accounting information and firm’s internal controls (Albrecht et al., 2018). Also, academics, researchers and compilers of accounting standards and stock exchange requirements usually concentrate on increasing the level of professional competence and provide error-free financial reports. In fact, the results of previous research such as Aier et al. (2005) show that accounting experience is generally associated with a reduction in misinformation and internal control weaknesses. However, accounting competence has a dual nature in a way that, on the one hand, it can lead to the improvement of accounting knowledge and internal management approaches (lennox, 2005) and on the other hand, the competence of managers can cause to decrease in board of directors’ independence (Hermalin and Weisbach, 1998) due to the manager’s justification power from the financial aspect.

Managers with accounting competence use their superior abilities to hide distortions or prevent adjustments. Accounting competence is considered as an important factor in recognizing management’s ability to override internal controls as an opportunity to distort financial reports (PCAOB, 2015a, b). Accounting qualification does not lead to distortion and misrepresentation of information, because it represents the manager ability to report reliable financial information, and there is no reason to expect changes in honest presentation from managers with accounting competence compared to managers who lack this feature. In contrast, some researches state that accounting competence can interact to other factors of fraud risk (including motivation) to increase the abuse and financial misstatement risk (Albrecht et al., 2018). The existence of a common executive background between managers and auditors can lead to excessive trust in managers and the possibility of discovering mistakes and corrections by the firm is reduced. High trust in managers with accounting competence allows for more flexibility in financial reporting, which indicates the reciprocal relationship between accounting competence and distortion-based incentives (Bergstresser and Philippon, 2006).

The main responsibility of the firm’s senior managers is to set macro and strategic policies, which is accompanied by the motivation and ability to influence accounting decisions (Cheng et al., 2016). The negative aspect of accounting competence is considered as a state of opportunity for fraud and distortion in financial statements. This means that management’s ability to ignore internal controls can be considered as an opportunity to abuse financial reports. Previous experience in the field of auditing as a manager or partner leads to increased knowledge and experience of auditing and recognition of internal controls as a result of interaction with multiple and different clients over time, which leads to the discovery or concealment of deviations in financial statements and reports. Every one of these skills can provide opportunity for managers to manage earnings with no adjusting audit report (at least in the short-term). However, accounting competence could not necessarily cause to distortion, and the potential negative aspects of accounting competence can be seen only when it is combined with other risk factors (Albrecht et al., 2018). Managers with high competence and ability have more skills in cover misstatements due to having experiences in the field of auditing as a manager or partner (Lev et al., 2010). Manipulated financial reports are often prepared by active, professional and innovated people with the high understanding of the firm’s corporate controls and weaknesses. Also, this group of managers have high experience in how to analyze the financial reports, as a result, this category of managers has a high ability to correct distorted financial statements. On the other hand, the common background of managers with independent auditors and more trust in management reduces the discovery of distortion and abuse in the short term (Wolfe and Hermanson, 2004). In this regard, Lennox (2005) showed that the previous job dependence of the management to the auditing firm can lead to a decrease in accounting quality.

Financial and investment decisions made by participants who try to maximize the present value of their future acquired interests. Managers on behalf of the firm are responsible for creating suitable profitability for investors in each period. Based on the theory of conflict of interest, managers are optimistic and do not have loyalty to shareholders, but the existence of controlling shareholders prevents the opportunistic approaches of managers, and for this reason, managers consider their interests according to the limitations in the firm and the capital market (Lambrecht and Myers, 2012). In this regard, managers with accounting competence may use strategies that lead to changes in the quality of accounting information and dividends.

To compose research sample, we prepare overconfidence, manager accounting competence and dividend smoothing data from Codal (“Comprehensive Database of All Listed Companies”) over the years 2012–2022. The managerial accounting competence data is coming from “TSETMC” and same website which collect resume of managers for all the listed firms. Finally, we combine the all data and restrict our sample to firms with no lost data on any of the research variables. Financial firms are also excluded because of their different operate environment and subsequently follow distinctive accounting approaches. Further, we remove the firms that does not have the managerial accounting competence information. Our research sample consists of 1,320 firm-year observations of Tehran Security Exchange firms between 2012 and 2022 years.

For manager accounting competence variable, the definition of Albrecht et al. (2018) was used. Accounting competence is a dummy variable that takes the “one” if the president; any corporate vice president in charge of the major business division, unit or function; any other manager who has a decision-making tasks; or any other individual who performs same decision-making tasks, has previous experience as a manager or partner in an audit firm, and zero otherwise.

Considering that the investment plans of firm is based on managers' decisions and personality approaches, the first criterion was set based on the investment and managers' decisions; this approach is applied in some researches (e.g. Ahmed and Duellman, 2012).

Based on Ben-David et al. (2013), our measure of overconfidence is dummy variable that equal to “one” in case the capital expenditures in year t is more than the middle level of industry capital expenditures for that year, otherwise zero.

Lintner (1956) with respect to partial adjustment hypothesis show that firm managers know the non-persistent nature of firm earnings. Fliers (2019) show that managers tend to dividend uniformity. The previous levels of dividend and current earnings assess the current fluctuate in firm dividends. The optimal level of dividends (Div*) is represent by:

where E_t is the firm current earnings and r is the standard rate of firm payout. Fliers (2019) show that based on high risk of dividend decrease, fluctuate in dividends will be imperceptible. So, we estimate the Fliers (2019) following partial adjustment model:

where Div_t is the dividends’ changes. The target payout ratio of firms is determine by $β1β2$⁠, and the speed of adjustment for dividend is determine by $−β2$⁠.

To test the impact of managerial overconfidence and managerial accounting competence on dividend smoothing, we use the following regression model:

where $AD_DIV$ is the dividend smoothing, while Overconfidence and Acc_Competence are proxy of managerial overconfidence and managerial accounting competence, respectively. To test main hypothesis, research focus on β1 and β2 coefficient. In situation that managerial overconfidence increases the dividend smoothing of research firms, then, the coefficients on Overconfidence (β1) should be positive and significant. If managerial accounting competence increases the dividend smoothing, then, the coefficients on Acc_Competence (β2) should be positive and significant. Following some studies (Fliers, 2019; Rashidi, 2020), we control other variables that affect the dividend smoothing of Iranian firms, too. These variables consist firm size (Size) is the natural log of total assets’ book value, market to book ratio (MTB) is the market value to book value of firm’s total assets, leverage (Lev) is total debts divided by total assets, institutional ownership (Inst) is the sum of institutional ownership’ percentage, managerial ownership (Mgt) is the percentage of the shares owned by the managers, board size (B_size) is the number of the board of directors, board independence (B_Ind) is the independent directors divided by the total number of the board of directors and ROA is the income before extraordinary items adjust by lagged total assets. LOSS is a dummy variable that take the value of “one” for firm with negative income before extraordinary items of t and t−1 years, otherwise zero. Ret_Std show the share returns standard deviation during the t, t−1 and t−2 years. Cfo_Std show the standard deviation of cash flow from operating during the t, t−1 and t−2 years. The standard deviation of profitability change rates over the t, t−1 and t−2 years is used to measure environmental uncertainty (VIX). This measure has been used by researchers such as Dichev and Tang (2009). Finally, regression analysis control for the industry and year effect.  Appendix show the research variables definition.

Table 1 represents the sample descriptive statistics. It shows the descriptive statistics of managerial overconfidence, managerial competence and other variables used in regression analyses. It reports that dividend smoothing (⁠$AD_DIV$⁠) of sample is −0.039 that is in line to prior studies (Shahab et al., 2020), and indicates the change of dividends in order to control the firm’s credit position among investors. The average value of manager overconfidence (Overconfidence) in investing expenditure proxy is 0.537 which is highly comparable with the statistics reported by Ahmed and Duellman (2012).

The average value of managerial accounting competence (Acc_Competence) is 0.242 which is highly comparable with the statistics reported by Ahmed and Duellman (2012). In addition to this, Table 1 contains the descriptive statistics for control variables. The sample ownership structure consists of 71% institutional shareholders, and the variable of MGT is 66.6%. The environmental uncertainty shows the low sales process fluctuation. The mean of LEV (0.662), indicating that the research sample is highly leveraged. The mean of 0.213 for ROA show the return of 21.3 money unit for investment in 100 money unit assets. By analyzing the variation of the sample data, it appears that the variable distribution is normal (Xu et al., 2013).

Table 2 shows the correlation matrix, which represent a positive relationship between managerial overconfidence (Overconfidence) and managerial accounting competence (Acc_Competence). In addition, the values of variables’ correlation are not high, showing that multicollinearity is not a main issue of research. The table coefficients also have expected signs.

Whit respect to informative of descriptive statistics and correlation analysis, more conclusive result and evidence can be used through multivariate regression analysis that controls for special firm-specific variables (Bhuiyan and Hooks, 2019). Table 3 represents the results of our main regression where equation (1) estimated to examine the association between manager overconfidence (Overconfidence), managerial accounting competence (Acc_Competence) and dividend smoothing (AD_DIV) with respect to control variables consist of fixed effects of year and industry. In Model (1) we can see that the relation between Overconfidence and AD_DIV is positive after controlling for other variables. Also, the result shows the negative relation between Acc_Competence and AD_DIV. The positive relationship between Overconfidence and AD_DIV is statistically significant with strong coefficient. The negative relationship between Acc_Competence and AD_DIV is confirm statistically, too. So, the significant of Overconfidence and Acc_Competence in Table 3 provide high support to research hypothesis. About the other variables, Size, ROA, Loss and b_ind are positively associated while Inst, Mgt, Cfo_Std, Ret_Std, MTB, Lev, and B_Size, are negatively linked with dividend smoothing. The results of research control variables are linked to dividend smoothing studies (e.g. Fliers, 2019; Aier et al., 2005; Chen et al., 2022).

Finally, the results are in line with executives' accounting competence (Albrecht et al., 2018; Kalelkar and Khan, 2016; Matsunaga et al., 2013) and manager overconfidence aspects (Zhou et al., 2020; Cohen and Zarowin, 2010) which show the fact that the accounting background of managers and their overconfident characteristic, increase the tendency to dividend smoothing. In particular, they focus on behavioral proxies and experimental background of managers as these criteria are some of the most basic factors for the investing and reporting. Empirically, research findings develop the literature on the reporting consequences of managerial overconfidence and manager accounting competence from Iran context.

In this step, we examine whether information quality change the impact of Overconfidence, Acc_Competence on AD_DIV. So, we cut our sample in two groups consist of high and low information quality firms and estimate equation (1) with these new subsamples. The results in Table 4 show that Overconfidence and Acc_Competence are positively linked with dividend smoothing in low information quality sub-sample. As a matter of fact, this positive and significant association is only for the low information quality subsample. These results are in line with our expectations that information environment with ambiguity and uncertainty lead to reduce the investor’s sentiment to analyze and predict cause to increase in the decision-making error and create higher activity risk. Therefore, in low information quality environments, firms change the dividend payment trends (dividend smoothing) in order to reduce investment risk and maintain the attractiveness of stock (Du et al., 2010; Leary and Michaely, 2011). Accounting information allows investors to assess the firm value and its inherent risks. Theories show that when the environment information quality is higher, the amount of exploitation of private information by informed traders is lower because the information quality leads to reflection of share news in prices. So, it reduces the need to smooth dividends to protect prices (Easley and O’Hara, 2004). If there is a high-quality information system, the communication of high-quality information to the market makes it difficult for the smoothing of dividends. Trying to obtain information through unofficial available channels of capital market leads to increase the risk of unreal information, and excess expected return will be limited because of acquisition costs of information. Based on information cost literature, the quality cause to information asymmetry decrease and reduces the dividend smoothing (Dutta and Nezlobin, 2017). In summary, our results show that the relationship between overconfidence and accounting competence with dividend smoothing depends on the information quality of firms.

Based on the previous analysis, we measure manager overconfidence (Overconfidence) from capital expenditures proxy and accounting competence (Acc_Competence) from previous audit experience as a partner or manager in an audit firm, respectively. As a robustness test, we measure manager overconfidence (Overconfidence) from another way based on Rashidi (2020) that the overconfidence takes the value of “1” if predicted earning is more than real earning for 3 years, otherwise zero. Accounting competence is a dummy variable that takes the value of “1” if manager had previous audit experience as a partner or manager in a big audit firm, otherwise zero (Rashidi et al., 2024), and estimate main equation to review if main result of research is against using other proxies of the independent variables. We show the results of this different proxies in Table 5. Model 1 shows the results of main regression using alternate proxy of manager overconfidence (Overconfidence), it shows that manager overconfidence is positively associated with dividend smoothing (AD_DIV). Model 2 shows the results of main regression using another proxy of accounting competence (Acc_Competence), it shows that accounting competence is negatively associated with the dividend smoothing (AD_DIV). Finally, these outcomes show that main results of research are not sensitive to the use of another proxies of manager overconfidence and accounting competence.

The research findings show that the hypothesis analysis could be subject to the endogeneity bias. So, based on current studies (Gull et al., 2020), we estimate instrumental variable (IV) method and propensity score matching (PSM) approach to review endogeneity of models.

Propensity-score matching approach creates sets of groups consist of control and treatment. A matched set include at least one participant in the treatment and control group with similar propensity scores (Lunceford and Davidian, 2004). The goal of this approach is to approximate a random experiment, decrease many of data analysis problems (Hesarzadeh, 2020).

It allows to ensure that the association between dividend smoothing and independent variables consist of overconfidence and accounting competence is not originating from the firm specific control variables, to be more precise, the sample do not self-select to control the result and improve the dividend smoothing. Clearly, we create a dummy variable equals “1” while the manager overconfidence is greater than the median of sample, zero otherwise. The results of the first part in Table 6 show that the subsamples significantly differ based on control variables. So, PSM let us to have a subsample consist of firms with similar characteristics in terms of control variables but different based on manager overconfidence. In the next stage, we estimate main equation using PSM sample to examine if main results of research are effected by control variables. In Table 7, the results of Model 1 in the form of second stage PSM outcomes are the same as results under basic analysis in Table 3. In other words, verifying that research results are not affected by the difference in control variables.

At last, endogeneity need to test because one of research independent measures (Overconfidence) is based on earning that linked to dividend smoothing. IV approach is a way to solve the endogeneity issue. The main challenge of IV approach is to identify related variables that influence managerial overconfidence (independent variable) but are not associated with the dividend smoothing. Following prior studies and use the nature of properties and market share as valid instrumental variables, the IV estimation at first stage shows that instrumental variables are significant at 5% level. At the next stage, Table 7 shows that endogeneity test is significant, too and shows that chooses the instrumental variables are reliable. In relation with the link between manager overconfidence and dividend smoothing, the IV estimation result are same as main result in Table 3. So, research results are not effected by endogeneity issue.

If the managers expect the firm’s future performance to improve and cash resources to grow, they increase the dividend payment. In other words, in order to compensate for the current unfavorable conditions, relying on the ability to improve resources in the future, the firm is changing the dividend process. In this regard, Jagannathan et al. (2000) showed that if the ability of firms to create cash resources is limited, it is not possible to smooth dividends. Investors have different ability to process dividend information. Hence, changing the process of profit sharing can lead to creating unfavorable conditions for uninformed investors in information analysis and as a result increasing information asymmetry in financial markets (Baum et al., 2017). In light of dividend smoothing issues and change in firms’ expectation regarding the use of capable managers, we show a new approach for manager accounting competence in practices. Finally, this research makes some contributions to the manager competence literature.

At the first stage, the effect of manager overconfidence was review and that how it can change firms’ dividend. Based on prior literature, overconfident managers overestimate their ability to deal with incoming shocks and smooth dividends. This group of managers believe that they can control the relevant fluctuations in the long run and achieve their desired earning (Lambrecht and Myers, 2012). Overconfident managers have the ability to persuade the board to adopt aggressive approaches. In other words, overconfidence of managers leads to a decrease in cash retention in order to do aggressive projects, which results in a decrease in financial resources for distribution. In this regard, managers try to smooth the dividend in order to control the negative effects of their optimistic decisions (Yang and Kim, 2020). So, manager overconfidence can lead to enhances dividend smoothing.

Next, the result show that the manager accounting competence and dividend smoothing relation is significantly positive. This result is linked to related literature (e.g. Albrecht et al., 2018; Aier et al., 2005; Demerjian et al., 2012) and show that previous experience of managers (manager competence) in the field of auditing has led to the manager’s extensive knowledge of accounting standards and operational complexities, and the possibility of financial statements misrepresentation is reduced, and therefore, with respect to low possibility of financial misrepresentation, the conditions for dividend smoothing are reduced. Moreover, the results of subsamples show that research findings are not driven by accounting quality proxy. Finally, research result support hypothesis and are robust to the applicant of other measures of variable and tests for endogeneity.

The results have personal and public implications for regulator, stockholder and market practitioners. The findings cause to controlling and directing the ability of managers to make optimal decisions leads to an improve in the financial reporting quality and a reduction in the dividend smoothing. The current approach of accounting competence is based on the assumption that accounting competence has led to an efficiency of the management’s performance and reduce the self-interesting motives of manager. Eventually, however this research presents some contributions and findings, but we can see few limitations, which can motivate to future research. There are limitations, including the method of measuring the accounting competence of managers. Considering that the accounting competence of managers is one of the personal and internal characteristics of managers and cannot be directly observed and measured, the presented criteria have limitations.

Table A1