Factors Affecting Dividend Payment Decisions: The Case of Libya

This research paper will investigate the views of executives and top management of firms listed on the Libyan Stock Market [LSM] regarding the main factors affecting dividend payment decisions. A questionnaire has been used to gather primary data from a range of industries. The evidence reports that the most influential factors in making dividend payment decisions involve: earnings [level of current earnings], Patterns of past dividends, investment opportunities, and liquidity [availability of cash]. Furthermore, the viewpoints of the respondents from three industry sets show no difference in the importance of the dividend factors among industries groups, except the level of reserve. Therefore, the industry type issue appeared to influence the importance that respondents placed on determinants of dividend payment decisions. The findings also show that respondents strongly believe that dividend payment is an important issue in Libyan firms, and these results are consistent with signaling and Pecking order explanations for paying cash dividends. The originality of this paper lies in presenting new evidence to the body of knowledge regarding dividend decisions in a pre-emerging market; this is the first study to document the perceptions of managers of dividend-paying LSM-listed firms about the factors affecting dividend payments.


INTRODUCTION
Dividend decisions have been a contentious issue in financial literature for many decades, and remain the greatest dilemma for both academics and practitioners.
Ample empirical studies have attempted to highlight this issue from different perspectives and have followed various approaches to identify the fundamental factors that affect dividend decisions. According to , the main reason for this continual debate is the extensive research that has adopted the economic modeling approach describing the factors behind dividend decisions. Relying on this approach alone, however, to merely model market data while neglecting behavioral approaches which measure how managers behave when making such decisions cannot completely identify factors affecting dividend payment decisions. Chiang et al. [2006,p.78] deduced that the cardinal thrust of academic research must turn toward learning about motivation and on what perceptions this motivation is based.
Due to dividend behavior being "a cultural phenomenon" [Frankfurter and Wood,1997, p. 31], dividends policy in pre-emerging markets is significantly different from that in advanced markets[characterized by higher degrees of efficiency, competitiveness and liquidity, and by having greater transparency]. Libya is a pre-emerging economy in which the stock market has only recently been established and most of its listed firms are in state-ownership. Added to that, some firms were only privatized in the last few years, and these facts will undoubtedly influence respondents' views.
In light of the limited number of studies on dividend policy in developing economies in general, and the issue of dividends from the recently established Libyan Stock Market [LSM] in particular, this paper has three aims. The primary aim is to identify the most influential factors The total number of investors registered in the market reached around 250,000 investors and in 2010 the average daily trading value was 304,005 LD [according to Libyan Central bank 1$ = 1.24 LD] with about 7,465 transactions. There were12 listed brokerage companies in 2008,rising to 25 in 2010. However, from 22 nd February, 2011,five days after the start of the Libyan Civil War, the market closed for one full year, reopening on 22 nd February 2012. This closure led to the delay in the performance of the market [op.cit.,2012, p. 46].

LITERATURE REVIEW
The corporate financial literature contains voluminous research studies, including surveys dating back more than half century. It presents a large number of theories, hypotheses, and explanations for the dividends policy. In this section, two main topics will be discussed. First, explanations for paying cash dividends. Second, the main survey studies on dividends policy.

explanations for paying cash dividends
Many of the dividend theories that are commonly discussed today were originally proposed in literature in an attempt to explain the reasons for paying dividends. Below are eight broad categories, which are not necessarily mutually exclusive, for paying cash dividends: 1-Bird in hand theory: the theory claims that due to the uncertainty associated with the business environment, investors prefer current dividend payments as a bird in hand rather than capital gains as a sparrow in the tree because of lower risk [Gordon, 1959[Gordon, , 1963Bhattacharya, 1979]. 2-Information signalling theory: the theory argues that there is a gap between managers and investors, this gap can be reduced via dividend payments, therefore, investors consider an increase in dividend payments as a positive signal for a firm's future profits and its promising performance [Bhattacharya, 1979[Bhattacharya, , 1980John and Williams, 1985]. 3-Agency cost theory: the theory proposes that due to separation between ownership and management, the dividend payment is a useful way for lowering the agency cost of free cash flow [ Jensen and Meckling, 1976;Easterbrook, 1984;Jensen, 1986]. 4-Transaction cost theory: this theory posits that firms have limited internal fund resources to finance their investment needs, therefore, the firms pay low or nondividends to avoid the transaction cost associated with dividend payments and external financing [Rozeff, 1982]. 5-Tax-related effect theory: the supporters of this theory believe that dividend payments are taxed more than capital gains, thus, managers of firms should retain earnings [no dividend payments] and transfer them to capital gains [Elton and Gruber, 1970;Scholes, 1978, 1982]. 6-Pecking order theory: the theory claims that the priority to financing investment opportunities must be via internal cash flow, then debt and finally equity issuance. Therefore, firms should adopt a low dividend policy [Myers,1984;Myers and Majluf,1984]. 7-Life cycle theory: the theory assumes that a young firm in its high growth phase has many positive investment options but limited cash flow, therefore no dividend payments. While mature firms have a large amount of free cash flow but poor of investment opportunities, as a result, paying high dividends ratio [Fama and French, 2001;DeAngelo et al., 2006]. 8-Catering theory: this theory argues that investors' decision about dividend is determined by their sentiment which inconstant over the time, thus, managers of firms cater to investors what they currently require [Baker and Wurgler, 2004].
Survey Studies on dividends financial scholars have attempted to identify the main factors affecting dividend payment decisions employing survey approach. For instance, Lintner [1956] is best known for using a behavioral approach to ascertain the views of managers regarding determinants of dividend policy. He found that the most important determinants of dividends are the stability of earnings in the case of young firms, the long-term target payout of dividends, and current earnings. Baker and Powell [2000] investigated the views of managers of NYSE-listed US firms about factors determining dividend policy. Their study updates and extends previous survey research on dividends policy and provides a longitudinal comparison for the years1983-1997. They conclude that the most important factors include the level of current earnings, the level of expected earnings, and the pattern of past dividends. The key factors remain stable over the time period; however, industry type appeared to influence the importance that respondents placed on some determinants of dividend policy. Baker et al. [2001] examined the perception of managers about twenty-two factors [based on results from a 1999 survey] that influence dividends policy of financial and non-financial firms listed on NASDAQ. Their findings showed that the most important factors are the pattern of past dividends, the stability of earnings, and the level of current and expected future earnings. The study also identified significant differences between managers of financial firms and non-financial firms attached to nine of 22 factors in terms of importance. Frankfurter et al. [2004] carried out research in five countries on three continents to discover the perception of chief financial managers about dividends payout. They concluded that culture dimensions play a vital role in the perception of dividends policy. Brav et al. [2005] employed surveys and in-depth interviews with the executives of 384 public and private firms in the US to identify factors that affect dividend payouts and repurchase of shares. Their findings suggest that the target payout is no longer a preeminent decision variable and that managers express a strong desire to avoid cutting dividends and only consider increasing the level of dividends after investment and liquidity needs, and the requirements for stable future earnings are met. Bancel et al. [2005] surveyed the managers of 1131 firms in 16 European countries to find out determinants of dividend payouts in order to compare their views with those of US managers. The results showed that smooth dividends and the difficulty in cutting dividends are the most important factors. The study also revealed that although the European and American managers followed different targets, their views were similar.

THEORETICAL FRAMEWORK AND HYPOTHESES DEVELOPMENT
Based on the literature review and the researcher's knowledge of the Libyan economic environment in respect to the recent development of the LSM, the lack of public awareness of dividends, and the lack of adequate financial disclosure, the researcher speculates that the following factors play a vital role in dividend payment decisions of Libyan listed firms.

Earnings [Profitability]
There is almost unanimous agreement in the financial literature that profitability of firms is one of the most significant factors impacting on dividend payment decisions [Baker and Powell,2000 ;French, 2000,2002;Baker et al., 2006;Denis and Osobov, 2008;El Sady et al., 2012;Hossain et al., 2014].Since the amount of dividend is often distributed from annual income, it can be argued that more profitable firms have a tendency to increase their dividend payments. Bhattacharya [1979], Miller and Rock [1985] Mullins [1983, 1986] and Baker, Mendel and Wurgler [2015] explained higher dividend payment as a strong signal of future cash flow; because, management of firms [insiders] are likely to have more private knowledge about current situations and bright future prospects than investors [outsiders],managers use dividends as a device to convey useful information to the financial market about future profitability and growth, which in turn reduces asymmetric information [John and Williams, 1985].However, the question emerges: are dividend decisions motivated by current, future or stability of earnings? Several studies conducted in advanced and emerging markets reported that the largest factor influencing dividend decisions is the level of current earning [Linter,1956;Baker and Powell,2000;Baker et al., 2007;John, 2013;Alzomaia& Al-Khadhiri, 2013]. These studies claim that cash flow [current earnings] reflects the ability of firms to pay dividends.
The expected future earning is also, categorized as a key factor that affect dividends decision [Baker & Powell, 2000;Baker et al., 2001;Brav et al., 2005;Baker et al.,2007;Baker &Jabbouri,2016].The importance of this factor stems from the stability of future cash flow. Babiak and Fama [1968] found that managers have an aversion to increasing dividends which cannot be compensated by future earnings.
As such, the stability of earnings is documented as the main factor in deciding on dividends payment [Pourheydari, 2009;Archold and Vieira, 2010;Baker and Powell,2012;Baker and Kaboor,2015;Baker and Jabbouri, 2016].This suggests that managers recognize the importance of maintaining the magnitude of cash dividend from reducing in the future. The higher the volatility of earnings, the greater the likelihood is that a firm will experience a decrease in earnings which could result in a forced decline in the cash payouts [Baker et al., 2001]. Accordingly, the hypotheses are formulated as follows : Hypothesis 1: Ceteris paribus, Earnings significantly affect dividend payment decisions in Libya: Hypothesis 1A: Ceteris paribus, Level of current earnings significantly affects dividend payment decisions in Libya.

Pattern or Continuity of Past Dividends
Patterns of past dividends play a vital role in setting dividends policy; firms tend to keep cash dividends at least at the previous year's level. The earliest study was undertaken by Linter [1956] who suggested that firms' payout patterns depend on the dividends of the previous period. He argued that managers believe investors prefer firms which follow a steady dividends payout ratio, thus firms avoid changing dividend rates and try to maintain a constant level of dividend payouts. Brittan [1966] applied Linter's model and confirmed that firms follow a Constant Payout Ratio Policy. Similarly, Shevlin [1982] examined Linter's dividends model on a sample of Canadian and Australian firms, and the results suggested that dividend decisions depend on past patterns of dividends. Many other studies assert Linter's findings that past dividend patterns influence dividend payment decisions [Baker et al., 1985;Pruitt and Gitman, 1991;Benartzi et al., 1997;Baker et al.,1999;Benzinho, 2004;Bancel et al., 2005;Al-Twaijry,2007;Mizuno,2007;Szomko, 2015].
Conversely, Glen et al. [1995] argued that the behavior of dividends policy in developing countries is different from that in developed countries. They contend that dividend decisions in emerging markets are unrelated to past patterns of dividend payouts. Frankfurter and Wood [1997] deduced that patterns of firms' dividend payouts are a cultural phenomenon. More recently, Khan [2011] revealed that Pakistani managers do not take into account past patterns of payouts when setting dividends policy. Rahaman [2015] investigated the perception of 124 chief financial officers of firms listed on the Dhaka Stock Exchange [DSE] of Bangladesh. He found that target payout ratio follows current earnings and patterns of dividends changes over a firm's life cycle. ]. Accordingly, the hypothesis is formulated as follows : Hypothesis 2: Ceteris paribus, Pattern of past dividends significantly affects dividend payment decisions in Libya.

Investment Opportunities
Since early research into the dividends phenomena, firms' financing needs for investment intentions have been addressed in many studies as a leading determinant of dividends policy [Miller and Modigliani,1961;Higgers,1972;Macabe, 1979]. A review of financial literature discloses several explanations in regard to the correlation between investment opportunities and dividend payouts. For instance, the residual theory suggests that after all profitable investment opportunities have been undertaken, the surplus amount distributes as dividends. In this context, firms will totally depend on internally generated funds for growth opportunities; thus, dividend payouts will be low and often zero. Miller and Modigliani [1961] confirm that firms pay cash dividends after exhausting all profitable investments. More recently, focusing on emerging markets, Baker and Jabbouri [2016] surveyed Moroccan managers about setting dividends policy and found that 54% of respondents believed firms should pay dividends only after financing future investment opportunities and those managers generally support a residual theory.
Another explanation can be seen in life cycle theory which assumes that firms in advanced stages tend to pay higher dividend payments, while firms in the early phases of development retain earnings for new investments [Grullon et. al., 2002;DeAngelo and DeAngelo,2006].The logical notion of this assumption is that young firms have many positive investment opportunities with limited funding resources; thus, firms are forced to hold onto earnings rather than pay dividends. Fama and French [2001] revealed that firms with low growth options tend to pay higher dividend ratios, while high growth firms resort to keeping profits for new projects. Other studies documented that investment opportunities affect dividend decision [Chang and Rhee,1990;Smith and Watts, 1992;Holder et al., 1998]. Further, the pecking order theory, presented by Myers [1984] and Myers and Majluf [1984] claims that firms fund their projects and expand according to a particular order: first, internal funds [retained earnings]; second, external financing [debt] short-run over long run; and finally, equity issuance. In this respect, rapidly growing firms finance future profitable investments via external funds [if financing requirements for growth opportunities exceed internal resources]; thus, firms are more likely to pay low or no dividends. Therefore, the hypothesis is formulated as follows : Hypothesis 3: Ceteris paribus, Investment opportunities significantly affect dividend payment decisions in Libya.

Current Degree of Financial Leverage
Financial leverage has been considered as a primary factor that affects dividend decisions [Crutchley and Hansen,1989;Jensen et al., 1992;Gugler and Yurtoglu, 2003;Aivazian et al., 2003;Al Malkawi,2005;Al Twajiry, 2007;Al Kuwari, 2009]. Leverage is defined by Ward and Price [2006] as the percentage of capital that is financed by external funds [debt] instead of equity.
Numerous explanations can be found in financial literature about the main reasons for making firms pay low dividends. For instance, firms with high financial leverage tend to decrease dividend payouts due to constraints of debt-suppliers which impose rules upon the firms to maintain internal cash to cover financial obligations toward debt providers [Agrawal & Jayaraman, 1994;Manos, 2003;Naser et al., 2004;Rahaman, 2015]. In other words, highly leveraged firms have fixed financial commitments [interest payments] stemming from external debt. In such cases, firms will resort to retain internal cash resources to meet these liabilities towards creditors, and failure to meet these obligations may result in bankruptcy. Rozeff [1982] revealed that firms with a high level of debt tend to pay low dividends to avoid the transaction costs associated with external financing.
Other literature suggests that debt can be used as a device to reduce agency costs of free cash flow. Jensen [1989], Jensen & Meckling [1976], and Myers [1977],among others, argue that debt is another mechanism to minimize the agency problem of free cash flow; that is, when a firm borrows money, it will bear a large burden of fixed costs towards the lender, which will put management under control and monitoring in order to meet these commitments, and thus avoid keeping large free cash flow in the firms and, as a result, mitigate agency costs [risk of waste free cash flow] and maintain lenders' confidence. Hence, the hypothesis is formulated as follows : Hypothesis 4: Ceteris paribus, Level of current financial leverage significantly affects dividend payment decisions in Libya.

Liquidity Constraints
Liquidity has been described in financial literature as a major factor influencing dividends policy [Alli et al.,1993;Ho,2003;Kanwal and Kapoor, 2008;Kumar and Waheed, 2015]. The reason for the importance of this factor is that a strong liquidity position indicates [signals] the extent to which a firm is able to meet its financial obligations. Firms with higher cash availability tend to distribute higher cash dividend ratios as compared to firms with a poor liquidity. Baker, Mukherjee and Paskelian [2005] interpreted the importance of liquidity in affecting dividend payment decisions, suggesting that paying cash dividends depends on cash availability, not on earnings based on accrual accounting. As a result, the availability of cash has a crucial impact on the ability of firms to pay cash dividends. Forti et al. [2015] concluded that a large liquidity position correlates with larger firms' propensity to pay dividends to shareholders. Therefore, , the hypothesis is formulated as follows : Hypothesis 5: Ceteris paribus, Liquidity significantly affects dividend payment decisions in Libya.

Legal Rules and Constraints
Legal restrictions such as capital requirements influence the amount that firms payout as dividends to shareholders [Mayne, 1980;Boldin and Leggett, 1995;Baker et al., 2001;Baker et al.,2006]. Cash dividend payments should not be greater than the total of retained profits plus net income for the current fiscal year. This so-called capital impairment rule prohibits the distribution of annual profits in cases of impairment of capital [Ballantine & Hills, 1935, p. 246].
Laws and regulations affect the process of setting dividend payout policy. For instance, the Basel Committee on Banking Supervision [BCBS, 2011] points to the increasing regulatory relevance of banks' dividend policy [Brunnermeier et al., 2009]. Mehta et al. [2014] conducted research on determinants of banks' dividend policies in Pakistan. They revealed that the level of reserves has a major impact on dividend payments. Hsia and Tseng [2016] found that capital requirement regulation has a significant impact on banks' dividend payouts. Accordingly, , the hypotheses are formulated as follows : Hypothesis 6: Ceteris paribus, Legal rules and constraints significantly affect dividend payment decisions in Libya:

Hypothesis 6A: Level of reserves significantly affects dividend payment decisions in Libya.
Hypothesis 6B: Impairment of capital significantly affects dividend payment decisions in Libya.

METHODOLOGY AND DATA SAMPLE
This study adopted the quantitative approach, and employs a survey as the primary means of collecting data, similar to that used in many previous studies [Baker and Powell 2001;Brav et al. 2005;Baker et al. and 2016. The survey [questionnaire] is designed with two pages which translate text from English to Arabic [a copy of the Arabic version is available from the author] and consists of three sections: the first section contains four questions that provide a profile of respondents. The second section concerns specific information to identify whether dividend payment decisions are important in Libyan firms and if the respondents took part in dividend payment decisions during their career. The third section contains 11 closedend statements to obtain information about firm dividend policy. The section asks respondents to indicate the importance of 11 factors that each firm used in determining its dividend policy.
These main factors were derived from literature and previous empirical studies on dividends. A five-level Likert scale was used :[1= very un-important, 2= un-important, 3= neither important nor unimportant, 4= important and 5= very important]. The questionnaire also contains open questions which give respondents a chance to provide information about any other factors which might affect their dividend payment decisions. The questionnaire was piloted by academic colleagues and their feedback was used to make modifications before the final version of the questionnaire was given to the survey group.
Using the Libyan Stock Market [LSM] web page, the researcher selected listed firms which met one criterion: having paid cash dividends in two years during [2009][2010][2011][2012][2013][2014][2015].To confirm this criterion, financial statements of the firms were viewed. The sample contained 36 firms 1 from three industries [banking, insurance, and manufacturing].The questionnaires were distributed to the board of directors, chief executives and chief accountants of each firm 2 . Each questionnaire was presented with a cover letter which informed potential respondents that the result would be anonymous and published in summary. The 93 questionnaires were distributed in mid-April, 2016. By the end of July 2016, 80 responses had been received, but only 77 of these were usable [83% responses rate]. Summary information is presented in

STATISTICS TESTS
Several statistical tests using SPSS software were conducted in order to analyze data collected from the questionnaires to achieve the aim of the study. The descriptive statistics used to show whether respondents agree with the importance of the dividend payment issue in Libyan firms, and if respondents took part in dividend decisions during their career as well as the factors that influence the dividend policy, considered by managers in setting their firm's dividend policy. The researcher used the sign tests to determine the influence of each factor on the dividend policy of the firms paying dividends. Chi-square [ * -] analysis was used to test for the differences in responses among three industry groups, regarding 11 factors influencing dividend policy.

LIMITATIONS
As with any survey, some potential limitations of this study should be noted. First, the study concentrated only on the main factors [depending on results from prior empirical studies] affecting dividend payment decisions because these factors are familiar to decision makers of Libyan firms: the recent establishment of Libyan Stock Market; the lack of public awareness of the issuing of dividends; inadequate disclosure; and absence of effective financial intermediaries. Second, the researcher restricted the sample to three industries represented on the LSM. Third, the survey methodology measured beliefs and not necessarily actions. Finally, the researcher limited the length of the survey to increase the responses. Therefore, caution should be taken in generalizing the results.

SURVEY FINDINGS AND DISCUSSION
This section provides respondents' views concerning the importance of dividend payment decisions, the factors [in the Libyan context] affecting dividend payment decisions, the most important factors affecting dividend decisions, and industry influence on dividend decisions. This information will be followed by a discussion of the results.

Specific Information about Dividends
This section highlights whether respondents agree with the importance of the dividend payment issue in Libyan firms, and if respondents took part in dividend decisions during their career. As shown in table [2], 81% of respondents agree with the importance of the dividend payment issue, while only 19% do not agree. This result indicates that a high proportion of respondents are aware of the importance of dividend payments despite the fact that the dividend payment issue is a new development in Libya. Add to that, around 87% of respondents have been involved in dividend payment decisions. The Summary information is presented in table [2].

Factors Affecting Dividend Payment Decisions
This section focus on two areas: which factors are affecting dividend payment decisions and Which the most important factors affecting dividend decision. Table 2 the descriptive statistics showing the factors that affect dividend payment decisions, considered by managers when setting their firm's dividend policy. The sign tests has been used to determine the effect of each factor on the dividend payment decisions.. In total, factors were statistically significant at 0.001 levels [H1-H1a, H1c-H2, H4-H5, H6, H6a, and H6b] and one factor at the 0.05 level [H3]. The factors that were rejected were: H1b [Level of expected future earnings affects dividend payment decisions of firms]. Table 2 also presents the significance level of each of 11 factors. The following discussion focuses only on the five most important factors affecting the dividend policy. In total, 10 factors were important in affecting dividend policy. The factors were clearly ranked and the highest one was for earnings, specifically "level of current earnings" affects payment dividend decisions [H1a], which reflects the strong relationship between cash flow and dividend payment decisions. Notes: ** and * Significant at the 0.001 and 0.05 levels, respectively. The descriptive statistics show the level of importance of various factors considered by managers in setting their firm's dividend policy. The sign test are used to determine the significance of the each factor. The next important factor was Patterns of past dividends affects dividend payment decisions of firms [H2]. The importance of this factor may mean that a firm's past dividend decisions may restrict its current decisions. Baker et al. [1985] found a strong relationship between past and current dividend decisions.

Depicted in
The fourth important factor in setting the firm's dividend policy was investment opportunities affect dividend payment decisions of firms [H3]. Baker and Powell [2000] found a link between investment opportunities and dividend payment decisions.
The fifth most highly-ranked factor influencing dividend policy was liquidity or "availability of cash" [H5]. The findings indicate that firms with higher liquidity pay dividends which convey a positive signal to outsiders that they are able to readily pay its liabilities; this is consistent with the signaling theory explanation for paying dividends.

Industry Affecting Dividend Payout
This section discusses whether factors affecting dividend payment decisions differ by industry type.
In order to determine the differences in responses among three industry groups, regarding 11 factors affecting dividend payment decisions *test was employed. To perform the test and avoid inadequate cell size, the researcher collapsed five interval importance scales in two classes. The first class contains very unimportant, unimportant, and neither important nor unimportant; and the second class included important, and very important. The findings suggest that there are no significant differences in the importance of the dividend factors among the three groups, except for the level of reserves factor [H6a).
Further *test was performed using pair-wise comparisons between the industry groups on all 11 factors. The findings show that the manufacturing and insurance firms had no significant differences in responses at the .05 level for 11 questions with adequate cell sizes. Hence, the differences occurred primarily as a result of the banks' responses relative to either manufacturing or insurance firms.
The reported differences between the banks and the other firms may be due to regulation. Banks, for instance, pay careful attention to "reserves for non-performing loans" because, by law and regulations imposed by the central bank of Libya, they must meet specific [level of reserve] requirements for absorbing any future loan losses. Thus, the executives and top management of banking firms agree about the impact of the level of reserves on dividend payment decisions, whereas the executives and top management of manufacturing firms and insurance firms disagree about the impact of the level of reserves on dividend payment decisions. The Summary information is presented in table [4]. Notes: * Significant at the 0.05 level, respectively CONCLUSIONS This is the first study to investigate the views of respondents from LSM-listed firms about the main factors affecting dividend payment decisions. The results of this survey led to several conclusions. First, the vast majority of respondents believe that dividend payment decisions are an important issue. Second, the most important factors influencing dividend decisions include earnings [level of current earning], Patterns of past dividends, investment opportunities, and liquidity [availability of cash]. Third, the findings indicate no difference in the importance of the dividend factors among industries groups, except the level of reserve. Therefore, the industry type issue appeared to influence the importance that respondents placed on determinants of dividend payment decisions.

Notes:
1-Many research studies conducted in developed economies mainly the US markets, are analyzed based on large sample and cover long period.