Lup.lub.lu.se



Determinants of dividend policyThe effect of the global financial crisis on Swedish firmsbyHenrik Svensson & Victor ThorénMay 2015Master’s Programme in Corporate and Financial ManagementAbstractDividend is a much-debated subject and has been compared to a puzzle with no sole solution to explain why firms pay dividend. Many academics have tried to determine what company variables determines a firm’s dividend policy. These studies have provided different explanatory variables to explain why firms pay dividends. Therefore, the main purpose of this study was to fill in a piece of the dividend puzzle by investigating the determinants of dividend policy for Swedish firms. The global financial crisis in 2008 caused the Swedish economy experienced the fastest deceleration in demand and production since World War II. This challenging event forced many firms to take considerable action to avoid the negative effect that followed. The lack of studies regarding the impact of the financial crisis made us want to test whether Swedish firms experienced any significant changes in the determinants of dividend policy.Using a quantitative and deductive approach, this study used the Irrelevance, Bird-in-hand, Signalling, Agency, Catering, Life Cycle and Pecking Order theory to examine the determinants. The sample was drawn from the NASDAQ OMX Stockholm Large cap list. There were three time periods employed, pre & post the financial crisis as well as one whole period. By using a random effects model we tested the relationship between the selected company determinants and the dividend payout ratio for the three time periods.The study found that profitability, size, risk, retained earnings, and firm value can be seen as determinants of dividend policy for Swedish firms since they are significant. This implies that Life cycle, Agency, Bird-in-hand, Catering and Pecking order theory is applicable to the Swedish firms. The study does not find any supporting evidence for Signalling or the Irrelevance theory. The results showed that there were no statistically significant changes in the determinants between the pre & post period, except for a minimal difference in firm value. This implies that the financial crisis did not affect Swedish firms dividend policy. Acknowledgements We would like to thank out supervisor Naciye Seckerci for her support throughout the research process and Lund School of Economics and Management for providing the necessary means for conducting the research. Table of content TOC \o "1-3" 1.0 Introduction PAGEREF _Toc294267921 \h 61.1 Purpose and research question PAGEREF _Toc294267922 \h 81.1.1 Hypothesis PAGEREF _Toc294267923 \h 81.2 Research limitations PAGEREF _Toc294267924 \h 101.3 Thesis outline PAGEREF _Toc294267925 \h 102.0 Literature review PAGEREF _Toc294267926 \h 112.1 Dividends PAGEREF _Toc294267927 \h 112.2 Irrelevance theory PAGEREF _Toc294267928 \h 112.3 Bird-in-hand theory PAGEREF _Toc294267929 \h 122.4 Signalling theory PAGEREF _Toc294267930 \h 122.5 Agency theory PAGEREF _Toc294267931 \h 132.6 Catering theory PAGEREF _Toc294267932 \h 142.7 Life cycle theory PAGEREF _Toc294267933 \h 142.8 Measures of dividend PAGEREF _Toc294267934 \h 152.9 Previous studies of determinants PAGEREF _Toc294267935 \h 162.10 Selected dividend policy determinants PAGEREF _Toc294267936 \h 202.10.1 Growth opportunities PAGEREF _Toc294267937 \h 202.10.2 Profitability PAGEREF _Toc294267938 \h 212.10.3 Cash flow PAGEREF _Toc294267939 \h 212.10.4 Size PAGEREF _Toc294267940 \h 222.10.5 Risk PAGEREF _Toc294267941 \h 222.10.6 Retained earnings PAGEREF _Toc294267942 \h 232.10.7 Firm value PAGEREF _Toc294267943 \h 233.0 Methodology PAGEREF _Toc294267944 \h 243.1 Research method choices PAGEREF _Toc294267945 \h 243.2 The data PAGEREF _Toc294267946 \h 243.2.1 Sample and sample size PAGEREF _Toc294267947 \h 243.2.2 Data collection & time period PAGEREF _Toc294267948 \h 253.2.3 Variables definition PAGEREF _Toc294267949 \h 253.3 Regression model PAGEREF _Toc294267950 \h 273.3.1 Average values of the errors is zero PAGEREF _Toc294267951 \h 273.3.2 Variance of the errors is constant PAGEREF _Toc294267952 \h 273.3.3 Covariance between the error terms is zero PAGEREF _Toc294267953 \h 273.3.4 Independent variables are non-stochastic PAGEREF _Toc294267954 \h 273.3.5 Disturbances are normally distributed PAGEREF _Toc294267955 \h 283.3.6 No multicollinearity PAGEREF _Toc294267956 \h 283.3.7 Fixed & Random effects PAGEREF _Toc294267957 \h 283.3.8 T-test PAGEREF _Toc294267958 \h 293.3.9 Z-score PAGEREF _Toc294267959 \h 293.3.10 R-squared PAGEREF _Toc294267960 \h 303.4 Validity & Reliability PAGEREF _Toc294267961 \h 303.4.1 Validity of the independent variables PAGEREF _Toc294267962 \h 303.4.2 Reliability PAGEREF _Toc294267963 \h 304.0 Empirical findings PAGEREF _Toc294267964 \h 324.1 Regressions PAGEREF _Toc294267965 \h 324.1.1 Whole period regression (2003-2013) PAGEREF _Toc294267966 \h 324.1.2 Pre-crisis period regression (2003-2007) PAGEREF _Toc294267967 \h 334.1.3 Post-crisis period regression (2009-2013) PAGEREF _Toc294267968 \h 344.1.4 Z-score PAGEREF _Toc294267969 \h 355.0 Analysis and discussion PAGEREF _Toc294267970 \h 365.1 Determinants of the dividend policy PAGEREF _Toc294267971 \h 365.1.1 Growth PAGEREF _Toc294267972 \h 365.1.2 Profitability PAGEREF _Toc294267973 \h 375.1.3 Cash flow PAGEREF _Toc294267974 \h 375.1.4 Size PAGEREF _Toc294267975 \h 385.1.5 Risk PAGEREF _Toc294267976 \h 385.1.6 Retained earnings PAGEREF _Toc294267977 \h 395.1.7 Firm value PAGEREF _Toc294267978 \h 395.1.8 What are the determinants of the dividend policy in Swedish firms? PAGEREF _Toc294267979 \h 405.2 Were the determinants of dividend policy amongst Swedish firms affected by the financial crises? PAGEREF _Toc294267980 \h 416.0 Conclusion PAGEREF _Toc294267981 \h 436.1 Summary PAGEREF _Toc294267982 \h 436.2 Contribution PAGEREF _Toc294267983 \h 446.3 Further research PAGEREF _Toc294267984 \h 447.0 References PAGEREF _Toc294267985 \h 46List of tables TOC \t "Figurf?rteckning" \c Table 2.1 List of previous studies PAGEREF _Toc294267986 \h 17Table 3.1 Sample selection PAGEREF _Toc294267987 \h 24Table 4.1 Regression results for whole period (2003-2013) PAGEREF _Toc294267988 \h 32Table 4.2 Regression results for pre-crisis period (2003-2007) PAGEREF _Toc294267989 \h 33Table 4.3 Regression results for post-crisis period (2009-2013) PAGEREF _Toc294267990 \h 34Table 4.4 Z-scores for the independent variables PAGEREF _Toc294267991 \h 35Table 5.1 Determinants of dividend policy PAGEREF _Toc294267992 \h 40Table 5.2 Difference in coefficients (pre and post crisis period) PAGEREF _Toc294267993 \h 41 TOC \c "Tabell" 1.0 Introduction “The harder we look at the dividend picture, the more it seems like a puzzle, with pieces that just don't fit together.” - Fischer Black, 1976Dividend is and has been a debated subject among academics for many years. Does it create value or is it totally irrelevant? And why do firms pay dividend? The Irrelevance theory, a groundbreaking theory first introduced in 1961 by Miller & Modigliani (M&M) states that dividend has no affect on the value of shares. This has since then been much debated. Opposing M&M, Lintner in 1956 mentions the Bird-in-hand theory, which is contradictive to the Irrelevance theory and indicates that investors are inclined to pay a premium on dividend paying shares. It states that investors would rather have one bird in the hand, than two in the bush (Lintner, 1956). Jensen (1986) joined the debate by introducing the Agency theory, which tries to explain why firms pay dividend. Jensen looked at dividend from the agency cost perspective and examined the possibility of using dividend payments as a means of decreasing agency costs (Jensen, 1986). Dividend also has the benefit of conveying information of future earnings and financial stability, which brings us to Signalling theory. In 1956 Lintner first introduced this theory, connecting managers dividend payout ratio to the share price. It was further mentioned by M&M (1961). The theory was then discussed by Akerlof (1970) where he stated that information asymmetry (lemons problem) makes the firm want to signal they are not a lemon. Fischer Black (1976) further argued that dividend policy says something the managers don’t say explicitly. It signals that a raise in dividend is only feasible if the firm can continue paying dividend in the future and managers will only lower the dividend if the outlooks for a quick recovery seem dismal. A contemporary contribution to the dividend discussion is the Life cycle theory presented by DeAngelo & DeAngelo (2006), which states that the need to distribute free cash flow is the driver for the optimal dividend policy. The theory states that firms pay out different amounts depending on if they are a young firm with the need for internal financing or a more mature one where they can shift out a larger portion of retained earnings. This associates to the financial crisis, were firms were put under a lot of stress, mainly from the difficulty of financing themselves during the crisis. Therefore, the need for internal financing increased and in turn affected dividend. The global financial crisis has had a significant impact on the global business environment in recent years. The event is of high relevance as it impacted many firms financially and as a result firms had to take action in order to limit the negative effects. Few attempts have been made to investigate the impact that this significant event has had on the determinants of the firm’s dividend policy. The origins of the global financial crisis can be traced back to the US housing and mortgage market. The fall in prices combined with highly leveraged consumers, thanks to subprime loans, resulted in a dramatic increase in credit losses. This together with the difficulty in assessing the risk due to securitization of these loans lead to an interbank lending failure. When Lehman Brothers filed for bankruptcy, it showed clear evidence of the severity of the impact that these subprime loans and structure credit products had on banks. The financial unease therefore turned into an acute global financial crisis (Riksbanken, 2009, translated).The rapid spread of the financial crisis as a result, lead to sharp declines in the global economy, which resulted in many countries having to revise their GDP forecasts by 4 to 5 percentage points (Braunerhjelm, Djerf, Frisén, Ohlsson, 2009, translated). Sweden, which is a small open economy country with extensive foreign trade and a financial market that is well integrated with the rest of the world, was directly impacted by the global economic downturn (Riksbank, 2009, translated). The global economic downturn caused a large decline in the demand for Swedish exports goods, causing significant effects on employment and production in the country, as the portion of exports accounted for 40 percent of GDP in 2007 (Riksbank, 2009, translated). This caused Sweden to experience the fastest deceleration in demand and production since World War II (Braunerhjelm, Djerf, Frisén, Ohlsson, 2009, translated). In addition to the sharp drop in demand for Swedish exports, many firms had to reassess their investment policies as they found it increasingly difficult to finance these investments from banks. Since this had such a large effect, we feel it is relevant to investigate how an event of this magnitude can affect firm’s dividend policy. Furthermore, we can clearly state that despite all of the theories mentioned there is no sole solution to the dividend puzzle that was first discussed by Fisher Black (1976). The questions he asked; “why do firms pay dividends” and “why do investors pay attention to dividends” are still relevant today. 1.1 Purpose and research questionThe aim of our empirical research is to investigate the determinants of dividend payout policy and contribute to the on going debate why firms pay dividend. This will be accomplished by looking at a 11 year period, 2003-2013. In addition to this, we will try to clarify the effects made by the financial crisis on the determinants of dividend, which is an event that has received little attention. We contribute to the research surrounding the financial crisis by measuring the dividend policy determinants pre & post the global financial crisis of 2008. These years will be split in two 2 separate five year periods between 2003-2007 and 2009-2013. Our research contributes to the dividend policy research in three distinct ways. Firstly, we will seek to address the reasons why Swedish firms pay dividends, which to our understanding hasn’t been extensively researched. Furthermore, we will try to determine whether previous studies on dividend determinates can be applicable to the Swedish market.RQ1: What are the determinants of the dividend policy in Swedish firms?Secondly, we will try to answer whether these companies’ dividend policy was or wasn’t affected by the macroeconomic environment. This is achieved by examining how one of the most challenging crises in recent years affected the determinants of dividend policy. RQ2: Were the determinants of dividend policy amongst Swedish firms affected by the financial crises?Thirdly, we will introduce a new variable in the form of price to earnings to the regression and test it’s affect on dividend policy. 1.1.1 Hypothesis Based on our research questions and theoretical framework, we have composed the following hypotheses and sub-hypotheses against the outcome of the data analysis. H1: The determinants of dividend policyH1a: Growth will have a positive effect on dividend policyWe believe that growth will have a positive affect on dividend policy following the Signalling theory were firms use dividend to signal growth opportunities to investors. This would imply that a firm with high growth opportunities would increase their dividend payout ratio.H1b: Profitability will have a negative effect on dividend policyThis hypothesis is based on the assumption that firms with low profitability are in most need of signalling to the markets that future profits lie ahead.H1c: Cash flow will have a positive effect on dividend policyWe assume that higher cash flows lead to an increase in dividend payout ratio as a tool to prevent agency conflicts. H1d: Size will have a positive effect on dividend policyThe Life cycle theory suggests that mature firms are more likely to pay dividend due to their access to external financing. H1e: Risk will have a negative effect on dividend policyBased on the Pecking order theory, firms with higher risk will chose internal financing before external financing. This will lead to firms reducing the amount of dividend paid to investors.H1f: Retained Earnings will have a positive effect on dividend policyA mature firm that has been historically profitable should be able to distribute larger proportions of retained earnings. H1g: Firm value will have a positive effect on dividend policyAccording to Catering theory and Bird-in-hand theory investors are willing to pay a premium on shares that pay dividend. H2: There will be a change in determinants between the two periodsBased on the significant impact that the financial crisis had on the Swedish economy we assume that firms will take action that will affect the dividend policy, due to the considerable macroeconomic effects experienced.1.2 Research limitationsA closer look at the dependent variable shows that there are two ways of measuring dividend and thus the results may differ between the different measurements. This is something we are aware of and have therefore chosen the most appropriate formula based on previous research. Furthermore, investor preference is an aspect that this research cannot account for and will always be a factor that could contradict all relevant theory. Looking at the Swedish NASDAQ OMX Stockholm Large Cap list, there are a lot of financial firms, which we have chosen to remove from the sample. Thus, removing financial firms excludes an important industry for the Swedish market. The reasons for removing them is that the nature of the firms aren’t comparable to the same extent, however we are aware that the number of observations was affected negatively. Stock repurchase is a aspect that could have a explanatory factor, however in this study we have chosen to discard it as we focus on dividends as a means of distributing wealth. Furthermore we have excluded taxes both on a corporate and personal level, due to its complex nature. These could be seen as explanatory factors of a firm’s dividend policy.1.3 Thesis outlineThe first chapter introduces the reader to the subject and relevant background setting, as well as stating the purpose and research questions. In chapter 2 the relevant theories and previous empirical findings are presented, finishing off with the definition and description of our selected determinants of dividend payout ratio. Chapter 3 presents the data and methodology as well as the regression model used. In chapter 4 the empirical findings are presented and in chapter 5 the analysis and discussion connects the empirical findings back to the theoretical framework. Lastly, chapter 6 summarizes and concludes our research. 2.0 Literature review2.1 DividendsDividends can be defined as payments made to a firm’s shareholders out of past and current earnings (Pyles, 2014). The amount of earnings that a firm decides to distribute to its shareholders is set at a specific date by the board of directors, known as the dividend declaration date (Damodaran, 2010). The declaration date is important in the sense that the announcement to increase, decrease, or to maintain dividend puts across information upon which the market reacts to the changes that is most likely to occur (Damodaran, 2010).From the perspective of an investor the degree of reward they receive for investing in a firms shares is of high importance. Dividend can be categorized as the source of secured return that an investor receives (Pyles, 2014). Consequently, this leads to investors paying considerable attention whether a stock is purchased with or without dividends. For an investor to have the right to receive dividends, from the invested firm, there is a need to own the stock before the ex-dividend date. Purchase of shares post the ex-dividend date will result in forgoing the dividend received for the period, and as a result the market will reflect this loss with a fall in the stock price (Pyles, 2014).2.2 Irrelevance theoryMiller & Modigliani’s theory on dividend uses the perfect market assumption as a starting point under which the Irrelevance theory is developed. The Irrelevance theory states that a firm’s dividend policy is irrelevant since it doesn’t affect the value of the company under perfect market conditions (Miller & Modigliani, 1961). Thereby stating that whether the firm pays dividend or how much doesn’t have an effect on the value of the firm. M&M further states that investors can construct their own dividend by simply selling stocks to receive the preferred amount. If the firm pays to much dividend the investor just buys back shares and should therefore be indifferent between dividend and capital gains. As a result, shareholders are then unwilling to pay a premium for dividend paying firms hence the Irrelevance theory (M&M). However, DeAngelo, DeAngelo & Skinner (2009) argue that in a rational market the current market value equals the discounted future expected dividend. Thus the only way for an investor to get full value today is if the market expects the firm to fully distribute the value generated by the investment policy (DeAngelo et al, 2009). One might now consider the issue with the different taxation on capital gains but Black & Scholes (1974) and Miller & Scholes (1978) found that even this doesn’t affect the results. As well as developing the irrelevance theory in the groundbreaking article, M&M also acknowledge the information content of dividend, contributing further to Signalling theory first mentioned by Lintner (1956).2.3 Bird-in-hand theory“Better a bird in the hand than two in the bush” – John Lintner, 1956The Bird-in-hand theory first mentioned by Lintner (1956) states why a firm should pay dividend. It contradicts Miller & Modigliani’s Irrelevance theory since it states that investors are inclined to pay a premium for dividend stocks. Gordon in 1959 & 1962 elaborated on this theory by stating that investors have a likeness to shares that pay dividend today since there is higher uncertainty to future dividends and investors therefore use a higher discount rate for companies who don’t pay dividend. Paying dividend has other benefits such as a higher credit rating and this is useful since the firm will improve their ability to raise capital (Purmessur & Boodhoo, 2009). The theory is an equally controversial as M&M and has many opposes. 2.4 Signalling theoryIn 1970 Akerlof introduced his paper on information asymmetry regarding quality and uncertainty in “The market for lemons”, in which he develops the theory and uses the automobile market as an example. The theory splits the automobile market into four types of cars where there are new cars (which are lemons or cherries) and used cars (which are lemon or cherries). The problem arises when the owner after owning the new car for a period of time decides to sell. Since only he or she and no one else know if the car is a lemon, information asymmetry arises. If the car is a good car, the seller will not be able to get the price he or she wants and therefore keeps the car instead. Then lemons and cherries will sell at the same price thus establishing a new lower market equilibrium (Akerlof, 1970). This results in the lemons cars gradually driving out the cherries. Therefore the sellers of cherries need to signal to the market that their car isn’t a lemon, thus creating signalling. The same problem of information asymmetry arises when managers have more information than the market. They therefore want to signal they aren't a lemon. Signalling theory and dividend originates from Lintner who in 1956 discusses the change in stock price when managers change the dividend. Miller & Modigliani (1961) argue that dividends has no effect on share price in the Irrelevance theory, but they still mention that in the real world (excluding their perfect market assumption) there is a form of signalling effect, which they describe as information content. Fischer Black (1976) also argues that dividend policy says something the managers don’t say explicitly, in the sense that they will only raise the dividend if they are sure that the company can continue paying dividend in the future and will only lower it when the outlook for a quick recovery are poor. This gives investors a unique insight, different from other open sources, to how managers view the company’s future prospects. This is exactly what Miller & Modigliani (1961) mention when they describe the information content i.e. dividend polices conveys information. DeAngelo & DeAngelo (2006) further state that the literature sees signalling as the reason behind why companies have increased dividend despite the fact that this has resulted in higher taxes for investors (DeAngelo & DeAngelo, 2006). The theory connecting dividend and signalling is supported by Baker & Powell (2000) & Brav, Graham, Harvey, & Michaely (2005).2.5 Agency theoryIn Agency theory managers are seen as agents and shareholders as principals. The agent’s main task is to make decisions that are in the interest of the principals. The two debate over the investment policy and the optimal size of the company (Rozeff, 1982; Jensen, 1986). If managers are free to make choices in their own best interest, agency costs arises. Dividend can reduce these costs. Keeping excess cash in the company and thereby using internal financing avoids monitoring, which isn’t a good thing for principals since there is a correlation between the size of the firm and the compensation of the manager. This indicates that the firm might grow beyond their optimal size. Agency costs can be reduced by paying dividend since when there are higher free cash flows the agency costs increase and excessive free cash flow can be paid as dividend and thereby reduces the problem. Another solution to this discussed by Jensen & Meckling (1976) is increasing debt since it increases monitoring and results in managers having to hold their word of paying out parts of the free cash flow. Jensen finds that cutting dividend results in a large drop in the stock price, which would be consistent with Agency theory since this will leave managers with greater free cash flow and principals with increased costs for managerial risk aversion and empire building (Jensen, 1986). Academics have conducted research regarding Agency theory as a determinant of dividend policy. Rozeff (1982), Lloyd, Jahera, & Page (1985), Holder, Langrehr, & Hexter (1998) found significant evidence that agency theory can be seen as a determined of dividend policy used to monitor management of the company. The previous research motivates that a firm with a dispersed ownership structure will use dividend as a method to monitor agents/managers. In our research we will look at the correlation between FCF and the dividend payout ratio to see if Jensen (1986) findings apply to the Swedish stock market before and after the financial crises.2.6 Catering theoryThe Catering theory was developed by Malmcom Baker and Jeffrey Wurgler (2004) where they view dividend policy as being driven by investor preferences. Managers need to cater to the investors needs, i.e. their preference regarding cash generating stocks. Baker & Wurgler (2004) state that “further analysis confirms that these results are better explained by catering than other theories of dividends”. An important detail to notice is that the theory only focuses on the decision of instigating or not instigating dividend and should therefore not be mixed up with investment policy (Baker & Wurgler, 2004). 2.7 Life cycle theoryDeAngelo & DeAngelo (2006) develop the Life cycle theory were they view the need for the firm to distribute free cash flow as the driver of the optimal dividend policy. The theory summarizes many of the facts surrounding dividend, which they have collected from previous studies such as Lintner (1956), Fama & French (2001) and Jensen (1986). The theory cultivates how dividend is paid or not paid out during different life cycles of the firm. The model predicts that young high growth firms pay no dividend while mature firms with stable cash flows pay out a large part of the retained earnings. This relates to the financing decision that face firms and can be associated with Pecking order theory, which was summarized by Myers & Majluf (1984). It states how firms should prioritise their financing needs when managers have superior information. Internal financing is highest in the hierarchy and external financing is the lesser-preferred choice. If external financing becomes necessary then debt is higher in the hierarchy than equity (Myers & Majluf, 1984). They further state that no dividend should be paid if the firm has to finance themselves through other risky securities like selling stock. In their model dividend is highly correlated with manager’s view on the value of the assets, i.e. information asymmetry and dividend can help convey this information. Miller & Modigliani also mention the information content of dividend but establish in their perfect market assumption that dividend is irrelevant for the investment decisions made by the firm and therefore shouldn’t be a determinant (Miller & Modigliani, 1962). Fama (1974) set out to test this theory and found results indicating a relationship between dividend and investment policy. However Fama (1974) could not statistically confirm that this relationship holds and as a result could not reject M&M’s statement. Myers & Majluf (1984) discusses the investment decisions managers face for example when a firm has to issue equity to peruse a positive NPV project. They then risk ending up in the “financing trap” where managers with insider information don't want to issue new equity since it is a negative signal at the expense of the old shareholders. Passing on this good investment opportunity though, leads to a decrease in firm value (Myers & Majluf, 1984). 2.8 Measures of dividendIn order to conduct a study that investigates a firm’s dividend policy there is a need to discuss the different measurements of dividends. One of the most common measures of a firm’s dividend policy is the dividend payout ratio. The measurement allows us to observe the percentage of the firm’s earnings that is paid as dividends to the shareholders (Damodaran, 2010). Dividend Payout Ratio=Dividends Per ShareEarnings Per ShareFurthermore, the dividend payout ratio is a measurement that can be seen as highly independent to external factors, as the inputs of the formula are accounting figures from the firm (Penman, 2009). McManus, Gwilym & Thomas (2004) discuss that this measure of dividends is more informative regarding signalling effects due to the formula taking into account only internal factors.The second widely used measurement is the dividend yield. The measurement takes into consideration the firms dividends in relation to the stock price. The dividend yield measures the return that the investors make from dividends (Damodaran, 2010). The measurement takes into consideration the stock price of the firm, which make it subject to external factors that influence the measurement of the dividend policy. This means that influences that are not factors of the dividends policy can overstate or understate the firm’s policy of distributing its earnings to shareholders. Dividend Yield=Dividend Per ShareShare PriceHowever, Campbell and Shiller (1988), and Fama and French (1988, 1989) document that dividend yields is a measurement that can forecast a firm’s stock return and is therefore more informative than the dividend payout ratio. For our research we see that the dividend payout ratio is a more informative measurement of the firms dividend policy as it illustrates the percentage of earnings that is distributed to the shareholders. Furthermore, the measurement only takes internal factors into consideration, which we believe will be more informative in our analysis of the determinants of firm’s dividend policies. This is evidenced by the majority of studies that use the dividend payout ratio as a measurement of dividend policy (Rozeff, 1982; Baker & Smith, 2006; Gupta & Banga, 2010, Hellstr?m & Inagambaev, 2012; Abidin, Singh, Agnew, & Banchit, 2014; Bisschop, 2014).2.9 Previous studies of determinantsOver the years academics have tried to establish a set of determinants that impacts a firms dividends policy. Throughout these studies they have analysed various financial variables and their relationship with a firms dividends policy. We have observed that a majority of studies have analysed firms in the United States. However, recently academics have broadened the geographical area conducting research on dividend policy in various countries. Table 2.1 presents a sample of academic research on the topic and their finding of factors that influence a firm’s dividends policy. We can clearly see that there is a shortage of studies that analyse the Nordic region, a gap that we want to fill. Table 2.1 List of previous studiesAmong the discussions regarding dividends payout policy, Rozeff (1982) contributed with a widely recognized study that investigated what type of determinants that impact a firm’s motivation to implement a dividend policy. Rozeff’s (1982) study sampled 1000 US firms and analysed a variety of variables as explanatory factors of dividends. He found that the number of shareholders had a positive relationship with dividends. Furthermore, he found a negative relationship between insider ownership and dividends. These explanatory factors support the agency theory, as he argues that a firm with a high number of shareholders will pay higher dividends in order to reduce agency costs, whereas a firm with high percentage of insider ownership require less monitoring of the management of the firm (Rozeff, 1982). In addition to these factors Rozeff (1982) found that growth had a negative relationship with dividends, as high growth leads to higher investments. External financing will be costly and a high growth firm will avoid a dividends payout policy as this would lead to lower internal funds. Rozeff (1982) also found that high-risk firms avoid dividend payments due to their financial uncertainty. Firms with high risk will retain their earnings, as a constant dividends policy will be difficult to preserve. Many academics have replicated and expanded Rozeff’s (1982) work in order to confirm his result and also to strengthen his evidence. Lloyd et al (1985), Holder et al (1998) and Baker & Smith (2006) all found supporting evidence for Rozeff’s determinants. Lloyd et al (1985) included firm size as a variable in their study. They found that larger firms have a higher tendency to pay dividends. They argue that larger firms have easier access to capital markets resulting in less dependence on internal funds to finance investments. Additionally, Holder et al (1998) included a firm’s free cash flow as variable to their 477 firm sample. Their results demonstrated that firms with high free cash flow pay higher dividends. Their findings provided further support to Rozeff’s (1982) findings, as they argue that dividends are used as a method of monitoring management and reducing agency conflict of over-investment or empire building. In addition to these new variables Baker & Smith (2006) concluded that firms with low leverage levels pay higher dividends than firms with high degree of leverage. These studies indicate that dividend is used as a substitute to debt in order to monitor the firm’s management. Baker & Powell (2000) presented a study that examined the different theories regarding dividend payments to shareholders. They investigated corporate managers view on the relationship of dividends to value, optimal dividends payout ratio, and dividends relevance to the different theories such as, Signalling, Bird-in-hand, and Agency theory. They found that 77% of the respondents viewed dividends as a factor of the firm’s value, supporting the Bird-in-hand theory. In addition, the study found that managers followed Linter’s (1956) model of partial adjustment of dividends policy. They observed a consensus view that the dividend policy was an effective method of signalling information to shareholders. However, the study is consistent with Fisher Blacks (1976) statement that dividend is a puzzle as Baker & Powell (2000) found no theory that dominated among managers. Gupta & Banga (2010) conducted a study of 150 dividend-paying firms in India over a period of seven years, 2001-2007. They investigated the firm’s leverage level, profitability, liquidity, growth rate, and ownership structure as explanatory factors of dividends decisions. Their study showed that only the leverage level and liquidity of a firm could be seen as determinants of a firm’s dividend decision. They discuss that a firm with high level of debt will incur high interest payments and as a result decrease the level of dividends in order to lower the firms fixed charges. Furthermore, they found that liquidity has a positive relationship with dividends. They consider that a firm with high liquidity will be able to consistently pay a set dividend ratio.Osobov & Denis (2007) study investigated the determinants of dividends policy from an international perspective. They conducted an analysis of six countries, US, Canada, UK, Germany, France, and Japan. Osobov & Denis (2007) found that in all six countries large, profitable, and high earning firms tend to pay higher dividends. They argue that these findings support Jensen’s (1986) Agency theory and DeAngelo & DeAngelos (2006) discussion of firm life cycle as reasons for firm’s dividends decision. They argue that their results cast doubt on Signalling theory as their result contradict that small and less profitable firms use dividends as a means of signalling to investors. In addition, they found mixed results regarding growth opportunities as a variable, as dividend paying firms in US, Canada, and the UK tend to have less valuable growth opportunities unlike dividend paying firms in Germany, France, and Japan. Hellstr?m & Inagambaev (2012) conducted a study on large and medium cap Swedish firms. Their study investigated the dividends determinants of financial and non-financial firms. They couldn’t find any significant determinants for Swedish financial firms. However, from their sample they concluded that non-financial dividend paying firms had a positive relationship with free cash flow. They argue that investors prefer firms to distribute the excess cash in form of dividends rather than retain the cash internally. They found that high-risk firms with high growth tend to pay low dividends, consistent with Rozeff’s (1982) findings. Moreover, their findings support previous studies that large and profitable firm pay higher dividend. This supports DeAngelo & DeAngelo (2006) statement that mature firms tend to pay dividends.Abidin, Singh, Agnew, & Banchit (2014) studied industrial UK firms during the financial crisis of 2008, a period of economic adversity. They found that during this period dividends became more concentrated as small firms either reduced or omitted dividends. They argue that the later a firm in the life cycle the more able they are to pay dividends. They discuss that their results support the Life cycle theory and Signalling theory.Bisschop (2014) conducted one of the most recent studies regarding the determinants of dividends payout policy. He studied 78 Dutch firms over the period 2006-2012, and analysed the financial crisis impact on firm’s dividend policy. Bisschop (2014) found that profitability and firm size have a positive relationship with dividends, as previous studies have shown. Contrary to academic research he found a positive relationship between growth opportunities. In addition to this, the result showed a negative relationship between liquidity and dividends payout ratio. 2.10 Selected dividend policy determinantsIn this section we will discuss the different determinants of dividend policy and their relationship to the discussed theories. In addition, we will discuss the various measurements used to reflect the determinants. The presented determinants will be the basis of our analysis of Swedish firms and it is therefore important to illustrate their relevance to our study.2.10.1 Growth opportunitiesIt is argued that the firm’s growth opportunities is an important determinant of its dividend policy. A firm’s value can be categorized as the value of assets in place and the sum of the firm’s growth opportunities (Shin & Stulz, 2000). It is therefore an important variable that a firm takes into consideration in its decision making process. As a result, growth opportunity is a key determinant of a firm’s dividend policy as it can be used to convey information about the firm’s outlook of the growth to investors, known as signalling effect. On the other hand, the firm’s growth opportunities are also an important component in DeAngelo & DeAngelo (2006) theory of life-cycle period of a firm. They state that a firm in the early stages with valuable growth opportunities will avoid dividends, as internal financing is important to finance these opportunities. Due to the importance of growth opportunities and its interlinkage to different theories regarding dividends we see this as an important variable to analyse. A firm’s growth opportunities can be measured in various ways. Among academics, sales growth has been used as a measurement of the firm’s growth opportunities (Rozeff, 1982; Lloyd et al, 1985; Holder et al, 1998; Baker & Smith, 2006, Gupta & Banga, 2010; Hellstr?m & Inagambaev, 2012). A common measure of sales growth among previous studies has been to use the past sales growth of the firms (Rozeff, 1982; Lloyd et al, 1985; Holder et al, 1998; Baker & Smith, 2006, Gupta & Banga, 2010; Hellstr?m & Inagambaev, 2012). However, the drawback with this measurement is that it is backwards looking and does not accurately reflect the growth opportunities in the future. This has been evidenced in Rozeff’s (1982) and Lloyd et al (1985) studies, which have also used forecasted sales growth to reflect a forward looking measure. We should note that the forecasted sales growth measurement technique is subject to over and underestimations and can also be seen as inaccurate measurement technique. An alternative to sales growth as a measurement that can minimize these issues is the market value of equity to total assets of the firm. We observe that this measurement technique is the preferred choice in recent studies (Osobov & Denis, 2007; Bisschop, 2014).2.10.2 ProfitabilityProfitability of a firm is an important aspect of a firm’s dividend policy. The profitability of a firm is a good indicator of their ability to generate earnings to its shareholders. Furthermore, it is an important determinant as it can reflect the firm’s operational efficiency. The discussed theory above, states that dividends can be used to convey information to the market. A firm that may have low profitability is a candidate most in need of signalling to the market that the management has a positive outlook of the firm’s performance. For this reason we see the need to include this variable in our analysis. In order to assess firm profitability there are a variety of different measurement techniques that researchers have used. Baker & Smith (2006) and Gupta & Banga (2010) used the profit margin to measure the firm’s profitability. The profit margin is a measurement technique that shows how much the firm keeps as earnings for the sales. However, an alternative that measure earnings in relation to total assets has shown to be the preferred measurement of profitability, used in studies by Gupta & Banga (2010), Osobov & Denis (2007), Abidin et al (2014), and Bisschop (2014). Return on asset is a good indicator of the manager’s ability to use the company’s assets to generate earnings for shareholders. For our research we see that the profit margin is a useful measurement of the firms profitability. The measurement technique demonstrates the ability to create earnings from sales, which will in turn be used for dividends. This will also be useful in comparing the variable with growth opportunities, which are important explanatory variables for the Signalling theory.2.10.3 Cash flow A firm’s cash flow is an important determinant of the dividend policy. The cash flow of a firm illustrates the amount of cash that a firm can generates from its operations. The determinant is an important explanatory variable of Jensen (1986) theory of agency conflicts. A firm that can generate excess cash is subject to conflicts between managers and shareholders, where manager could use the excess cash for their own personal interest. As a result, investors of firms with high cash flows are in need of monitoring and as discussed above paying dividends is a form of monitoring mechanism.In order to measure a firm’s cash flow there are many different techniques that we could use. One of the preferred measurements of cash flow seen in previous studies is the free cash flow of a firm. Jensen (1986) explains that the free cash flow of the firm is the excess cash after the firm’s required investments. This measurement has been used in previous studies to investigate the relationship with the dividend policy (Holder et al, 1998; Baker & Smith, 2006; Hellstr?m & Inagambaev, 2012). An alternative measurement used by Gupta & Banga (2010) and Abidin et al (2014) is the cash flow from operations to demonstrate the cash available for the firm to expand and grow with investments in projects. The drawback of this measure is that it does not represent the excess cash of a firm, as it is pre-required investment. As a result, the measurement does not accurately reflect the issue discussed in agency theory.For our research we see that the free cash flow measurement technique as the most accurate technique as it measures the excess cash of the firm, which a better variable to relate to the Jensen (1986) Agency theory. 2.10.4 SizeFirm size is an important factor for the investment decision. It can be measured in different ways, using different proxies (Bisschop, 2014). These proxies can range from sales, number of employees, total value to total assets. Bigger firms are often in the mature part of the life cycle discussed by DeAngelo & DeAngelo (2006) and usually pay out a larger part of their retained earnings. We speculate that larger firms might also have done better in maintaining a higher dividend payout ratio. Looking at different ways of measuring firm size, Fama & French (2001) measure firm size as the percent of listed firms on NYSE that have a smaller market cap than the research object. Osobov & Dennis (2007) use the book value of total assets as a proxy for firm size. Hellstr?m (2012) uses another alternative, which is market value. Our research will use the natural logarithm of total sales as a measurement for size, which has been used by Lloyd et al (1985), Holder et al (1998), Baker & Smith (2006), DeAngelo & DeAngelo (2006) and Bisschop (2014). This is more appropriate than for example market value since we will be looking only at large cap firms, which differs less than looking at the total stock market.2.10.5 RiskRisk can be measured using different variables and there have been some alternatives in dividend research. Rozeff (1982) uses the Beta of a firm as a proxy for operating and financial leverage, since he argues a firm with higher leverage will have a higher alternative cost of dividend and therefore pay out a smaller portion of its retained earnings. This measurement technique is further supported by Lloyd et al (1985) study, which uses the Beta as a proxy for financial and operating leverage. Holder et al (1998) study uses an alternative measure of risk with standard deviation of the firm’s monthly returns as a proxy for risk, connecting it to transaction costs. We want to use risk as a variable to check against Pecking order theory, observing if higher risk firms will use retained earnings, internal financing prior to paying dividend, i.e. the investment decision. To test for this we will be using Beta as a proxy for financial and operating leverage. 2.10.6 Retained earningsWe want to look at Retained earnings to Total assets since this is a measure of how much internal financing the firm generates. Fama & French (2001) found that firms, which have higher earnings than investment-needs, in the past used to pay dividend but have now become non-payers. DeAngelo et al (2009) found a view consistent with Fama & French (2001) in that the number of firms with positive retained earnings has decreased and that this could explain the decreasing number of dividend paying firms since earnings serve as a basis for dividend (Lintner, 1956). This is a variable that could be used to show past profitability and need for external financing as it shows the ability to generate internal funds. There is different ways of measuring retained earnings. Osobov & Denis (2007) use something they call earned equity. This is derived by dividing retained earnings by total book equity, which is also used by Bisschop (2014). We want to look at this variable to see if the findings of the above are consistent with the Nordic market and in particular the Swedish market. We also want to connect the variable to the Pecking order theory, which looks at the different preferences of financing the firm. 2.10.7 Firm valueDividends impact on the firm value is a debated subject and connects to the Irrelevance, Bird- in-hand and Catering theory. The Miller & Modigliani Irrelevance theory states that the firm value isn’t affected by dividend whereas the Bird-in-hand theory states that investors will pay a premium on dividend paying stocks. In addition to the Bird-in-hand theory firms can cater to investors needs by paying dividends, which results in a premium on the share price. We can observe that no previous studies have taken this variable into consideration. Therefore we think it will be an interesting contribution to existing research. Our variable of choice to estimate the firm value to equity holders will be the Price to Earnings (P/E) ratio. The price to earnings ratio measures the market value per share divided by the earnings per share. It can be measured as trailing or forward looking. A firm with a high P/E is expected to showcase high growth and companies that aren’t earning will not have a P/E ratio. Liu, Nissim & Thomas (2002) find the historical P/E ratio to be a good indicator of frim value, in fact, better than models based on dividend or cash flow and therefore we will use a trailing P/E ratio.3.0 Methodology 3.1 Research method choices Our study will be an explanatory and descriptive one, since we are using a deductive approach and hypothesis testing (Bryman & Bell, 2011).Epistemology The research will follow the natural science methods and therefore we have chosen a positivistic approach. The advantage is that the author’s personal opinions are excluded given that we can observe the research object without affecting it (Bryman & Bell, 2011). OntologySince we will look at historical data and won’t use a subjective interpretation method we instead have an external approach and use an objective approach when conducting the research (Bryman & Bell, 2011). These choices reflect a common approach in quantitative finance and we will follow in that tradition (Bryman & Bell, 2011).3.2 The data 3.2.1 Sample and sample sizeOur research analyses the determinants of dividend payout policy in the Swedish market. Due to the availability of data we have chosen to analyse publicly listed firms. We have chosen to analyse the Swedish NASDAQ OMX Stockholm Large Cap list (as the original sample) due to the higher probability of dividend paying firms among the larger corporations. The original sample consisted of 94 firms. An important factor to adjust for in the sample, is the fact that the firm had to be listed during our time periods, which resulted in removing 14 firms. After adjusting and removing financial firms (-18), due to their characteristics, as well as non-dividend paying firms (-17), the sample size was reduced to 45. Next we adjusted the sample size by removing one of the stocks for firms who had dual class listed shares. This resulted in a total of 35 firms that we analysed. Table 3.1 Sample selection180213159690003.2.2 Data collection & time periodOur time period can be divided into three periods. Firstly, we looked at the time period between 2003-2013 in order to define the determinants of dividend policy of Swedish firms. Secondly, in order to analyse the effects of the financial crisis on Swedish companies, we compared the determinants of two separate time periods. The first time period, which we categorize as pre crises period is 2003-2007 and the second time period listed as post crises between 2009-2013. In our research we consider 2008 to be the crisis year, which means that we have excluded it from the pre and post periods. The data was collected from Thomson Reuters Datastream database. It was then summarized in a spreadsheet before manually constructing the variables and running the various regression models. 3.2.3 Variables definitionIn order to conduct our research on the determinants of dividend payout policy of Swedish firms we conducted a regression analysis on the following variables. The different determinants of dividend policy have been discussed above and are based on previous studies. Firstly, the dependent variable of our research is the dividend payout ratio, which reflects a firm’s dividend policy. The dividend payout ratio is measured in this study as total dividend paid by total earnings.Dividend Payout Ratio=Dividends Per ShareEarnings Per ShareSecondly, we analyse seven independent variables that are seen as determinants of the firm’s dividend policy. These independent variables have been selected on the basis of the discussed literature review. One of our independent variables will represent the firm’s growth opportunities. This will be measured as the total market value of equity divided by the total asset of the firm. Growth Opportunities=Market CapitalizationTotal AssetsProfitability is the second independent variable that will be measured. We have defined profitability as the firm’s profit margin. This is measured as the firm’s net income divided by total sales.Profitability= Net income Total SalesThe third variable represents the firm’s cash flow. We measured the firm’s cash flow as the free cash flow divided by total assets. Free cash flow was computed by subtracting capital expenditures from operating cash flow of the firms.Cash Flow=(Net Cash Flow From Operations-Capital Expenditure)Total AssetsAnother variable that will be analysed is the size of the firms. In order to reflect this in our analysis we have used the natural logarithm (ln) of sales as a proxy. Size=ln?(Total Sales)To reflect the firm’s risk we have used the firm specific betas (β) as an independent variable. This has been calculated as the volatility of the firm’s share price in comparison to the market. The firm specific beta has been calculated manually from the firm’s share price and market index weekly returns. We have used the OMX Stockholm 30 index to represent the market.β=Cov(ri,rm)Var(rm)Where: ri= return of stockrm= return of market indexThe sixth independent variable that we investigated is the retained earnings of firms. We have measured the firm’s retained earnings variable as the total retained earnings divided by total assets.Retained Earnings=Total Retained EarningsTotal AssetsLastly, the firm value will be measured as an independent variable. This is a new variable in studies of determinants of dividend payout policy. We have measured the firm’s value by using the price to earnings ratio. In order to compute this ratio we have divided the firm’s share price by their earnings per share.Price To Earnings Ratio=Share PriceEarnings Per Share3.3 Regression modelSince our research consists of both a time series and a cross sectional dimension we have a dataset known as panel data (Brooks, 2008). In order to achieve unbiased and consistent estimates our model needs the Ordinary Least Squares (OLS) five assumptions to hold. These assumptions are as follows:3.3.1 Average values of the errors is zero The first assumption for OLS is that the mean value of the errors is zero (Brooks, 2008). Since we’ve included a constant term in our regression, this assumption isn’t violated (Brooks, 2008).3.3.2 Variance of the errors is constant The second OLS assumption is that the variance of the errors is constant, also known as the assumption of homoscedasticity (Brooks, 2008). Using the Breusch-Pagan-Godfrey test we tested if the variance of the errors are constant. The null hypothesis under the test states that there is homoscedasticity in the variance of the errors. The test shows that we cannot reject the null hypothesis implying that there is no evidence of heteroscedasticity.3.3.3 Covariance between the error terms is zero The third assumption implies that the covariance between the error terms, which over time or cross-sectionally, must be zero (Brooks, 2008). This is known as autocorrelation and we tested for this using the Durbin-Watson (D-W) test, which has a null hypothesis that there is no evidence of autocorrelation. Since our D-W statistic was found to be 2, we cannot reject the null hypothesis and therefore there is no evidence of autocorrelation. 3.3.4 Independent variables are non-stochasticThe forth assumption is that the independent variables are non-stochastic. This implies that the error terms are not correlated with the independent variables (Brooks, 2008). We tested this by conducting a correlation matrix between the independent variables and the error terms. The correlation matrix showed no evidence of higher correlation, therefore we can conclude that the independent variables are non-stochastic. 3.3.5 Disturbances are normally distributed The fifth and last assumption is that the disturbances are normally distributed (Brooks, 2008). Testing this data for normal distribution, we used the Jacque-Berra test and found the coefficients to be normally distributed at a significant level of 99 % with the exception of risk, which was significant at the 90% level.3.3.6 No multicollinearityAn additional assumption when using an OLS model is that the independent variables are not significantly correlated with each other. From conducting a correlation matrix we can conclude that there is no significant correlation between any pair of variables. The highest correlating pair is observed between CASH FLOW and GROWTH with 0.725, which is below 0.8.3.3.7 Fixed & Random effectsOur datasets embodies information both cross-sectionally and over time, which means there is a need to apply different estimator approaches to analyse the data. The most broadly used are pooled, fixed and random effects regressions.The first regression that we conducted was a pooled least squared regression. This model assumes that the intercepts are the same for each firm and this is perhaps an inappropriate assumption, since the pooled regression neglects the cross-sectional and time-series nature of the data. We therefore used the Redundant Fixed Effects test in order to determine whether pooled OLS or fixed effects should be used. The test showed that we should use cross-sectional fixed effects. Thereafter, we conducted the Hausman test in order to test whether fixed or random effects should be used. The null hypothesis under the Hausman test states that random effects are the preferred estimator approach. Our test showed that we could not reject the null hypothesis, which implies that random effects should be used to analyse our dataset. As a result the regression that we have chosen to run is:yit= α+ βxit+ ωitDividend Payout Ratioit= α+ βGrowthit+βProfitabilityit+βCash_Flowit+βSizeit+βRiskit+βRetained_Earningsit+βFirm_Valueit+ ωit3.3.8 T-testTo be able to test if our independent variables are significant, as well as testing our various hypothesis stated we have used the t-test. Following the fact that our data is normally distributed the t-test is appropriate. The null hypothesis of the t-test is Beta = 0. The alternative hypothesis is Beta does not equal = 0. The formula for the t-test is given as follows (Brooks, 2008):t= β-β*SE(β)Where: β = sample meanβ*= population meanSE(β) = standard deviation of estimatorIn order to test the hypothesis a critical value for a two-tailed test is needed. This is provided by a t-table. Rejecting the null hypothesis at the 95% level indicates that the test is statistically significant (Brooks, 2008). 3.3.9 Z-scoreIn our research we have investigated the effects that the financial crisis had on the determinants of dividend payout policy. We have conducted two separate regressions, pre and post crisis, to illustrate the determinants of the specified time period. In order to compare the coefficients of these two regressions we need to assess the significance of the difference using the z-test (Clogg, Petkova & Haritou, 1995). z=β1-β2SEβ12+SEβ22Where: β1 = coefficient from regression oneβ2 = coefficient from regression twoSE(β) = standard deviation of estimatorThe null hypothesis under the z-test states that the coefficients are comparable (Paternoster, Brame, Mazerolle, & Piquero, 1998). This means that rejecting the null hypothesis we can state that there is a significant difference in the coefficients of the two regressions. The z-table illustrates that the critical value for a 5% two-tailed test is 1.96. The z-test was conducted on all seven determinants of dividend payout policy.3.3.10 R-squaredThe R-squared is a measure of how well the dependent variable is explained by the independent variables (Brooks, 2008). If R-square is high then the model fits the data well and inversely if its close to zero it isn’t a satisfactory fit to the data (Brooks, 2008). Our R-square using the random effects model were all above 0,90, which indicates that the independent variables explain more than 90% of the dependent variable in all of the three time periods.3.4 Validity & Reliability3.4.1 Validity of the independent variablesValidity is one of the more important research criteria and examines if the stated variables, created in the purpose to measure the dependent variable, really do measure this (Bryman & Bell, 2011). Internal validity refers to the causality of the independent variables as well as the accuracy of the measurement techniques (Bryman & Bell, 2011). To check the validity of our independent variables we tested the determinants of dividend policy on the whole time period between 2003-2013. It is important that the independent variables explain the dependent and we have done a thorough examination of previous researchers use of determinants to make sure the measures are correct. However consistent our variables may be, there is a possibility that there are other independent variables that could explain the dependent variable that are not included in our research. External validity refers to the generalization of the research to other markets and contexts. Concerning the generalization of our results we feel that because of the number of companies and long time period our research is applicable to Swedish companies. However, since we have analysed the NASDAQ OMX Stockholm Large Cap firms, this can be seen as a simplification of the Swedish market and this could be seen as a weakness. 3.4.2 ReliabilityReliability is especially important in quantitative research since it tells the reader that he or she can rely on the results (Bryman & Bell, 2011). The variables (determinants) we have used are collected from examined previous studies where other researchers have tested them. This shows that the selected variables are reliable in the study and thereby making our research reliable. Data was collected from Thomsom Reuters Datastream, which in turn relies on annual reports. Annual reports can be seen as biased due to the fact that the firms publish themselves, however since the annual reports are subject to auditing rules and accounting principles we view them as unbiased. As the reader can both confirm the sources of data and the methods of using them, this makes our research consistent and auditable.In addition, reliability tells us if it would be possible to replicate the study and get the same results. Since we have clearly stated the methods used to analyse our data and statistically tested the results we can conclude that our research can be replicated. As a result, our research can be seen as reliable.4.0 Empirical findings4.1 RegressionsIn this section we will present our results that we produced from the Random Effects Panel Data regression. From the discussed testing of the sample data we can state that we meet the various OLS assumptions, which means that our results are unbiased and consistent. Furthermore, we will demonstrate the relationship between the selected company determinants and the dividend payout ratio. In addition, the statistical significance of the relationships will be presented in the tables below. 4.1.1 Whole period regression (2003-2013)In order to determine the determinants of dividend policy for Swedish firms we have run a regression with a time frame of eleven years. The results from the regression for the Swedish firms listed on the large cap are presented in Table 4.1. The table illustrates that the regression has a R-square of 94 percent. This indicates that 94 percent of the changes in Dividend Payout Ratio (DPR) can be explained by the independent variables in the regression. In addition, we can state that the regression is a good fit for the data. Table 4.1 Regression results for whole period (2003-2013)Cross sectional Random effects, period 2003-2013VariableCoefficientStd. Errort-StatisticProb.??C-1.9280800.672284-2.8679540.0044Growth0.0329010.0519600.6332040.5270Profitability0.2924470.1730051.6903980.0918Cash_Flow-0.0622910.784636-0.0793880.9368Size0.1446900.0405463.5685200.0004Risk-0.1882800.098309-1.9151820.0562Retained_Earnings-0.5464820.308590-1.7709010.0774Firm_Value0.0148490.00018381.121580.0000R-squared0.946596???Looking at the whole period with the intent of determining our determinants, we can observe that two variables are significant at the 5 % level, Size and Firm Value. A further three variables are statistically significant at the 10 % level, Profitability, Risk and Retained Earnings. The table shows that the firm’s growth opportunities have a positive relationship with the dividend payout ratio. However, the relationship with the DPR is insignificant, indicated by the p-value of 0.527. Profitability has a significant and positive relation to DPR, which indicates that the more profitable a firm is the higher dividends they paid. Cash flow shows a slight negative relationship with DPR but is not statistically significant. On the other hand, a firms size has a positive and significant effect on DPR, which shows that a larger firm can pay higher dividend. Risk has a negative and significant effect on DPR and leads us to believe that higher risk leads to a lower DPR. Furthermore, retained earnings shows a negative and significant effect on DPR, which indicates that higher retained earnings leads to firms paying lower dividend. Firm Value has a significant coefficient of 0.01, which indicates that firm value has a minimal relationship with DPR.4.1.2 Pre-crisis period regression (2003-2007)Table 4.2 shows the regression results for Swedish firms prior to the financial crisis that was experienced in 2008. The table shows that the regression’s has a R-square of 0.96, which indicate that the model fits the sample data. Additionally, the R-square result suggests that the selected independent variables can explain 96 percent of the variation in the dividend payout ratio. Table 4.2 Regression results for pre-crisis period (2003-2007)Cross sectional Random effects, period 2003-2007VariableCoefficientStd. Errort-StatisticProb.??C-2.3365410.889031-2.6281880.0094Growth0.0888790.0789851.1252600.2621Profitability0.3171920.2749131.1537910.2502Cash_Flow-0.6646881.063219-0.6251650.5327Size0.1761830.0546093.2262360.0015Risk-0.2634960.148864-1.7700420.0785Retained_Earnings-0.9971070.521457-1.9121580.0576Firm_Value0.0144570.00023362.140870.0000R-squared0.960450???The regression output indicate the more growth opportunities that a firm has the higher the dividend payout ratio, shown by the positive coefficient. However, the positive relationship is not statistically significant. Profitability also has a positive but not significant relationship with DPR, though the effect is larger, meaning higher profits could lead to higher DPR. However, the results show that cash flow has a negative relationship with DPR, but is not statistically significant. The results from the table above show that the size of a firm has a positive and significant relation with DPR, demonstrated by the coefficient of 0.176. This implies that larger firms pay higher dividend during the period. Alternatively, firms that are perceived to possess higher risk, tend to pay fewer dividends to shareholders, indicated by coefficient of -0.26. Moreover, the results show that retained earnings has a negative and significant relationship with DPR, leading us to believe that when firms have higher retained earnings they pay less dividend. Lastly, firm value has a statistically significant coefficient of 0.01, which indicates that the firm value of firms has minimal effect on the dividend payout ratio. 4.1.3 Post-crisis period regression (2009-2013)Table 4.3 presents the regression results for the post-crisis period, 2009 to 2013. The results show that the model fits the data and that the explanatory variables can explain 91 percent of the variations in the dividend payout ratio. This is demonstrated by the R-square result of 0.916. Table 4.3 Regression results for post-crisis period (2009-2013)Cross sectional Random effects, period 2009-2013VariableCoefficientStd. Errort-StatisticProb.??C-1.9677420.999361-1.9690000.0506Growth0.0642200.0727820.8823710.3788Profitability0.4128840.2423551.7036370.0903Cash_Flow-0.2135381.095445-0.1949330.8457Size0.1274510.0590732.1575060.0324Risk-0.0250960.145525-0.1724500.8633Retained_Earnings-0.2935270.471003-0.6231960.5340Firm_Value0.0176910.00040843.311590.0000R-squared0.916326???In this period, growth has a positive but insignificant relationship with DPR. Profitability on the other hand has a significant and positive relationship with DPR, indicated by the coefficient of 0.41. This means that higher profitability of firms during the post-crisis period caused the dividend payout ratio to increase. Cash flow of a firm has a negative but not significant relationship with DPR. However, firm size has a positive and significant relationship with DPR indicating that larger firms pay higher dividend during the post-crisis period. The riskiness of a firm has a slight negative relationship with DPR but is statistically insignificant. Furthermore, we see that an increase in retained earnings would decrease a firm’s DPR, however the result is insignificant. Firm value affects DPR in a slightly positive and significant effect. 4.1.4 Z-scoreThe Z-score is used in order to determine whether there is a statistically significant difference between the independent variables of the pre and post crisis periods. The null hypothesis under the z-test states that the coefficients are comparable (Paternoster, Brame, Mazerolle, & Piquero, 1998). The z-test critical value of 1.96 is used for a 95 percent confidence level and 1.64 for a 90 percent confidence level.Cash flow, Risk and Retained Earnings showed considerable differences in beta between period 1 and 2 although they were not significant as seen in the table 4.4. Table 4.4 demonstrates the changes in the independent variables and whether they are statistically significant. Since we can reject the null hypothesis for Firm_Value we can state that there is a significant difference in Firm_Value between the two regressions. However, the test implies that the coefficient changes of the other variables cannot be rejected and thus there was no significant difference in the other 6 coefficients between the two periods. Table 4.4 Z-scores for the independent variablesVariableBeta period 1Beta period 2Beta diffSE^2 period 1SE^2 period 2Z-scoreReject/Don't rejectGrowth0.088880.06422-0.024660.006240.005300.22959Don't rejectProfitability0.317190.412880.095690.075580.05874-0.26111Don't rejectCash_Flow-0.66469-0.213540.451151.130431.20000-0.29553Don't rejectSize0.176180.12745-0.048730.002980.003490.60576Don't rejectRisk-0.26350-0.025100.238400.022160.02118-1.14517Don't rejectRetained Earnings-0.99711-0.293530.703580.271920.20340-1.02052Don't rejectFirm Value0.014460.017960.003500.000000.00000-7.45780RejectNote: the critical value is 1.96 two-tailed test (α=0.05)Note: the critical value is 1.64 two-tailed test (α=0.10)5.0 Analysis and discussionIn this section we will analyse and discuss the various hypotheses stated in the introduction and by doing so, answer our two research questions. Moreover, we will elaborate on the fact if the existing research and literature can be applicable to the Swedish market and if so, what they imply. Furthermore, we will analyse whether the investigated determinants changed in any significant way, as a result of the financial crisis.5.1 Determinants of the dividend policyIn this part of the section we will analyse and discuss the results from the regression analysis of the whole period, 2003-2013, to determine which of the selected company variables can be seen as the determinants of the dividend payout policy of Swedish firms.5.1.1 GrowthH1a: Growth will have a positive effect on dividend policy Our empirical findings found that growth opportunities and dividend payout ratio has a positive relationship, which is equivalent to the findings of Osobov & Denis (2007). The result suggests that Swedish firms with valuable growth opportunities tend to pay higher dividends. This is consistent with Signalling theory, where Black (1976) argues that dividend is used to convey information that manager’s don’t explicitly communicate. As a result, we can assume that Swedish firms use dividends as an effective tool to communicate to shareholders that the managers see valuable growth opportunities in the future, as Baker & Powell (2000) argue in their study. However, the results are contradictive to the findings of Rozeff (1982), Lloyd et al (1985), Holder et al (1998), and surprisingly Hellstr?m & Inagambaev (2012). These previous studies found that growth had a negative relationship to dividend payout ratio. Their results suggest that the Life cycle theory explains the relationship, where young firms with high growth avoid paying dividends due to the need of financing of future investments. In spite of the opposing result that our research found, the relationship between growth opportunities and dividend payout ratio was statistically insignificant and therefore we cannot confidently confirm our hypothesis. 5.1.2 ProfitabilityH1b: Profitability will have a negative effect on dividend policy Profitability showed a positive and significant relationship, which was contradictive to our hypothesis. Therefore, the hypothesis can be rejected, which indicate that the Signalling theory does not hold. Signalling theory argues that more profitable firms are less inclined to signal future profits. Previous research states that this effect would be applicable to firms experiencing lower profitability during a shorter period. These firms would be the ones in dire need of signalling higher future profits as mentioned by Akerlof (1970). Our results are not in line with Signalling theory and cast a doubt on signalling as a determinant of dividend payout policy in the Swedish market especially since our sample consists of large dividend paying companies who according to signalling theory are the ones least in need of signalling as also found by Osobov & Denis (2007) Abidin, Singh, Agnew & Banchit (2014).Though contradictive to Signalling theory, our results show that profitability has a positive effect on dividend and is therefore inline with DeAngelo & DeAngelo (2006) Life cycle theory since a more mature firm with higher profitability can pay higher dividend. These findings coincide with Abidin, Singh, Agnew, & Banchit (2014).5.1.3 Cash flowH1c: Cash flow will have a positive effect on dividend policy Jensen (1986) argue that the higher the free cash flow, the higher probability of agency conflict. They suggest that firms can use dividend as a monitoring tool to avoid agency conflicts between managers and shareholders. Holder et al (1998) study suggested that the Agency theory could be applied to dividend paying firms in the US, as he found a positive correlation between the cash flow and dividend payout ratio of US firms.In contrast to Holder et al (1998), our empirical finding indicate that cash flow has a negative relationship and that the Agency theory is not applicable to the Swedish market. This is surprisingly different to the Hellstr?m & Inagambaev (2012) study that found a positive relationship between Swedish firm’s cash flow and dividend policy. However, our result were found to be statistically insignificant, which does not enable us to determine if the negative relationship holds. As a result, we cannot argue against the findings of Hellstr?m & Inagambaev (2012).5.1.4 SizeH1d: Size will have a positive effect on dividend policy The Life cycle theory suggests that mature firms are more likely to pay dividend due to their access to external financing (DeAngelo & DeAngelo, 2006) (Abidin, Singh, Agnew, & Banchit, 2014). A positive and significant relationship was found and this result concurs with previous studies that confirm that size has a positive effect on dividend (Hellstr?m & Inagambaev, 2014) (Osobov & Denis, 2006). Thereby this theory can be applied to the Swedish market and thus confirms our hypothesis. Because of relatively low manager ownership the goals of the agent could differ from the principal and larger size could potentially lead to empire building, thus creating an agency cost (Jensen, 1986). Our result concurs with Hellstr?m & Inagambaev (2014) in that to avoid agency costs, firms pay higher dividends (Jensen, 1986). 5.1.5 RiskH1e: Risk will have a negative effect on dividend policy An examination of our result show that the significant and negative relationship between risk and dividend payout ratio is consistent with existing research by Rozeff (1982), Lloyd et al (1985), Holder et al (1998), and Hellstr?m & Inagambaev (2012). This indicates that Swedish firms dividend payout ratio decreases when they experience higher operational and financial leverage. The higher operational and financial leverage can result in volatility of the firm’s cash flows, as explained by Rozeff (1982). If firms experience volatility in the cash flow and future uncertainty, preserving a stable dividend payout ratio will be difficult and would lead firms to target a lower DPR, which they could meet. In addition to the operational and financial leverage, the relationship can also be explained by Myers & Majluf (1984) Pecking order theory. Firms with high risk can experience limited access to external financing and decide to retain earnings and lower their DPR to finance investments. Moreover, banks and debt capital markets will demand higher interest rates due to the higher risk of the firms, which in turn will increase the firms financial risk. For this reason Swedish firms that have high risk would prefer to retain a larger proportion of their earnings as an alternative to external financing. As a result, the findings are in line with our hypothesis. 5.1.6 Retained earningsH1f: Retained Earnings will have a positive effect on dividend policy A negative and significant relationship was found between retained earnings and dividend policy. This finding was unexpected and suggests that the Life cycle theory doesn’t hold since it states that when a firm becomes more mature and stable it should have higher retained earnings and therefore be able to pay out a higher dividend (DeAngelo & DeAngelo, 2006). The Agency theory is also falsified by our result since when a firm has higher retained earnings they pay dividend to avoid agency costs (Jensen, 1986). It also opposes Osobov & Dennis (2007) whose research on the topic in different western capitalistic countries all concluded positive relationships between retained earnings and dividend. Thus, it can be said that their results are not applicable to the Swedish market and that our hypothesis is rejected. A possible explanation could be that Swedish firms with higher retained earnings engage in stock repurchases instead of dividend.5.1.7 Firm valueH1g: Firm value will have a positive effect on dividend policy A positive and significant relationship was established between firm value and dividend policy. The results are in line with the theory by Baker & Wurgler (2004) that show dividend is driven by investor demand. The Catering theory and Bird-in-hand theory states that investors are willing to pay a premium on dividend paying shares because managers cater to investors needs and that investors prefer a dividend today because of the higher uncertainty in the future (Baker & Wurgler, 2004; Lintner, 1956). Our results are therefore inline with the Bird-in-hand theory developed by Lintner (1956). Although firm value has a positive effect, we note that it has a relatively little effect on the dividend payout ratio in our research. Therefore we feel that the irrelevance theory cannot for certain be falsified since a large change of 1 unit in the Price to Earnings ratio only results in an increase of 0.01 in the dividend payout ratio. 5.1.8 What are the determinants of the dividend policy in Swedish firms?An examination of our results show that we can confirm that the following determinants of dividend policy, profitability, size, risk, retained earnings and firm value, can be applied to the Swedish firms. Table 5.1 Determinants of dividend policyDeterminants of dividend policyTheoryProfit.SizeRiskFirm V.R.EIrrelevance Theory???O*?Bird-in-hand Theory???C*?Signalling TheoryO**????Agency Theory?C*??O**Catering Theory???C*??Life cycle TheoryC**C*??O**Pecking order Theory??C**??C= Confirms?????O= Opposes?????**=Significant at 5% level?????*=Significant at the 10% level????The results provide evidence that Swedish firms apply DeAngelo & DeAngelo (2006) Life cycle theory, which is indicated by profitability and firm sizes positive relationship to the dividend payout ratio. We assume that mature firms with higher profitability have greater ability to distribute larger proportions of earnings to its shareholders. In addition, the relationship between profitability and dividend payout ratio questions whether dividends is used to signal information content since Signalling theory suggests that lower profitable firms are in greatest need of conveying information to investors. The positive and significant relationship between size and dividend payout ratio suggest that firms DPR increase when the firm size grows. From the empirical findings we can assume that Swedish firms use dividends to counter agency conflicts of empire building by paying dividend. Moreover, risk showed that Rozeff (1982), Lloyd et al (1985), Holder et al (1998), and Hellstr?m & Inagambaev (2012) findings of higher operational and financial leverage has a negative effect on dividend. The result also supports Myers & Majlufs (1984) Pecking order theory that firms with higher risk prefer internal financing in order to limit the financial risk.Our research further showed that firm value had a positive and significant relationship with dividend payout ratio and thus confirms Lintner (1956) Bird-in-hand theory as well as Baker & Wurgler (2004) Catering theory. The results imply that investors in Swedish firms are willing to pay a premium on dividend paying shares. However, the change in firm value causes a minimal adjustment to the dividend payout ratio, which might suggest that the Bird-in-hand theory is not extensively applied in the Swedish market, and thus the Irrelevance theory cannot be completely disregarded.Unexpectedly, our results show that retained earnings contradict the Life cycle theory by DeAngelo & DeAngelo (2006), which states that mature firms with high past and present earnings are more likely to distribute these earnings to its shareholders. Therefore the negative relationship disputes this assumption. Furthermore, the result goes against the Agency theory by Jensen (1986). The relationship indicates that firms retain a larger amount of earnings and as a result avoids monitoring and this relationship is there not in line with Agency theory.5.2 Were the determinants of dividend policy amongst Swedish firms affected by the financial crises?H2: There will be a change in determinants between the two periods Analysing the results from the pre and post crisis periods we can observe some surprising changes in the coefficient of the selected determinants of the dividend payout policy. Table 5.2 Difference in coefficients (pre and post crisis period)VariableBeta period 1Beta period 2Beta diffZ-scoreReject/Don't rejectGrowth0.088880.06422-0.024660.22959Don't rejectProfitability0.317190.412880.09569-0.26111Don't rejectCash_Flow-0.66469-0.213540.45115-0.29553Don't rejectSize0.176180.12745-0.048730.60576Don't rejectRisk-0.26350-0.025100.23840-1.14517Don't rejectRetained Earnings-0.99711-0.293530.70358-1.02052Don't rejectFirm Value0.014460.017960.00350-7.45780RejectNote: the critical value is 1.96 two-tailed test (α=0.05)??Note: the critical value is 1.64 two-tailed test (α=0.10)??The largest changes in coefficients were seen in retained earnings, cash flow, and risk. These results indicate that in the post crisis period retained earnings, cash flow, and risk had a weaker negative relationship compared to the pre crisis period. Although the regression results show some changes in the coefficients of the selected determinants, the z-test implies that six of the coefficient changes are statistically insignificant. The results of the z-test are consistent with Bisschop (2014) and Abidin et al (2014) studies. These studies found that the financial crisis did not have any statistically significant change in the determinants of dividend policy. Conversely, our study showed that firm value had a statistically significant change in the coefficient. However, in table 5.2 we can observe that the change is minimal. Thus, we can claim that the financial crisis did not alter the determinants of Swedish firms dividend policy, even though the financial crisis had an impact on the Swedish economy and forced many firms to take actions to avoid the impact. 6.0 Conclusion6.1 SummaryThe main purpose of the research was to investigate the determinants of dividend policy in the Swedish market and to determine whether the financial crisis had any significant effect on these determinants. In order to answer the two research questions, we employed a Random effects regression model using a selected sample of 35 firms from the NASDAQ OMX Stockholm Large Cap list. Three time periods were examined, 2003-2013 to examine the determinants of dividend policy and 2003-2007 & 2009-2013 to empirically test the difference in the determinants pre and post the financial crisis. After examining existing literature and previous research the following determinants were chosen: Growth, Profitability, Cash Flow, Size, Risk, Retained Earnings and Firm Value. The study found that profitability, size, risk, retained earnings, and firm value can be seen as determinants of dividend policy for Swedish firms. The empirical result showed that profitability and size established a positive relationship, which supports DeAngelo & DeAngelo (2006) Life cycle theory stating that mature firms with higher profitability are more inclined to pay dividend to its shareholders. Moreover, size also suggests that larger firms will pay higher dividends in order to avoid empire building as Jensen’s (1986) Agency theory proposes. Furthermore, risk showed a negative relationship, which coincides with Myers & Majluf (1984) Pecking order theory since internal financing is the preferred choice when firms have limited access to external financing. The study found that firm value established a positive relationship, which is inline with Linter (1956) Bird-in-hand theory and Baker & Wurgler (2004) Catering theory stating that investors would pay a premium for dividend paying firms because of the uncertainty of the future and preference for cash generating shares. Conversely, the study found that retained earnings had a negative relationship, which casts doubt on the Life cycle and Agency theory’s applicability to the Swedish market.However, some questions were left unanswered as the study found that growth and cash flow relationships with dividend payout ratio were statistically insignificant.Furthermore, this study investigated whether the determinants of dividend policy were affected by the financial crisis. The results showed that there were no statistically significant changes between the pre and post period, except for firm value. However, the change observed was minimal which does not alter the assumption of the effect that firm value has on the dividend policy.Overall, the study found that most of the selected company determinants had a significant relationship with dividend payout ratio and that the financial crisis did not cause changes in the relationship. In addition, the study provided supporting evidence for Bird-in-hand, Catering, Agency, Life cycle, and Pecking order theory as well as casts doubt on Signalling theory.6.2 ContributionThis study has contributed to the discussion of why firms pay dividend, by investigating the relationship between the selected company determinants and the dividend payout ratio. The relationships are analysed to determine whether existing theories and previous research are applicable to Swedish firms. The outcome of the results provide a piece of the dividend puzzle mentioned by Black in 1976. Furthermore, the study has focused on Swedish firms, a market that has not been extensively researched. This study can therefore be used by academics and professionals to understand how Swedish firm’s dividend policy differs from other researched markets.In addition, the study has included a new variable, firm value, which was measured by the price to earnings ratio. The study found a significant relationship, which indicates that the variable can be used in future research. The study has also clarified the impact of one of the most challenging financial crises in recent years. The results provide evidence, for academics and professionals, that Swedish firm’s dividend policy were not significantly impacted by a critical event such as one experienced in 2008.6.3 Further researchThe study showed that most determinants could be related to existing theory except for one independent variable, Retained Earnings. It is therefore interesting if future research could investigate why retained earnings opposes the existing theories. Moreover, future research could be conducted to demonstrate the different outcomes using the dividend yield instead of the dividend payout ratio as a dependent variable.We would also recommend increasing the sample size to include NASDAQ OMX Medium and Small cap listed firms in order to better reflect the Swedish market. By covering all listed firms the researcher might get a more accurate result.Another potential addition to existing research could be achieved by investigating the differences between other Nordic countries to get a complementing picture of the Nordic region. 7.0 References Abidin, S., Singh, V., Agnew, M., & Banchit, A. (2014). Determinants of Dividend Payout Policy for UK, Twenty-First Annual Conference of the Multinational Finance Society Proceedings, Session 15, Available online: [Accessed 10 March 2015]Akerlof, G. A. (1970). The Market for "Lemons": Quality Uncertainty and the Market Mechanism. The Quarterly Journal of Economics, Vol. 84, No. 3, pp. 488-500Baker, H. K., & Powell, G. E. (2000). Determinants of Corporate Dividend Policy: A Survey of NYSE Firms. Financial Practice and Education, Vol. 10, No. 1, pp. 29–40.Baker, K. H., & Smith, D. M. (2006). In Search of a Residual Dividend Policy. Review of Financial Economics, Vol. 15, No. 1, pp. 1–18.Baker, M., & Wurgler, J. (2004). A Catering Theory of Dividends. The Journal of Finance, Vol. 59, No. 3, pp. 1125–1165.Bisschop, B. (2014). The Impact Of A Financial Crisis On The Dividend Payout Of Dutch Publicly Listed Firms. Bachelor Thesis, Faculty of Management and Governance, University of Twente, Available Online: [Accessed 15 April 2015].Black, F. (1976). The Dividend Puzzle. Journal of Portfolio Management, Vol. 2, pp. 5-8.Black, F., & Scholes, M. (1974). The Effects of Dividend Yield and Dividend Policy on Common Stock Prices and Returns. Journal of Financial Economics, Vol. 1, No. 1, pp. 1-22.Braunerhjelm, P., Djerf, O., Frisén, H., & Ohlsson, H. (2009). Finanskrisens effekter p? svensk industri. Industrins Ekonomiska R?d, Available Online: $File/IER%20-%20Finanskrisens%20effekter%20p%C3%A5%20svensk%20industri%20fullst%C3%A4ndig.pdf [Accessed 19 April 2015].Brav, A., Graham, J. R., Harvey, C. R., & Michaely, R. (2005). Payout policy in the 21st century. Journal of Financial Economics, Vol. 77, No. 3, pp. 483–527.Brooks, C. (2008). Introductory Econometrics for Finance. 2nd edn. New York: Cambridge University Press.Bryman, A., & Bell, E. (2011). Business Research Methods. 3rd edn. Oxford : Oxford University Press.Campbell, J. and Shiller, R. (1988). Stock Prices, Earnings and Expected Dividends. The Journal of Finance, Vol. 43, No. 3, pp. 661-676.Clogg, C. C., Petkova, E., & Haritou, A. (1995). Statistical Methods for Comparing Regression Coefficients Between Models. American Journal of Sociology, Vol. 100, No. 5, pp. 1261-1293Damodaran, A. (2010). Applied Corporate Finance: A User’s Manual. 3rd edn. New York: Wiley, John & SonsDeAngelo, H., & DeAngelo, L. (2006). Payout Policy Pedagogy: What Matters and Why. European Financial Management, Vol. 13, No. 1, pp. 11-27.DeAngelo, H., DeAngelo, L., & Skinner, D. J. (2009). Corporate Payout Policy. Foundations and Trends in Finance, Vol. 3, No. 2–3, pp. 95–287.Fama, E. F. (1974). The Empirical Relationships Between the Dividend and Investment Decisions of Firms. American Economic Review, Vol. 164, No. 3, pp. 304-318.Fama, E. F., & French, K. R. (1988). Dividend yields and expected stock returns. Journal of Financial Economics, Vol. 22, No. 1, pp. 3-25.Fama, E. F., & French, K. R. (198). Business Conditions and Expected Returns on Stocks and Bonds. Journal of Financial Economics, Vol. 25, No. 1, pp. 23-49.Fama, E. F., & French, K. R. (2001). Disappearing dividends: changing firm characteristics or lower propensity to pay?. Journal of Financial Economics, Vol. 60, No. 1, pp. 3-43.Gordon, M. J. (1959). Dividends, Earnings, and Stock Prices. The Review of Economics and Statistics, Vol. 41, No. 2, pp. 99-105.Gordon, M. J. (1962). The Savings Investment and Valuation of a Corporation. The Review of Economics and Statistics, Vol. 44, No.1, pp. 37-51Gupta. A., & Banga. C. (2010). The Determinants of Corporate Dividend Policy. Decision, Vol. 37, No.2, pp. 63-77Hellstr?m, G., & Inagambaev, G. (2012). Determinants of Dividend Payout Ratios: A Study of Swedish Large and Medium Caps. Master Thesis, Ume? School of Business and Economics, Ume? University, Available Online: [Accessed 18 April 2015].Holder, M. E., Langrehr, F. W., & Hexter, L. J. (1998). Dividend Policy Determinants: An Investigation of the Influences of Stakeholder Theory. Financial Management, Vol. 27, No. 3, pp. 73-82Jensen, M. C. (1986). Agency Cost of Free Cash Flow, Corporate Finance and Takeovers. The American Economic Review, Vol. 76, No. 2, pp. 326-329.Jensen, M. C., & Meckling, W. H. (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, Vol. 3, No. 4, pp. 305-360.Lintner, J. (1956). Distribution of Incomes of Corporations Among Dividends, Retained Earnings, and Taxes. The American Economic Review, Vol. 46, No. 2, pp. 97-113.Liu, J., Nissim, D., & Thomas, J. (2002). Equity Valuation Using Multiples. Journal of Accounting Research, Vol. 40, No. 1, pp. 135–172.Lloyd, W. P., Jahera, J. S., & Page, D. E. (1985). Agency Costs and Dividend Payout Ratios. Quarterly Journal of Business and Economics, Vol. 24, No. 3, pp. 19–29. McManus , I. D., Gwilym , O. A., & Thomas , S. (2004). The Role of Payout Ratio in the Relationship between Stock Returns and Dividend Yield. Journal of Business Finance & Accounting, Vol. 31, No. 9-10, pp. 1355-1387.Miller, M. H. & Modigliani, F. (1961). Dividend Policy, Growth, and the Valuation of Shares. The Journal of Business, Vol. 34, No. 4, pp. 411-433.Miller, M. H., & Scholes, M. S. (1978). Dividends and Taxes. Journal of Financial Economics, Vol.6, No.4, pp. 333–364.Myers, S. C., & Majluf, N. S. (1984). Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have. Journal of Financial Economics, Vol. 13, No. 2, pp. 187–221.Osobov, I. & Denis, D. J. (2007). Why Do Firms Pay Dividends? International Evidence on the Determinants of Dividend Policy. Available at SSRN: [Accessed 14 April 2015].Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. (1998). Using The Correct Statistical Test For The Equality of Regression Coefficients. Criminology, Vol. 36, No. 4, pp. 859–866.Penman, S. (2010). Financial Statement Analysis and Security Valuation. 4th edn. New York: McGraw-Hill Education.Pyles, M. K. (2014), Applied Corporate Finance: Questions, Problems and Making Decisions in the Real World, [e-book]. New York, NY: Springer New York, Available through: LUSEM University Library website [Accessed 10 April 2015]Purmessur, R. D., & Boodhoo, (2009). Signalling Power of Dividend on Firms’ Future Profits: A Literature Review. International Interdisciplinary Journal.Riksbank (2009). ?berg: Sverige och Finanskrisen, Speech by Svante ?berg, Available Online: [Accessed 21 April 2015].Rozeff , M. S. (1982). Growth, Beta and Agency Costs as Determinants of Dividend Payout Ratios. Journal of Financial Research, Vol. 5, No. 3, pp. 249-259.Shin, H.-H., & Stulz, R. (2000). Firm Value, Risk, and Growth Opportunities. Dice Center Working Paper No. 2000-8. Available at SSRN: [Accessed 20 April 2015]. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download