1 Bachelor Thesis Economics & Business Economics
Faculty of Economics and Business
Is there symmetry in the effect on stock prices of dividend increases and decreases?
An investigation into the change in share prices around announced dividend raises and cuts in AEX listed firms
Author: Yannick G. Buijs Student number: 11580070
Supervisor: Drs. P.V. Trietsch, M.Phil.
Specialization: Finance
Date: 30/06/2021
2
Statement of Originality
This document is written by Yannick Buijs, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
UvA Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
This paper investigates the short-term effects of dividend announcements on share prices on the Dutch stock market from 2016 to 2020. The sample consists of 120 dividend increases and 25 dividend decreases of 25 AEX listed firms from 1 January 2016 until 31 December 2020. The benchmark used to calculate abnormal returns of a single stock was the Single Index Model, using the AEX index as market index. This study found that there are significant positive cumulative average abnormal returns in a [-3; +3] event window of 0.62% when a dividend was raised. On the contrary, the negative Cumulative Average Abnormal Returns for dividend cuts during the same event window were -4.29%. This gives reason to believe that dividend announcements, especially negatives ones, contain substantial information that lead to abnormal returns.
4
Table of Contents
Statement of Originality ... 2
Abstract ... 3
1. Introduction ... 5
2. Theoretical framework ... 7
2.1 Dividend announcement effect ... 7
2.2 Dividend theories ... 8
2.2.1 Dividend relevance theory ... 8
2.2.2 Signaling theory (information content of dividends) ... 9
2.2.3 Catering theory ... 10
2.2.4 Free Cash Flow Theory ... 11
2.3 Research questions ... 12
2.4 Hypotheses ... 13
2.5 Previous empirical research ... 13
3. Data and methodology ... 15
3.1 Data and sample selection ... 15
3.2 Methodology ... 15
3.2.1 Estimation and event window ... 15
3.2.2 Model ... 16
4. Results ... 18
4.1 Results ... 18
5. Conclusions and limitations ... 21
5.1 Conclusions ... 21
5.2 Limitations ... 21
5.3 Recommendations for further research ... 22
Bibliography ... 23
5
1. Introduction
The effect of dividend announcements on stock prices has been studied for decades, for example by Miller and Modigliani in 1961, yet there are still many missing pieces to the dividend puzzle. In his famous paper, Black (1976) said about the dividend puzzle: “The harder we look at the dividend picture, the more it seems like a puzzle, with pieces that just don’t fit together.” Almost half a century after Black’s 1976 paper this is still the case. Even though more research on the topic has been done, different results from this research have been obtained. However, despite the uncertain implications of dividends, all public companies worldwide still paid out US$1.255 trillion in dividends in 2020 (Henderson, 2021). But worldwide dividends have decreased by 12.1% from a record high US$1.429 trillion in dividends in 2019. Dividends have thus not been stable in recent years, which means that companies have had to make announcements concerning their changing dividend payments. Such announcements often lead to changes in the share price of a company as shown by Pettit (1972), Aharony and Swary (1980), Michaely et al. (1995), among others.
To determine whether asymmetry exists in the reaction to dividend announcements, both dividends raises and cuts are investigated. The ensuing change in stock price for both
announcements will help determine whether investors react stronger to one than to the other.
This paper investigates whether the share prices of firms react differently in the short-run to dividend raises as opposed to dividend cuts. This is done by conducting an event study around dividend announcements, similarly to MacKinlay (1997). The Cumulative Average Abnormal Returns (CAAR) and Average Abnormal Returns (AAR) for both dividend raises and cuts will be determined in the period from 1 January 2016 up to 31 December 2020 for the 25 companies that make up the Amsterdam Exchange Index (AEX). This has yielded a sample of a total of 145 dividend announcements, of which 120 were dividend raises and 25 were dividend cuts. The AEX index will be used as the benchmark index. A main event window of three days before until three days after the announcement is used. The
significance of the abnormal returns for both raises and cuts will be tested using a statistical t- test. Furthermore, to assess whether both a dividend raise and cut significantly differ from one another, a two-sample t-test is conducted.
The following section consists of the existing literature about the implications of
dividends and the asymmetry between dividend raises and cuts. The hypotheses of this paper will also be proposed in this section. Section 3 will elaborate on the data selected for this
6 study and the methodology used to conduct the research in this paper. The outcome of this research is shown in Section 4, where the results will be analysed. Finally, in Section 5, the final conclusions are drawn from the study.
7
2. Theoretical framework
This section will explain existing theories that help explain the effects of a dividend announcement. At first, there will be a thorough explanation of the dividend announcement effect. Once it has been made clear what the dividend announcement effect entails, this paper will elaborate on a couple well-studied theories about dividend announcements. These are the signalling theory, the catering theory, the dividend relevance theory and the free cash flow theory. Following the behavioural theories are the research questions that this thesis tries to answer. Then the hypotheses of this paper are this discussed. This section concludes with an overview of previous empirical research.
The following conceptual framework (Figure 1) is a visual representation of all the theories about what dividend announcements signal. It also shows which problems lead to these theories. It also shows two other variables that have an effect on stock prices, namely firm earnings and macro-economic factors.
Figure 1
2.1 Dividend announcement effect
This paragraph will explain in further detail what the dividend announcement effect means, followed by an explanation of how the effect is measured in this study.
When a firm wants to pay a dividend to their shareholders, this has to be made public through an official announcement. There are multiple forms of dividends, but this study will
8 solely focus on the most common type: cash dividends. When a firm declares that they will raise or cut their dividend, this will lead to a change in their share price (Ball and Brown, 1968). This change in share price is known as the dividend announcement effect. An increased dividend is often followed by a higher share price (Laabs, 2013). This could have multiple reasons, which will be elaborated upon in the “Dividend theories” section. Firms’
dividends are sometimes cut as well, which leads to a negative security price reaction (Spangler, 1973).
Dividends can also be completely omitted, or be paid for the first time: a dividend
initiation. Michaely et al. (1995) found that dividend omissions yielded a three-day return of - 7.0%, whereas initiations resulted in a 3.4% increase in share price after three days on the NYSE and AMEX. Healy and Palepu (1988) found a significant average 2-day return of 3.9% for dividend initiating firms, and a -9.5% 2-day return for firms that omitted dividends.
This indicates that investors react differently to a dividend initiation compared to a dividend omission. Dividend initiations and raises are often responded too with an increase in share price, whereas dividend omissions and decreases lead to a lower share price than before.
2.2 Dividend theories
The following subsections will consist of the multiple theories that try to explain what a dividend announcement really implies.
2.2.1 Dividend relevance theory
The dividend relevance theory was developed by Gordon (1963) and Lintner (1964). The dividend relevance theory suggests that a firm’s dividend policy does have an effect on its share price (Walter, 1963). The theory states that investors are actually not indifferent about profit distributions by firms. Investors tend to be risk averse and thus would prefer to receive dividends today over dividends tomorrow or a possible higher future share price (Gordon, 1963). Dividend payments are considered more certain than potential future capital gains.
Bhattacharya (1979) states that investors are more likely to desire dividends when their planning horizon is shorter. They have a higher “urgency” to realize wealth for consumption, through the accumulation of dividend payments. This means that there is a higher demand for dividend payments when investors have a short investment horizon. Contrastingly, investors
9 with a long(er) investment horizon would prefer profits to be reinvested in the firm, leading to a higher share price in the future.
Research by DeAngelo and DeAngelo (2006) suggests that payout policy is not irrelevant. They note that Miller and Modigliani (1961) assume a payout of all free cash flows, thus restricting a mix between some dividend payment and some cash retention. When all options are available, payout (and investment) policy of a firm do affect shareholder wealth (DeAngelo and DeAngelo, 2006). They found that this effect does not solely come forth from project choices or market imperfections.
A survey by Brav et al. (2005), where CFOs were asked about their motives and opinions of their payout policies, shows that managers find the level of dividend payout just as
important as new investment decisions. This was not the result for share repurchases, which were seen as a less important payout policy. This indicates that managers consider dividend payment more important than share repurchases.
2.2.2 Signaling theory (information content of dividends)
The signaling hypothesis states that management can inform its shareholders about the growth potential of the firm through dividends (Miller and Rock, 1985). In public firms there is so-called separation of ownership between the shareholders and the management. This leads to separation of knowledge, where the managers (insiders) might know more about what is going on inside the company than its shareholders (outsiders). Thus, dividends can be the solver of this problem by correcting this information asymmetry. Management can emit a credible signal of potential future growth of the firm to its shareholders by raising the
dividend (Bhattacharya, 1979). A survey conducted by Brav et al. (2005) shows that most managers themselves carry the belief that they convey information with their dividend announcements.
Pettit (1972) came to the conclusion that there was an information effect in dividend announcements, but very little to be found in earnings announcements. On the other hand, however, Watts (1973) concluded that dividend announcements contain very little
information once the effect of the earnings report was taken into account. This was also concluded by famous research conducted by Miller and Modigliani (1961). They did propose
10 that dividends, in fact, do contain certain information. This is because a change in a share’s dividend rate was often causing its market price to change. Miller and Modigliani, however, came to the conclusion that it was not merely the dividend change that caused the price change, but rather that the dividend change was a response to information concerning future profits.
Shareholders should still consider dividend announcements as important information, because of their close relationship to current and future profits of a firm (Lintner, 1956).
Black and Scholes (1974) also notice this information effect of changes in dividend. They state that when dividends are cut when the directors of a firm are expecting troubled times, but those troubled times do not happen, the effect of the declined stock price will only be temporary. Thus, if directors can credibly say that they are using the freed-up funds from lowering dividends for investments (with a positive Net Present Value), the stock price won’t be negatively affected by the dividend cut.
2.2.3 Catering theory
The catering theory, developed by Baker and Wurgler (2004), states that managers of firms cater to investor demand for dividends. The desire by investors to receive dividends, rather than share buybacks or firm (re)investments, changes from time to time. When
investors prefer dividend paying stocks over those that don’t, firm managers tend to increase (or even initiate) dividend payments. However, when investors don’t value dividend
payments very highly, rational acting managers tend to decrease (or even omit) dividend payments. The catering theory thus proposes that investor demands are a driving factor behind a firm’s decision to pay dividends.
A study by Fama and French (2001) found that in 1999 only 20.8% of U.S. firms pay dividends, compared to a peak in 1978 where two-thirds of all firms paid dividends. This decline in dividend-paying companies could partially be explained by an increasing number of listings of small, high growth firms. However, this does not tell the full story of why the amount of dividend-paying companies is declining. An important reason for why there are fewer dividend-paying firms, is that firms have become less likely to pay dividends, even when controlled for firm characteristics. It is suggested that the perceived benefits of dividends have declined in recent years. Firm managers notice this decreased demand for dividends and subsequently lower (or omit) their firm’s dividend. This could be viewed as a confirmation for the catering theory.
11 The catering theory gives reason to believe that there is a dividend premium on dividend- paying stocks as opposed to nonpayers (Baker and Wurgler, 2004). This premium means that dividends are an important part of a share price.
2.2.4 Free Cash Flow Theory
The Free Cash Flow Theory, also known as ‘the excess cash problem’, is a theory that aims to explain some of the information content of dividend announcements. The free cash flow hypothesis states that the announcement of a dividend increase will be followed by a positive abnormal return on its share price, whereas the contrary is expected to happen for a dividend cut. It combines the aforementioned signaling theory with the problem of agency costs.
Agency costs arise when an agent (corporate management) is in charge to make the decisions for a principal (a firm’s shareholders). This happens in large firms due to the separation of ownership. Rather than achieving the general objective of maximizing
shareholder value, decisions made by a manager with different interests could lead to agency costs for a firm. Mature research by Berle and Means (1932) already voiced some concerns that corporations might not be run primarily in the interests of its stockholders. Jensen (1986) states that most of the agency problems arise from managers that focus too much on
increasing the size of its firm, rather than the size of its profits. Some managers do this, because they might receive higher compensation from expanding the corporation. This is worrying for shareholders when a firm’s free cash flow is used by managers undertake suboptimal projects with a negative net present value (NPV). Thus, this so-called empire- building comes at the expense of the firm value.
The agency problems for a firm due to empire-building could be alleviated by issuing a higher dividend, which leads to less waste of free cash flow (Easterbrook, 1984). According to the overinvestment hypothesis, a higher dividend by a firm with free cash flow problems would then signal a reduction in wasted free cash flow on negative NPV projects (Lang and Litzenberger, 1989). This would subsequently lead to an increase in firm value (Jensen, 1986). Likewise, a decreased dividend could indicate that more negative NPV projects will be undertaken, followed by a decrease in share price (Park and Jang, 2013). This gives reason
12 to believe that dividend announcements can reduce agency costs related to the
overinvestment problem.
However, not all research on the free cash flow theory has found a significant relationship between dividend announcements and free cash flows. Chosiah et al. (2019) did not find significant results supporting the free cash flow hypothesis. Yoon and Starks (1995) even found results that contradict the free cash flow hypothesis. In the Netherlands, the dividend decisions do not hold a significant relationship with the intensity of agency problems (Renneboog and Szilagyi, 2006).
Firms that are most susceptible to the free cash flow problem are those with intermediate prior performance (Blau and Fuller, 2010). Intermediate firms that are expecting better performance in the future could signal this to the market with an increased dividend payment.
Firms that are perceived to be of high-quality pay dividends to eliminate their relatively high free cash flow problem, since they have a high free cash flow that could be used for
investments. Because they are already perceived to be of high quality, they are not concerned about signaling information through dividend announcements. Firms with the lowest
historical performance naturally have little free cash flow, and thus this does not lead to free cash flow problem. Intermediate performing firms pay dividends to solve their free cash flow problem and use dividend raise announcements to signal higher future performance compared to firms with similar performance in recent years.
2.3 Research questions
The central research question in this paper is:
Is the dividend announcement effect asymmetrical?
This will be investigated in the short-run by comparing the share price reaction after a positive dividend announcement (dividend raise) to the reaction of a share price after a negative dividend announcement (dividend cut).
In the tests that will be conducted in the next section, a thorough explanation of how the dividend announcement effect is measured will follow. Subsequently, an interpretation of the results will answer the questions why a dividend raise likely leads to positive CAAR. On the other hand, an explanation for the abnormal decrease in stock price after a lowering of dividends will be sought.
13
2.4 Hypotheses
Taking into account all previously conducted research with regards to the consequences on stock prices after dividend announcements, this paper will test the following hypotheses.
Hypothesis 1: A higher dividend will yield positive cumulative abnormal returns in the short-run.
Hypothesis 2: A dividend cut will lead to negative cumulative abnormal returns in the short-run.
Hypothesis 3: Announcing dividend raises will lead to a higher cumulative abnormal return than the announcement of a dividend cut.
2.5 Previous empirical research
This section shows an overview of the literature relevant for this research.
Author Year Country Variables Conclusion
Pettit (1972) 1964 - 1968 United States • Earnings performance
• Daily share prices
• Mutually exclusive dividend classes
Information effect in dividend announcements, but no
information in earnings.
Renneboog and Szilagyi (2006)
1996 – 2004 The Netherlands • Dividends
• Firm value
• Book value of assets
• Net income
• Cash flow
Dutch firms pay low and
moderately smoothed dividends, based on operating cash flows.
Watts (1973) 1947 - 1966 United States • Regular dividend
Information in earnings reports,
14
• Special dividends
• Expected future earnings
• Unexpected changes in dividends and earnings
but very little in dividend
announcements.
Laabs (2013) 2005 - 2013 United States • Risk-adjusted return of stock price
• S&P 500
• Alpha
• Beta
Significant positive market reaction prior to a firm’s increased dividend
announcement.
Michaely et al.
(1995)
1964 - 1988 United States • Dividend omissions
Omissions lead to -7% return in three days.
Initiations lead to +3.4%.
Figure 2
15
3. Data and methodology
This chapter explains which data and methodology was used for testing the asymmetry in Dutch dividend announcements. The type of study in this paper is an event study, which was founded by Ball and Brown (1968).
3.1 Data and sample selection
The data that was used for this study comes from the Wharton Research Data Services (WRDS). Here we can retrieve data of AEX listed companies from the dates when dividend announcements were made (declaration date), including the daily stock price. The AEX index return is used as the market benchmark.
The sample consists of the 25 firms that were listed on the AEX from 1 January 2016 until 31 December 2020. This will result in a sufficiently high amount of announced dividend raises and cuts to conduct parametric tests. Special dividends and stock-only dividends are not included in the sample. What should also be noted is that the composition of the AEX changes of the years. Only firms that were part of the AEX index on the time of the
announcement of their changing dividend are taken into account. Thus, over the investigated five years 34 different companies were researched.
3.2 Methodology
This section describes the methodology used for conducting this research. It will
elaborate on the chosen estimation window and event window. Furthermore, the model used for determining the Cumulative (Average) Abnormal Returns is introduced.
3.2.1 Estimation and event window
This paper will use an estimation window that starts 120 days before the observed
dividend announcement, which is deemed a sufficient estimator of a share price under normal circumstances (MacKinlay, 1997). The main estimation window closes 3 days before the change in dividend is made public. The main event window utilized in this research starts 3 days before and finishes 3 days after a dividend announcement, as can be seen in Figure 3.
This event window captures the hypothesized abnormal returns in the short-run.
16 Figure 3
3.2.2
Model
To determine whether dividend announcements have an effect on share price, the
abnormal return of a stock should be calculated. To calculate the abnormal return of a stock, a market model is needed. This study opts for the use of the Single Index Market Model
(SIMM). This single index market model is considered to be just as strong as a multi-factor model (MacKinlay, 1997). The SIMM is defined as:
𝑅𝑖𝑡 = 𝛼𝑖 + 𝛽𝑖𝑅𝑚+ 𝜀𝑖𝑡 𝑅𝑖𝑡 = 𝑟𝑒𝑎𝑙𝑖𝑧𝑒𝑑 𝑟𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝑠𝑡𝑜𝑐𝑘 𝑖
𝛼𝑖 = 𝑎𝑙𝑝ℎ𝑎 (𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡 𝑡𝑒𝑟𝑚) 𝛽𝑖 = 𝑚𝑎𝑟𝑘𝑒𝑡 𝑏𝑒𝑡𝑎 𝑜𝑓 𝑠𝑡𝑜𝑐𝑘 𝑖 𝑅𝑚 = 𝑚𝑎𝑟𝑘𝑒𝑡 𝑟𝑒𝑡𝑢𝑟𝑛
𝜀𝑖𝑡 = 𝑒𝑟𝑟𝑜𝑟 𝑡𝑒𝑟𝑚 (𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑡𝑜 𝑒𝑞𝑢𝑎𝑙 0, 𝑤𝑖𝑡ℎ 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝜎𝜀2)
To account for all stocks in the sample, the cumulative abnormal return (CAR) is
calculated. This is the sum of all abnormal returns. More useful measures for this research are the average abnormal return (AAR) and cumulative average abnormal return (CAAR). Using the CAR, the CAAR is calculated for the previously mentioned event window of 3 days before until 3 days after the event of a dividend announcement.
The results of the CAAR in and of itself can’t prove or disprove the hypotheses of this paper. Hence, a statistical t-test is used. Assuming that CAAR residuals are normally distributed the used test statistic is t-distributed with n-2 degrees of freedom, where n is the number of observations in the sample. All t-tests will be one-sided. Dividend raises undergo right-tailed tests; cuts left-tailed. From the t-test a conclusion can be drawn whether the
17 (C)AAR after a dividend raise and a dividend cut are significantly different from zero.
Subsequently, conclusions on different significance levels (1%, 5% and 10%) can be drawn using the p-values that follow from the t-tests.
To determine whether the CAARs of a dividend raise is significantly larger than those of a dividend cut, a right-tailed two-sample t-test is used. This test is executed to determine whether the market’s response to a dividend raise is asymmetric to a dividend cut.
18
4. Results
This section will show the results that were obtained from the sample, using the previously described methodology. At first, the effects of a dividend raise and dividend cut will be looked at separately. This is followed by the two different types of
announcements that are compared to one another.
4.1 Results
To test hypothesis 1 and hypothesis 2, the CAAR around dividend raises and cuts were tested for significance. The main event window that was considered in this research started three days before the dividend announcement and finished three days thereafter. This generally resulted in a positive CAAR around an announcement of an increased dividend, thus supporting hypothesis 1. When a dividend was cut, the CAAR were negative. As can be seen in Table 1, the dividend cuts led to more significant results than the dividend raises. This supports hypothesis 2.
Event window [-3, +3]
Announcement n CAAR Std. error p-value
Raise 120 0.62492315 0.386741 0.0543853*
Cut 25 -4.287770084 1.695978 0.00922787***
Table 1. *p<0.10, **p<0.05, ***p<0.01 To further investigate the distinct reactions between differing announcements, the Average Abnormal Returns (AAR) per type of announcement was observed separately for every single day in the event window. There are very significant differences in AARs between a raise and a cut one day before and on the day of a dividend announcement, as shown in Table 2. The AAR for dividend cuts on the event day, however, are much larger than for raises, indicating asymmetry. Remarkably, three days before an announcement, stocks that will undergo dividend cuts yield positive AAR. On that same event day, stocks that are about to have a higher dividend tend to see a negative AAR.
19 Event day Dividend Raise
AAR (%)
Dividend Cut AAR (%)
Difference (percentage point)
P-value
-3 -0.04893326 0.32230707 -0.37124033 0.801101
-2 -0.10583108 -0.19301 0.08717892 0.438676
-1 0.052995 -1.13681 1.189805 0.00562382***
0 0.505531 -2.07142 2.576951 0.00921439***
+1 0.063099 -0.31768 0.380779 0.252696
+2 0.077430936 -1.40977 1.48720094 0.0559248*
+3 0.097557983 0.335429428 -0.237871445 0.631525
Table 2. *p<0.10, **p<0.05, ***p<0.01
To test whether announced dividend raises led to significantly higher CAAR than dividend cuts, the two samples were compared to each other. To see whether the results of the main event window [-3; +3] are robust, two more event windows ([-2; +2] and [-1; +1]) were observed. In Table 3 the results of all observed event windows are showing very significant results for each event window. Thus, there is reason to assume that dividend raise
announcements lead to higher CAAR than the announcements of dividend cuts. Furthermore, the CAAR of dividend cuts were much farther away from zero than the CAAR after raises, indicating asymmetry in the stock market’s reaction to the different types of announcements.
The largest difference in CAAR was observed in an event window of a total of five days.
This window starts two days before and ends two days after the announcement.
Event window Dividend Raise CAAR (%)
Dividend cut CAAR (%)
Difference (%-point)
p-value
[-3; +3] 0.62492315 -4.287770084 4.91269323 0.00443985***
[-2, +2] 0.576298427 -4.94551 5.52180843 0.00160331***
[-1; +1] 0.604698575 -3.34273334 3.94743192 0.00290819***
Table 3. *p<0.10, **p<0.05, ***p<0.01 To further test the robustness of each type of announcement, they were also tested separately over multiple event windows for significance. Table 4 shows the results of those tests.
Dividend cuts were statistically significantly lower than zero on a 99% confidence level for
20 all three event windows. On the other hand, dividend raises were significantly higher than zero, albeit on lower confidence levels. Dividend cuts thus tend to see stronger (negative) reactions after an announcement.
Announcement CAAR (%) Std. Error p-value Event window
Raise 0.624923 0.386741 0.0543853* [-3; +3]
Cut -4.287770084 1.695978 0.00922787*** [-3; +3]
Raise 0.576298427 0.342113121 0.0473506** [-2; +2]
Cut -4.94551 1.666259 0.00334730*** [-2; +2]
Raise 0.604699 0.275478 0.0150490** [-1; +1]
Cut -3.34273 1.28558 0.00784915*** [-1; +1]
Table 4. *p<0.10, **p<0.05, ***p<0.01
21
5. Conclusions and limitations
The last chapter of this thesis will conclude the results of this research. First come the conclusions that are derived from all the previous chapters. Thereafter the limitations of this paper are discussed. Lastly, recommendations for further research on dividend
announcements is given.
5.1 Conclusions
The aim of this paper was to investigate whether positive and negative dividend
announcements have a significant asymmetric impact on the share prices of AEX listed firms.
This paper has showed that dividend increases are met by significant abnormal increases of share prices in the short-run. Even more significant results were found for announcements of decreasing dividends. Dividend cuts led to very significant negative abnormal returns in the observed event windows. Thus, the findings of this study provide evidence for both
hypothesis 1 and 2. This could have implications for firms that are willing to change their dividends. Namely firms should be careful with increasing their dividend too much, when there’s a risk that they won’t sustain that level of dividends in the future. Considering the negative impact of a dividend cut, firms should try to avoid decreasing their dividends.
Furthermore, the CAARs of dividend raises were significantly more positive than for cuts in several observed event windows. This supports hypothesis 3 that announcing an increase in dividend will lead to larger positive abnormal returns than announcing a decreased dividend.
The negative CAARs after a dividend cut were much larger (further away from zero) and more significant than the positive CAAR (closer to zero) after a dividend raise. The stock price reaction to announced dividend cuts is far greater than to dividend raises. Dividend cuts are understood by investors as a signal of poor future performance of a firm. To conclude, there exists asymmetry between dividend cuts and raises, since dividend cuts lead to a much greater change in stock price than raises.
5.2 Limitations
One of the restrictions to this study is that a dividend change is rarely announced by itself.
Many a time dividend announcements are accompanied by earnings reports, which could influence the share price of a firm. Thus, it is very tough to fully isolate the effect of the
22 dividend announcement. Moreover, since perfect markets do not exist, share price
fluctuations are never completely explicable.
Another limitation of this study was the relatively small sample of observed dividend cuts (25) compared to dividend raises (120). This because firms are much more likely to increase rather than decrease their dividends.
5.3 Recommendations for further research
A recommendation for future research on the effect of dividend announcements on share prices is to include simultaneously announced earnings in the model to estimate abnormal returns. Financial analysts often publish expectations for both upcoming earnings and dividends. The discrepancy between the actual announced amount compared to the expected amount could be used to determine the announcement effect on the stock price.
23
Bibliography
Aharony, J., & Swary, I. (1980). Quarterly Dividend and Earnings Announcements and Stockholders' Returns: An Empirical Analysis. The Journal of Finance, 35(1), 1-12.
Ball, R., & Brown, P. (1968). An Empirical Evaluation of Accounting Income Numbers. Journal of Accounting Research, 6(2), 159-178.
Baker, M., & Wurgler, J. (2004). A catering theory of dividends. The Journal of finance, 59(3), 1125-1165.
Berle and Means (1932), Modern, 19, 67-85, 108-9; Gordon, "Ownership," 369-70; TNEC Report.
Bhattacharya, S. (1979). Imperfect information, dividend policy, and" the bird in the hand"
fallacy. The Bell Journal of Economics, 259-270.
Black, F. (1976), The dividend puzzle, The Journal of Portfolio Management Winter 1976, 2 (2) 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, 1(1), 1-22.
Brav, A., Graham, J. R., Harvey, C. R., & Michaely, R. (2005). Payout policy in the 21st century. Journal of financial economics, 77(3), 483-527.
Chosiah, C., Purwanto, B., & Ermawati, W. J. (2019). Dividend policy, investment opportunity set, free cash flow, and company performance: Indonesian’s agricultural sector. Jurnal Keuangan dan Perbankan, 23(3), 403-417.
Daniels, K., Shin, T. S., & Lee, C. F. (1997). The information content of dividend hypothesis:
A permanent income approach. International review of economics & finance, 6(1), 77-86.
DeAngelo, H., & DeAngelo, L. (2006). The irrelevance of the MM dividend irrelevance theorem. Journal of financial economics, 79(2), 293-315.
Dividend Announcements, Security Performance, and Capital Market Efficiency Author(s):
R. Richardson Pettit Source: The Journal of Finance, Dec., 1972, Vol. 27, No. 5 (Dec., 1972), pp. 993-1007 Published by: Wiley for the American Finance Association
Easterbrook, F. H. (1984). Two agency-cost explanations of dividends. The American economic review, 74(4), 650-659.
Fama, E. F., & French, K. R. (2001). DISAPPEARING DIVIDENDS: CHANGING FIRM CHARACTERISTICS OR LOWER PROPENSITY TO PAY? Journal of Applied Corporate Finance, 14(1), 67–79.
Fuller, K., & Blau, B. M. (2010). Signaling, free cash flow and “nonmonotonic”
dividends. Financial Review, 45(1), 21-56.
24 Gordon, M. J. (1963). Optimal investment and financing policy. The Journal of
finance, 18(2), 264-272.
Healy, P. M., & Palepu, K. G. (1988). Earnings information conveyed by dividend initiations and omissions. Journal of financial Economics, 21(2), 149-175.
Henderson, J. (n.d.). Global Dividend Index, Edition 30,
https://cdn.janushenderson.com/webdocs/Final+30th+Global+Report.pdf
Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. The American economic review, 76(2), 323-329.
Laabs, Douglas S., "THE IMPACT OF INCREASED DIVIDEND ANNOUNCEMENTS ON STOCK PRICE: A TEST OF MARKET EFFICIENCY" (2013). Theses, Dissertations &
Honors Papers. Paper 146.
Lang, L. H., & Litzenberger, R. H. (1989). Dividend announcements: Cash flow signaling vs.
free cash flow hypothesis?. Journal of financial economics, 24(1), 181-191.
Lintner, J. (1956). Distribution of incomes of corporations among dividends, retained earnings, and taxes. The American economic review, 46(2), 97-113.
MacKinlay, A. (1997). Event Studies in Economics and Finance. Journal of Economic Literature, 35(1), 13-39.
Michaely, R., Thaler, R., & Womack, K. (1995). Price Reactions to Dividend Initiations and Omissions: Overreaction or Drift? The Journal of Finance, 50(2), 573-608.
Miller, M. H., & Modigliani, F. (1961). Dividend policy, growth, and the valuation of shares. the Journal of Business, 34(4), 411-433.
Miller, M. H., & Rock, K. (1985). Dividend policy under asymmetric information. The Journal of finance, 40(4), 1031-1051.
Park, K., & Jang, S. S. (2013). Capital structure, free cash flow, diversification and firm performance: A holistic analysis. International Journal of Hospitality Management, 33, 51- 63.
Pettit, R. (1972). Dividend Announcements, Security Performance, and Capital Market Efficiency. The Journal of Finance, 27(5), 993-1007
Renneboog, L., & Szilagyi, P. G. (2006). How relevant is dividend policy under low shareholder protection?. Tilburg University.
Spangler, C. (1973), "The Effects of Unanticipated Changes in Dividends on Security Returns," unpublished Master of Science Thesis, MIT, Cambridge, Mass., 1973.
Walter, J. E. (1963). Dividend policy: its influence on the value of the enterprise. The Journal of finance, 18(2), 280-291.
25 Watts, R. (1973). The information content of dividends. The Journal of Business, 46(2), 191- 211.
Yoon, P. S., & Starks, L. T. (1995). Signaling, investment opportunities, and dividend announcements. The Review of Financial Studies, 8(4), 995-1018.