The Relevance of Dividend Announcements - Research Paper Example

? Event Study - Using Econometrics Introduction Announcement of dividends acts as a signal originated by managers of different firm which differ with changes in economic conditions. The investors, in response, interpret these signals in different ways depending on the aspects of the dividend phenomenon. It would be of interest to investigate whether the behavior of investors when dividends are announced entails ‘surprise’ or whether their behavior has any relationship with earnings. Some of the researchers who examined a similar subject include Lonie et al. (1996), who studied the relationship between dividend announcements and earnings, both of which appear simultaneously. As such, in practice, investment analysts who analyze long-term economic prospects are to be expected to inquire about information related to top management as well as changes in corporate strategy among other issues that are believed to have an impact on the company’s earnings (Denis, 1990; Marsh, 1993). These considerations are also very important in share trading decisions. One of the most anticipated types of regular returns on stocks are dividends. These are payments made by firms to their shareholders as a compensation for their capital. If dividend policy is frequently adjusted, then the variation in the stock prices would be expected to remain high. Variation in share prices occurs due to recurrent investors’ reactions ahead of important announcement by the companies, dividends announcement being among them. Such announcements are floated through the stock exchanges as well as made through annual reports for the purpose of reaching the investors. Nevertheless, some companies release crucial information through the press release even before they publish them in the annual reports, if the need arises. Furthermore, there are other mediums that transmit information to the investors in effect changing the prices of the stocks. Some of the announcement that may influence prices of shares include new product launch, management changeover, changes of government policies, starting and shutting down of projects, mergers and acquisitions, political changes, downsizing of firms, as well as the announcement of dividend, which is the focus of this study. The prices of stocks tend to change immediately the news hinting about these changes is received by the investors. The relevance of dividend announcements is a subject of considerable interest for many researchers in the field of finance. This subject has ignited considerable controversy, considering the impact dividends announcement have before and after their announcement. This study builds upon different dividend literature by looking at the impact of dividend announcements for a multiple of firms that are listed in UK stock markets. The study will use an event study approach to examine abnormal returns for a total of 50 firms, which is potentially caused by the announcement of dividends for 21 days spreading between July 2012 and August 2012. The results indicate that the announcement of dividends that is accompanied with good news is received positively by investors; hence, more rate of increase of abnormal returns are experienced after the announcement. The results of this study also indicate that firms should be more prudent when making dividend policies as they have significant economic implications. This paper looks at the reactions of markets following a blend of dividend announcement by companies from the UK. These announcements are pointers that are originated by the managers of different companies in conditions of economic uncertainty, a scenario that is accompanied by varied investors interpretations and informational asymmetry. This study will be in accordance with the principles adopted in any empirical studies that investigate the impacts of share prices in relation to changes in dividends (for example, Charest 1978; Eades et al., 1985). Literature Review From the study of Ghosh and Woolridge (1988), it has been established that markets can respond both favorably and unfavorably towards dividend announcements depending on a variety of factors. Apparently, following the announcement of dividends, the effect on the markets can be a result of earnings and abnormal share returns as well as a multiple of other events that are hard to explain. Some authors such as Chowdhury and Miles (1987) and Eades et al. (1985) try to explain these events in different ways, especially focusing on dividends and omissions. Others, such as Christie (1990) and Dielman and Oppenheimer (1984), have added the unusual timing of dividend announcements as well as special rather than normal dividends. In the UK, however, cases of dividends initiations and omissions are rare events, and, hence, researchers have hardly studied such phenomena. Many researchers have adopted an event study approach to study the effects of dividend announcement. Even those who use different methods, also attribute the dividend announcement to a market signal. Ofer and Siegel (1987) tested how share prices and earnings forecasts changes when dividend is changed unexpectedly; they demonstrated that the analysts adjusted their forecasts in the same direction with the sign of the resultant change. Elsewhere, Jeong (1992) carried out an econometric study by using cointegration technique to establish the symmetry correlation between earnings and dividends and confirmed that information asymmetry moves in the same direction with the constancy of dividend policy. Generally, the dividends per share becomes more stable when there is a higher extent of information asymmetry. In Easton’s (1991) study on the Australian market, the interaction between earnings and dividends were found to have some effect on share returns, providing an insight that the investors’ buying and selling decisions were also affected by the interaction of different signals. In addition, Liljeblom's (1989) study about shares listed on the Stockholm stock exchange confirmed that interaction of dividends announcement and earnings played some role following the announcement of stock splits and stock dividends. Theobalds (1978) used UK data to study intertemporal dividend models with the aim of finding out the stability of the structure of such models instead of signals in isolation. He established that factors such as the introduction of advanced corporation tax, dividend controls in operation, and the significant changes in the rate of inflation, contributed to the poor prediction in each of the models that were tested. Most of the empirical studies conducted in finance in respect to the relationship between economic events and the stock return presupposes efficiency of the stock market. On this assumption, it is hoped that prices of stocks do not respond rapidly to new or unbiased information. Although this theory is very important in finance, many researchers have conducted studies that have increasingly shown instances of inefficiency (e.g. Joy, Litzenberger and McEnally, 1977; Charest, 1978). The reaction of the market in response to the announcement of dividends is a perfect example of these scenarios. As such, if the market does not react efficiently to announcement of dividends and other regular events, then it would be tasking trying to explain the market reaction to incidents that are highly unpredictable. Charest (1978), a researcher who has studied the stock market’s reaction to the dividend announcement, found that the market’s reaction is slow. The meaning of his findings was that stocks earn abnormally high returns immediately after it has been announced that dividends will be increased and earn abnormally low returns immediately after it has been announced that dividends will be reduced – this indicates an opportunity of profit trading. Kalay and Loewenstein (1985) examined the trading effect of the announcement effect, by testing the ability of the market to form impartial anticipations. However, contrary to what would be expected from the theory of market efficiency, these authors found a proof of positive response to an announcement effect. Sample data and Methodology This study will be conducted by a traditional even-study technique whereby firm-specific dividend announcement events are compared with the stock market reactions (Brown and Warner, 1985; Strong, 1992). In such a study, the expected stock return is compared with the actual stock, which should be immediate to the announcement period of the dividend. This analysis helps establish if any stock market reaction takes place around this time. A market model will be used to determine the Average Abnormal Returns (AAR), which are used in the study. Then, the average of AAR of the whole event window (21 days) is used, and each company is assigned the following 3 groups of events: Change in AAR is more than 2.5%, would be Good news Change in AAR is less than -2.5%, would be Bad news Change in AAR is between 2.5%, and -2,5% would be NO news The data are collected from the FAME database, whereby the inclusion criteria is based on the following: A UK listed company in London Stock Exchange Active Companies With Annual Sales above ? 50M After selecting the companies that met all the above inclusion criteria, the researcher ended up with 561 companies, after which companies that did not account for data up to December 2011 were filtered out, hence ending up with 271 companies. Finally, 50 companies were randomly selected from the study. The hypothesis of the study is that there is a significant impact following announcement of dividends, on the abnormal stock returns of the companies being studied. The study tests the hypothesis through an event study approach. This approach measures the effects of dividend announcement or any other events of the firm and relating it to the firm’s security prices. The event chosen for this study is dividend announcement. The event windows taken for the purpose of this study spreads between 16 July 2012 to 13 August 2012. Event window is the total time period surrounding the event that has been taken as the major time frame to investigate the impact of the event in question. Therefore, the total of the event days including the date of announcement is 21 days. This means that the total event windows is split into two divisions, the first being the stock prices prior to the announcement of dividends and the second division being the stock prices after the dividends were announced. The event date, which is the date that the dividends were announced, is represented as t=0, and it is situated in the middle of the event window. The first part of the window is composed of 10 days of stock prices (-10) while the second one is composed of 10 days of stock prices (+10). The common investors are reached by the information regarding the dividend announcement through the stock exchange markets. Just as it happens with other abnormal and normal announcements, investors will have varied expectations towards the announcement of dividends. This is particularly because dividends are major constituents of investors earnings besides capital appreciation. Following change in investors’ expectations and confidence in a certain stock, the stock is likely to experience some changes, which can potentially cause changes in its abnormal returns. The hypothesis has, therefore, been designed along these relationships. The presence or absence of abnormal share performance will be determined through analysis of the daily share return data. The announcement period is marked as starting from day t-1 all the way to day t, where t is the day that the dividend was initially announced. To draw the general inferences for the event in question, the abnormal returns are aggregated. This aggregation is done through stocks and time. Therefore, the concept of Agreegare Abnormal Returns (AAR) is very important as it will help analysis in several periods and stocks within the event period. The test of the null hypothesis can then be tested given the AAR. The reason why aggregating is important is because tests with one event observation cannot have any relevance. Table 1 illustrates the ARR across 50 companies. Plots of the ARR are also included in Figure 1. The results from this study are largely congruent with what has been discussed from the literature in regards to the effect of the dividend announcement on stock earnings. The information revealed from this analysis can as well be useful for valuation of companies. Results and discussions Looking from the time of announcement; that is, announcement day 0, the percentage of good news is 6.899% Table 1: Good, Neutral and Bad groupings Event Day Good Neutral Bad -10 0.657% 0.134% -0.578% -9 0.745% 0.072% -1.000% -8 1.762% 0.008% -1.820% -7 2.555% 0.352% -1.379% -6 2.341% 0.227% 0.250% -5 3.098% 0.317% 0.419% -4 2.338% 0.385% 1.054% -3 2.738% 0.053% 0.467% -2 2.678% -0.160% -0.503% -1 2.189% 0.108% -0.739% 0 6.899% -0.048% -5.430% 1 8.444% 0.025% -6.356% 2 8.923% 0.329% -7.523% 3 9.127% 0.160% -7.595% 4 9.231% -0.084% -7.331% 5 9.947% -0.089% -8.138% 6 10.835% 0.157% -7.605% 7 11.693% 0.124% -9.554% 8 11.871% 0.267% -10.556% 9 11.907% -0.061% -11.446% 10 12.426% -0.559% -11.558% Looking at the good news, which indicates that the investors received the highest amount of abnormal returns, it is clear that the percentage of abnormal returns increased continuously after the dividends are announced on day 0. However, for the events windows before the dividends are announced, the rate of increase in ARR is not only slower but also unpredictable. The ARR increased from 0.657% during event -10 to 2.189% during event -1, which is a difference of only 1.532%. This is significantly below the rate of change between event 1 and event 10, which recorded a whopping 3.982% (12.426%-8.444%). For the No News group, the figures seem to make no meaningful changes, which is an indication that the investors do not have any motivation to change their investment decisions whatsoever. For the bad new, the pattern seems to be directly opposite from the Good News. Under this group, the ARR changed from -0. 578% during the year -10 to -5.430% during the announcement date. This is a difference of -4.852%, which is quite substantial. For the period after the announcement of dividends, the ARR changed from -6.356% to -11.558%, which was a difference of -5.20%. This means that the rate of change after the announcement date is more than the period before the announcement date, which is also the case with the Good News. These patterns are more clear from Figure 1 below, which is a plot of the three groups of News against the Average Abnormal Return (ARR). From the graph, the chase group of No News appears to be moving constantly, right from the day -10 all the way to day +10. However, for the Good News group, the changes are substantial; the ARR is increasing at a slow rate between day -10 through day -1, but the rate of increase heightens from day 1 through day 10. Figure 1: a plot of Good, Neutral and Bad News against ARR The conclusions of use of Accumulated Abnormal Returns in the above model is congruent with different observations that have been made from different literatures. The ARR plot reveals that, somehow, the market slowly follows up the imminent announcement. The ARR of Good News gradually increases between day -10 and -1 while the ARR for the bad News gradually decreases over the same period. Following the announcement of dividends, the ARR is somewhat constant because the investors have already experienced an event that would cause them to adjust their investment behavior, as it would be expected. Conclusion The empirical study has demonstrated that post-dividend share announcement with Good News has a positive and significant impact on the stock prices. Apparently, all the companies that have been tested announced just about constant dividends every year based on the face value of the shares, hence the shareholders are already aware about the values of dividends, such that this factor does not have any substantial impact on the stock exchanges. Nevertheless, the amount of profits following declaration of dividends has a substantial impact on the prices of the stocks. Nevertheless, the changes observed in regards to the stock prices can also be contributed by other factors other than dividend announcements. References Brown, S.J. and Warner, J.B., 1985. Using daily stock returns: the case of event studies. Journal of Financial Economics, Vol. 14 No. 1, pp. 3-31. Charest, G., 1978. Dividend information, stock returns and market efficiency – II. Journal of Financial Economics, Vol. 6, pp. 297-330. Chowdhury, G. and Miles, D.K., 1987. An Empirical Model of Companies' Debt and Dividend Decisions: Evidence from Company Accounts Data, Bank of England Discussion Paper, No. 28. Christie, W.G., 1990. Dividend yield and expected returns. Journal of Financial Economics, Vol. 28 Nos 1/2, pp. 95-125. Denis, D.J., 1990. Defensive changes in corporate payout policy: share repurchases and special dividends. Journal of Finance, Vol. 45, pp. 1433-56. Dielman, T.E. and Oppenheimer, H.R., 1984. An examination of investor behavior during periods of large dividend changes, Journal of Financial and Quantitative Analysis, Vol. 19, pp. 197-216. Eades, K.M., Hess, P.J. and Kim. E.H., 1985. Market rationality and dividend announcements. Journal of Financial Economics, Vol. 14, pp. 581-604. Easton, S., 1991. Earnings and dividends: is there an interaction effect?. Journal of Business Finance and Accounting, Vol. 18, pp. 255-66 Ghosh, C. and Woolridge, J.R., 1988. An analysis of shareholder reaction to dividend cuts and omissions. Journal of Financial Research, Vol. 11, pp. 281-94. Jeong, J. , 1992. The effect of the information environment on the dividend payout policy. paper presented at the International Financial Management Conference, Boston, MA, 12-13 November. Joy, O. RLitzenberger, R. and McEnally, R., 1977. The adjustment of stock prices to announcements of unanticipated changes in quarterly earnings. Journal of Accounting Research, Autumn, PP. 207-225. Kalay, A. and U. Loewenstein, 1985. Predictable events and excess returns: The case of dividend announcements, Journal of Financial Economics Vol. 14, pp. 423-449. Liljeblom, E., 1989. The informational impact of announcements of stock dividends and stock splits. Journal of Business Finance and Accounting, Vol. 16, pp. 681-97. Lonie, A. A., Abeyratna, G., Power, D. and Sinclair, C., 1996. The stock market reaction to dividend announcements: A UK study of complex market signals. Journal of Economic Studies Vol. 23 No.1, pp. 32–52. Marsh, P., 1993. Dividend announcements and stock price performance. unpublished paper presented at the Financial Markets Group Conference on Dividends and Earnings Forecasting and Valuation, London School of Economics, 23 March. Ofer, A.R. and Siegel, D.R., 1987. Corporate financial policy, information, and market expectations: an empirical investigation of dividends. Journal of Finance, Vol. 42, pp. 889-911. Strong, N., 1992, Modelling abnormal returns: a review article. Journal of Business Finance and Accounting, Vol. 19, pp. 533-53. Theobald, M., 1978. Intertemporal dividend models - an empirical analysis using recent UK data. Journal of Business Finance and Accounting, Vol. 5, pp. 123-35 Read More