Dividend Announcements and Shared Prices in Australia - Essay Example

Introduction Information signaling has long been explored by financial analysts. Many have tried to determine the effects of certain information on the stock prices of shares. Some have analyzed the effects of earnings announcements on the stock prices while other practitioners have studied the effects of share repurchase announcements on the share prices. Also, analysts have long been interested in the effects of dividend announcements on share prices. Such studies have been conducted to test the validity of the information signaling hypothesis when applied to stock or asset prices. The basic principle of the information signaling hypothesis is that when companies announce earnings, dividend, or share repurchase plans, they convey certain information or 'signals' to the market regarding the current and future value of their shares. The term signaling is commonly used in economic analyses. It refers to "the idea that one party (termed the agent) conveys some meaningful information about itself to another party (the principal)." ("Signaling", 2006) Also, signaling is based on the principles of asymmetric information. This says that, "In some economic transactions, inequalities in access to information upset the normal market for the exchange of goods and services." ("Signaling", 2006) In such a situation the signaling hypothesis says that, two parties could get around the problem of asymmetric information by having one party send a signal that would reveal some piece of relevant information to the other party. As mentioned, dividend announcements are one way by which information is conveyed to investors as well as to the market. The information content of dividend announcements has long been explored by various researchers and analysts. Moreover, various studies have been conducted to determine how dividend announcements affect the price of the shares of the company in the market. Different theories have been created to explain how certain factors affect the information content of dividend announcements and how effect that such information may have on the stock prices. The subject of the information content of dividend announcements warrants research for it affects several parties. For one, investors and shareholders are directly influenced by the stock prices of shares. Also, previous research has suggested that when dividend announcements are made, abnormal returns are seen especially during the period surrounding the announcement. (Starks, 2004) In this paper, one area of dividend announcements will be explored. Particular focus will be given on the size effect or the effect of the size of the firm on the abnormal returns that are seen when dividend announcements are made. This research will concentrate on the size effect as made evident in the Australian stock market. By conducting such analysis, this research will be able to answer the following research question: What is the effect of dividend announcements on asset prices Theoretical Framework This research will follow the methodology utilized by Mozes and Rapaccioli (1995). Their study aimed to determine the role of dividends in explaining the size effect. The said study is a follow up on past researchers that have been conducted wherein it was discovered that, on average, the firm's security price increases around its announcement of an increase in dividends or a special dividend. Moreover, previous researches have shown that the converse likewise holds. This means that a firm's security prices tens to decrease when an announcement of a decrease in dividends or a discontinued dividend is made. Also, past studies have shown that reactions to dividend changes are greater for small firms than for large firms. Using empirical analysis, the study of Mozes and Rapaccioli (1995) aimed to investigate "the extent to which dividend announcements affect the relation between firm size and the amount of information provided by earnings announcements." However, since this research does not concern itself with earnings announcements, the methodology utilized by the said authors will be slightly altered. Instead of considering earnings announcements, this paper will concentrate on determining the effect of dividend announcements on small and large firms. Particular focus will be given on the share prices of the firms when dividends are announced and given. Sample The sample for this study is composed of the top 100 Australian firms. Data on the sample was gathered from DataStream. Such data includes the market value, share prices, traded volume, and dividends paid for each of the firms between the August 2, 1991 and August 4, 2006. However, this study will concentrate on the time between 1991 and 1995. Moreover, the classification of the firms as large or small is dependent on the median market value of the whole sample. The firms whose market value is less than 5020.26, the median market value, are classified as small firms while those whose market value is greater than or equal to 5020.26 are classified as large firms. There were 69 firms that paid dividends at least once in the said period. Of the said number, 37 are classified as large firms while 32 are small firms. The table below provides descriptive statistical data on the sample: TOTAL NO. OF FIRMS: 100 LARGE FIRMS SMALL FIRMS TOTAL PAYS DIVIDENDS 37 32 69 DOES NOT PAY DIVIDENDS 13 18 31 Furthermore, the listed below are the total number of dividend payouts made in each year in the time period being analyzed. Methodology Following closely the methodology used by Mozes and Rapaccioli, this study analyzes the share price increases or decreases in relation to dividend announcements. Specifically, this study will concentrate on analyzing of the firm size in relation to the effect of dividend announcements on the share prices. The hypothesis to be tested for this study is states as follows: H1: Firm size is inversely related to the effect of dividend announcements on share prices. H1 will be supported if the increase in share prices for smaller firms that give out dividend announcements will be significantly greater than the increase in share prices of larger firms that carry out dividend announcements. The implication of this hypothesis follows that which Mozes and Rapaccioli (1995) suggests, that dividend announcements of larger firms have lesser impact on share prices because investors have other sources of information other than the information conveyed by the dividend announcements. Specific statistical tests are used to determine if the difference in the effect of dividend announcements for small and large firms will be significant. Analysis and Results Share prices for each of the 100 firms were gathered and were then subjected to analyses using correlation and F-Tests. Significance level was set at 0.001. The share prices of the 69 dividend-paying firms were gathered beginning with 4 weeks prior to the announcement. This research followed the model utilized by Mozes and Rapaccioli. (1995) The market model which uses ordinary least squares regression on the weekly returns is states as follows: Rit= ai + biRmt (1) Where; i = firm index; t = week index; Rit = the return for firm i on week t; and Rmt = the return on the market portfolio for week t. From such model, the unexpected return, which refers to the abnormal returns that are expected due to the dividend announcement, is computed using the following equation: Uit = Rit - (ai + bi Rmt) (2) Since the concern of this research is not the increase of share prices when increase in dividends are announced or the decrease in share prices when a decrease in dividends are announced but the change in share prices with the announcement of dividends, the value of the unexpected returns is squared following the methodology employed by Mozes and Rapaccioli (1995). This leads to a standardized index (SI) which is equivalent to the square residual during the announcement week divided by the average square residual of the weeks leading to the announcement. From this point, the size effect is measured through the estimation of the following coefficients, a and b, in the equation: SIt = a + bi SIZEt (3) SIZEt is the natural log of the firm's equity value. From the said equation, it was found that the size effect does exist in the sample. bi was estimated at -0.082 with a significance level grater than 0.001. Furthermore, following Mozes and Rapaccioli's (1995) example, the mean SI for the sample was found to be 1.28. This value indicates that more information is conveyed to investors and to the market during the announcement period rather than during the weeks that precede it. This means that the announcement of dividends does, in fact, affect share prices. To test the aforementioned hypothesis, the following equation was utilized. SIt = a + b1EARLYt + b2 SIZEDEt + b3SIZEENDt (4) Since this research was only concerned with the size effect and its relevance to dividend announcements, only the coefficient b2 was measured. Based on the model of Mozes and Rapaccioli (1995), SIZEDE is the log of the firm's equity value. In essence, when one disregards the other valuables, the equation used is equation (3) stated earlier as: SIt = a + bi SIZEt As the earlier analysis suggested, the size effect is indeed present. This is further proven by the fact that the difference between the change in share prices of small firms and large firms that make dividend announcements is significantly greater than the value of 0.001 which was earlier set. In essence, the hypothesis, H1, is accepted. Furthermore, it can be stated that firm size is inversely related to the change in share prices due to dividend announcements. To put is simply, smaller firms experience greater change in share prices when dividend announcements are made. Just like previous researches, one can assume that the underlying reason as to why such effect exists is because investors can make use of other sources of information concerning large firms. Such other sources of information can dampen the effect of the dividend announcement on share prices. Summary and Conclusion This research aimed to determine whether the size effect is present in the top 100 Australian firms specifically with regards to the change in share prices when dividend announcements are made. Furthermore, this research had the aim of determining what the effect was of dividend announcements on share prices are. The model and methodology utilized by Mozes and Rapaccioli (1995) were closely followed in this research. Dividend announcements convey information regarding the status of operations of firms to the market. These affect the share prices owing to the signaling hypothesis discussed earlier. Such theory was shown and expanded in this research. Moreover, this study was able to consider such theory in relation to the size effect. Through the analysis conducted, it was found that dividend announcements do in fact affect share prices. Significant changes in the share prices were observed when dividend announcements were initiated. Furthermore, this research found that the size effect is indeed present in the top 100 Australian firms. Specifically, it was found that small firms experienced greater abnormal returns when they made dividend announcements. Although large firms experienced some form of abnormal return, it was observed that such abnormal return was significantly less than that of small firms. The findings of this study warrants further research for several reasons. For one, only the top 100 Australian firms were used as part of the sample. It is suggested that future researches should analyze all the Australian firms so as to be able to provide more conclusive results. Also, since this study did not concern itself with the specific effect of the type of dividend announcement on the share prices, future research may consider determining what effect specific announcements have on share prices. For instance, one can study how increase in dividends may affect share prices or how discontinued or decreased dividends may affect share prices. Bibliography Mozes, H & Rapaccioli, D. "The Relation among dividend policy, firm size, and the information content of earnings announcements", The Journal of Financial Research, Spring 1995, Vol XVIII, No 1, pp 75 - 88. Signaling. Wikipedia, The Free Encyclopedia. 2006. Retrieved 30 Sept 2006 from: http://www.wikipedia.org/. Starks, L. & Yoon, P. The Change in the Information Content of Dividend Announcements: Abnormal Returns and Operating Performance. March 2004. University of Texas at Austin. [online]. 29 Sept 2006 from: http://www.fma.org/Stockholm/Papers/informationcontentdividend.pdf. Read More