Second, it examines the trend in each stock's rate of return, which affects the number of days to reach the average highest CAR and the average lowest CAR of each cash dividend announcement. Third, it examines the industry effect on the number of days to reach the average maximum CAR and average minimum CAR of each cash dividend announcement. The study finds that the average number of days to average the highest or lowest CAR more quickly is a good event new due to the change in dividends.
They would know the average number of days for the average of the maximum cumulative abnormal return and the average of the minimum cumulative abnormal return to be reached after the cash dividend announcement. This article examines the average number of days from a good event to reaching the average of the maximum cumulative abnormal return, also known as 'peak'. And the average number of days from a bad event to reaching the average of the minimum cumulative abnormal return, called 'Bottom'.
DATA AND METHODOLOGY
Data Sources
News Classification
This table presents the number of cash dividend announcements by year which analysis covers the time period from January 1999 to December 2011. Studies conducted by Aharony and Swary (1980), Asquith and Mullins (1983) show that dividend growth is associated with a sound company that has healthy cash flows and strong future earnings. In developed markets it is generally observed that a positive dividend change (i.e. an increase in the dividend) generates an increase in the share price thus generating an abnormal (i.e.
An excess) return for the investor and a negative change in the dividend (i.e. a dividend cut) causes the share price to fall, again creating an abnormal return, but this time on the negative side for the investor. The dividend change for each company for each announcement date was calculated as follows: After calculating the change in the dividend, three possible results are observed, namely no change in the dividend, a positive change in the dividend and a negative change in the dividend.
No dividend change implies that the company has not increased or decreased its current dividend announced compared to the previous announcement date to shareholders, resulting in zero dividend change, excluded from the sample for calculation. Positive dividend change implies that the company has increased its dividend announced to the shareholders in the current announcement date compared to the previous announcement resulting in a positive increase in dividend change. Finally, negative dividend change indicates a decrease in the current dividend announced by the company compared to its announcement for the previous dividend that resulted in a negative dividend change.
Next, based on the results obtained from the above calculation, the sample is divided into these two groups as: positive change of dividend is good event and negative change of dividend is bad event which is shown in table 3.2 Summary statistics for sample data . This table presents the number of cash dividend announcement by dividend change effect from positive change as good event and negative change as bad event and by year the analysis covers the time period from January 1999 to December 2011.
Methodology
Next daily abnormal returns are calculated, I use market adjustment returns as previous study (Brown and Warner, (1985)). I calculate daily SET index returns by taking the difference between the index close on day t-1 and the index close on day t. There are no associated significant events within the event windows before and after the announcement period.
Where ARit is the abnormal return of stock i on day t Rit is the return of stock i on day t. Where CAR t1, t2 is the cumulative abnormal return for each firm between the period of day t1 and day t2. The top and bottom picks are taken from each event that has the maximum CAR and minimum CAR, respectively, along with the number of the ongoing event.
Then the average number of days it takes to reach the “peak” or “bottom” is calculated based on the average number of runs. Therefore, the average sequence number represents the number of days in the average to reach the maximum CAR and minimum CAR for each announcement. According to the result of good news is the average number of days to reach the peak, while the result of bad news is the average number of days to reach the bottom.
If the t-test is significant, it means that the null hypothesis will be rejected and there is a difference between both the LS mean of good news and the LS mean of bad news. Simple linear regression test is used to examine the linear relationship between the number of days reaching maximum CAR and minimum CAR and maximum CAR and minimum CAR.
The pre-event return effect is used to examine the trend of the rate of return of each stock impact on the number of days to reach the maximum CAR and minimum CAR of each cash dividend announcement. It is classified into 2 types as “positive trend” and “negative trend” based on the average return of each stock during the preceding event (event period -40 to -1) of each cash dividend announcement. Since the average return score is a positive sign, the trend of the rate of return of previous events will be classified as a positive trend.
In contrast to the sign of the negative result, the trend of the rate of return of previous events will be classified as a negative trend. Additionally, the trend results are based on the effect of dividend change, good event and bad event. So the result would be good event positive, good event negative, bad event positive and bad event negative compared to the peak and bottom result of table 4.1.
As described in the Peak and Bottom Score, the industry effect uses the industry ID as an index to examine the number of days to reach the peak and bottom of the maximum CAR and minimum CAR, respectively. The trading imbalance is calculated by subtracting the value of selling baht from the value of buying baht and dividing by the total value of buying and selling baht for each security type and investor.
DSELLDBUY
DSELL
Moreover, to study the trend of investor trading, the cumulative daily buy-sell imbalance (Timb) is essential as an indicator to find out the effect of different trading volumes at peak or bottom date after each cash dividend announcement.
EMPIRICAL RESULTS
- Peak and Bottom Result
- Regression Analysis
- Trend of Return during Pre-Event Period
- Industry Effect to Peak and Bottom
- Type of Investor Responding to PEAK and BOTTOM
This table shows the average number of days from good event and bad event reaching the average of maximum CAR (Peak) and minimum CAR (Bottom) respectively, and the result of LS average. Refer to the Table 4.1 Peak and bottom result, it is continued to investigate that the average number of days to reach the peak and bottom have the lining relationship. The positive and negative for pre-event trend result, it can indicate that the average number of days to reach the maximum CAR and minimum CAR is close to the average number of days of peak and bottom result.
The share price reaction to the cash dividend announcement was effected from the different industries which is shown in Table 4.4 Industry effect to peak and bottom. It spends an average of 18 days after the good event of cash dividend announcement to reach the average of maximum CAR at 13.2479%. It spends an average of 24 days after the good event of cash dividend announcement to reach the average of maximum CAR at 11.5166%.
It spends an average of 27 days after the bad event of cash dividend announcement to reach the average of minimum CAR at -12.7202%. It spends an average of 32 days after the good event of cash dividend announcement to reach the average of maximum CAR at -13.0012%. This table presents the effect of the variety of industries on the average number of days to reach the average maximum CAR and minimum CAR.
From Table 4.5 Investor Type Responding to PEAK and BOTTOM, the net trade imbalance is used to measure investor response. For the maximum result, after the cash dividend is announced on average of 21 days to reach the average of maximum CAR of 11.4158%, it influences local institutions, proprietary trading and foreign investors to execute the orders "Sell" in the market investing with a negative net trade imbalance of 6.6617%. For bottom result, after the cash dividend is announced on average of 29 days on average to reach the average of the minimum CAR on it affects the local institution to execute the orders "Sell" in the market invests with the negative net trade imbalance of -4.9576% which is significance at 99% confidence interval level shown.
Furthermore, Figure 4.1 Panel A type of investor reacting to the peak and Figure 4.2 Panel B type of investor reacting to the bottom show the significant difference in the trade of each investor type during the study period based on the cumulative imbalance of each investor type .
CONCLUSION
For the peak result, the study shows that most investors (local institution, real estate trading and foreign investors) execute the order to sell in the market, except the local individual who executes the order to buy in the market. For the bottom result, the study shows that local institution and property trading execute the order to sell and buy respectively in the market. Finally, the major limitation for the study is the focus on only the cash dividend announcement news.
While it is realistic, there are many news that will affect the market and stock price at the same time, such as the global news effect, the internal and external economic effect, the political effect, the financial effect as currency effect, etc. Therefore, the The article concludes that although the cash dividends have informational content, investment planning should receive more related information to analyze and develop the investment.
REFERNCES
