Firm Characteristics And Abnormal Returns During Share Repurchases

Research design

In this study, the quantitative research method with deductive reasoning approach was based on the research questions that we aimed to answer the research questions. Data were collected and analysed. Secondary data were used in the study due to the reliability of the data. The data were obtained from various sources, such as Bursa Malaysia’s website, Investing.com’s website, Thomson Reuters.

Sampling

In this study, judgmental sampling technique was used. There are 13,537 of share repurchase announcements from 1 ^st January 2013 to 31 ^st December 2017. There is a possibility that the sample becomes undue weighting if we consider every repurchase day as an event day because some firms made frequent repurchases. Besides, many useful repurchases information will be ignored if we simply take the first repurchase of each firm as the event day. Therefore, some criteria were used to select suitable data as final sample.

Based on the data obtained from the Bursa Malaysia, the initial sample consists of 264 public listed companies that have made 13,537 share repurchases from 1 ^st January 2013 to 31 ^st December 2017. The initial sample was sorted and screened according to the criteria and the companies without fulfil the criteria are eliminated from the sample. Therefore, the final sample consists of 493 observations from 153 public listed companies after fulfilling the criteria as mentioned. The final sample would be used in the analysis from years 2013 to 2017.

Data collection

In the beginning of the data collection, the information regarding the share repurchase announcement date and the public listed companies involved in share repurchases from 1st January 2013 to 31st December 2017 was collected. Then, the daily share prices and the FTSE Malaysia KLCI historical data (market prices) around the share repurchase day were collected from Yahoo Finance’s website and Investing.com’s website. The abnormal return of the stock for the day t was calculated based on the daily share price and the market price. The cumulative abnormal return of the stock on day t was calculated based on the abnormal return of the stock for the day t . Next, the data of the independent variables for every company were collected from Datastream. Then, the data were filtered and sorted based on the share repurchase days.

Measurement of variables

In order to examine the abnormal returns surrounding the share repurchase announcement, this study used the standard event study methodology. The day of the first immediate announcement of share repurchases was considered as the event day and it was designated as day “0” in the event period. The event window was 41 days which means 20 trading days before and after the day “0”. The choice of event window was similar to the studies conducted by Zhang (2005, 2002), Isa and Lee ( 2014) and Ong and Ng ( 2018).

The daily share prices were needed to calculate the daily returns and the daily returns were calculated using the following formula:

𝑅𝑖𝑡 = 𝑃𝑖𝑡- 𝑃𝑖𝑡-1/𝑃−𝑖𝑡−1 (1)

where

𝑅 _𝑖𝑡 = daily return for stock i on trading day t , 𝑃 _𝑖𝑡 = closing price for stock i on trading day t ,

𝑃 _𝑖𝑡−1 = closing price for stock i on trading day t -1.

Daily abnormal returns were calculated using the following formula:

𝐴𝑅 _𝑖𝑡 = 𝑅 _𝑖𝑡 − 𝑅 _𝑚𝑡 (2)

where

𝐴𝑅 _𝑖𝑡 = abnormal return of stock i for day t ,

𝑅 _𝑖𝑡 = actual return for stock i on day t ,

𝑅 _𝑚𝑡 = observed return of the reference market on day t .

Daily average abnormal returns, 𝐴𝑅 _𝑡 were obtained as follows:

𝐴𝑅𝑡 = 1/n ∑𝑛

𝑖=1 𝐴𝑅𝑖𝑡 (3)

where n is the number of firms on day t .

The cumulative abnormal returns, CAR were calculated as follows:

𝐶𝐴𝑅 𝐴𝑅 _𝑡 (4)

Other variables are explained in Table 01 below:

Moderating variable

Moderating variable is defined as a qualitative or quantitative variable that affects the strength of the relation between an independent variable and a dependent variable. In this study, firm size was used as a moderating variable that affects the strength of the relationship between firm characteristics and abnormal returns. The direction or magnitude of the relationship between firm characteristics and abnormal returns might change according to the firm size due to the existence of information asymmetric. Large firms usually have small information asymmetric, whereas small firms usually have large information asymmetric. Different firm sizes with different levels of information asymmetric will cause investors react differently. Investors might react positively when small firms implement share repurchases because small firms have large information asymmetric. By having the possibility that firm size causes the direction or magnitude of the relationship between firm characteristic and abnormal returns to change and by understanding the definition of moderating effect, this study hypothetically assumed that firm size has a moderation effect on the relationship between firm characteristics and abnormal returns. Therefore, the hypotheses were set to examine this moderation effect, but not the direct effect on abnormal returns.

Empirical model

The independent variables in this study were debt ratio, dividend yield, earnings per share, market-to-book value, pre-event abnormal return, and return on assets. The dependent variable in this study was abnormal returns. The regression equation is as follows:

CAR _(0,1) = 𝛼 ₀ + 𝛽 ₁(𝐷𝐸𝐵𝑇) + 𝛽 ₂(𝐷𝑌𝐼𝐸𝐿𝐷) + 𝛽 ₃(𝐸𝑃𝑆) + 𝛽 ₄(𝑀𝑇𝐵𝑉) + 𝛽 ₅(𝑃𝑅𝐸)+ 𝛽 ₆(𝑅𝑂𝐴) + 𝜀

CAR _(0,1) = cumulative abnormal return from days 0 to 1.

DEBT = debt ratio. This variable is used as a measure of the firm’s assets that are financed by debt.

DYIELD = dividend yield. This variable is used as a measure of growth opportunities. High dividend yield indicates less growth opportunities ( Latif & Mohd, 2013).

EPS = earnings per share. This variable can be affected by share repurchases. Investor might interpret share repurchases as an evidence of lower future earnings if the firm use share repurchases to affect the earnings per share ( Latif & Mohd, 2013).

MTBV = market-to-book value. This variable is used as a proxy for the general perception of market valuation. It is predicted to have negative coefficient ( Isa & Lee, 2014; Ong & Ng, 2018). High MTBV indicates overvaluation and low MTBV indicates undervaluation.

PRE = prior announcements return. This variable is used as a proxy for undervaluation. It is predicted to have negative coefficient ( Isa & Lee, 2014; Ong & Ng, 2018).

ROA = return on assets. This variable is used as a proxy for the availability of alternative investment opportunities. It is predicted to have negative coefficient. High ROA indicates the firm more likely to use cash for real investments and low ROA indicates the firm more likely to repurchase shares ( Isa & Lee, 2014).

The moderating variable was firm size. The independent variables in this study were debt ratio, dividend yield, earnings per share, market-to-book value, pre-event abnormal return, and return on assets. The dependent variable in this study was abnormal returns. The regression equation is as follows:

𝐶𝐴𝑅 _(0,1) = 𝛼 ₀ + 𝛽 ₁(𝐷𝐸𝐵𝑇) + 𝛽 ₂(𝐷𝑌𝐼𝐸𝐿𝐷) + 𝛽 ₃(𝐸𝑃𝑆) + 𝛽 ₄(𝑀𝑇𝐵𝑉) + 𝛽 ₅(𝑃𝑅𝐸)+ 𝛽 ₆(𝑅𝑂𝐴) + 𝛽 ₇(𝐹𝑆) + 𝛽 ₈(𝐷𝐸𝐵𝑇. 𝐹𝑆) + 𝛽 ₉(𝐷𝑌𝐼𝐸𝐿𝐷. 𝐹𝑆)+ 𝛽 ₁₀(𝐸𝑃𝑆. 𝐹𝑆) + 𝛽 ₁₁(𝑀𝑇𝐵𝑉. 𝐹𝑆) + 𝛽 ₁₂(𝑃𝑅𝐸. 𝐹𝑆) + 𝛽 ₁₃(𝑅𝑂𝐴. 𝐹𝑆)+ 𝜀

In this study, there were two different data analysis tools were used in the analysis: Microsoft Excel and IBM Statistical Package for the Social Science (SPSS). In the beginning of the study, the collected data and information were managed by using Microsoft Excel. Besides, the sorting and screening of the data as well as removing missing items were also done by using Microsoft Excel. Then, the data were transferred to SPSS for descriptive analysis, normality test and some other analyses. Other analyses include Pearson Product-Moment Correlation, multicollinearity test, Durbin-Watson statistic, heteroscedasticity test, t-test, and multiple regression analysis ( George, 2011).

Before we examined the moderating effect of firm size using multiple linear regression, the independent variables were centered in order to avoid potentially high multicollinearity with the interaction terms. Centering is a method to solve high multicollinearity issue when there is an interaction effect by subtracting the mean (a constant) from each case and then compute the interaction term and estimate the model ( Robinson & Schumacker, 2009).

Descriptive analysis

Descriptive statistic used to describe the basic feature of the data in the research. It provides a simple summary and measure of the data that used in this research. Besides that, it provides a simple large amount of data into more practical way to be present and easier to interpret the data finding.

Table 2 shows the findings of descriptive statistics. CAR (0,1) is cumulative abnormal return from day 0 to day 1 variable. CAR (0,1) from the sample of 426 observations had a mean of 0.0039 with a standard deviation of 0.0381. The CAR (0,1) had a positive mean indicating that most of the firms obtain a higher expected abnormal return from day 0 to 1. The standard deviation for CAR (0,1) was 0.0381. On average, the CAR (0,1) deviates from the mean by 0.0381.

PRE is prior announcement returns variable. It also represents the cumulative abnormal returns from day -20 to day -1 variable. The prior announcement returns variable had an average value of 0.0003 with a standard deviation of 0.0873. The prior announcement returns variable had a positive mean indicating that most of the firms obtain a higher expected abnormal return from day -20 to -1. The standard deviation for CAR (-20,20) was 0.0873. On average, this variable deviates from the mean by 0.0873.

Pearson correlation matrix

The correlation coefficients of cumulative abnormal return from day 2 to day 20 and cumulative abnormal return from day -20 to day 20 in Table 3 were negative 0.081 and positive 0.192 respectively and both were significant at the 5 percent level of significance. According to Zhang ( 2002), this indicates that there is a weak negative relationship between CAR (2,20) and CAR (0,1) and there is a weak positive relationship between CAR (-20,20) and CAR (0,1).

Based on Table 3 , the correlation coefficients were negative 0.093 and negative 0.081 respectively and both were significant at the 5 percent level of significance. Therefore, debt ratio and return on assets have weak negative relationship with cumulative abnormal return from day 0 to day 1. Besides, the correlation coefficient of prior announcement returns was negative 0.142 and significant at the 1 percent level of significance. Therefore, there is weak negative relationship between prior announcement returns and cumulative abnormal return from day 0 to day.

Regression analysis

Based on Table 4 , the t-statistic for debt ratio variable (DEBT) was -2.069 and statistically significant at 5 percent level of significance. Therefore, there is a relationship between debt ratio and abnormal returns surrounding the share repurchase announcements. The t-statistic for dividend yield variable (DYIELD) was -0.447 and not statistically significant at 10 percent level of significance. Therefore, this indicates that there is no relationship between dividend yield and abnormal returns surrounding the share repurchase announcements.

The t-statistic for earnings per share variable (EPS) was -1.536 and not statistically significant at 10 percent level of significance. Therefore, this indicates that there is no relationship between earnings per share and abnormal returns surrounding the share repurchase announcements. The t-statistic for market-to-book value variable (MTBV) was 0.147 and not statistically significant at 10 percent level of significance. Therefore, this indicates that there is no relationship between marketto-book value and abnormal returns surrounding the share repurchase announcements.

Based on Table 5 , the t-statistic for debt ratio variable (DEBT) moderated by firm size (FS) was 0.279 and not statistically significant at 10 percent level of significance. Therefore, this indicates that firm size has no significant moderating effect on the relationship between debt ratio and abnormal returns.

The t-statistic for dividend yield variable (DYIELD) moderated by firm size (FS) was 0.056 and not statistically significant at 10 percent level of significance. Therefore, this indicates that firm size has no significant moderating effect on the relationship between dividend yield and abnormal returns.

The results of the moderated multiple regression analysis show that the coefficient for interaction variable between earnings per share (EPS) and firm size (FS), and coefficient for interaction variable between return on assets and firm size (FS) were statistically significant at 10 percent level of significance. Therefore, it can be concluded that the firm size (FS) is a moderating variable. Besides, the results also show that the t-statistic of firm size was -1.172 and not statistically significant at 10 percent level of significance. Thus, firm size has no direct effect on abnormal returns.

Summary

This study was conducted to examine whether share repurchase can support and inflate the share price of the company in Malaysian stock market. Besides, this study was also conducted to add value to the literature by examining the relationship between firm characteristics and abnormal returns during the share repurchases.

The main objective of the study was to examine the relationship between shares repurchase announcements and abnormal returns. The results concluded that there is a positive relationship between share repurchase announcements and abnormal returns.

The second objective of the study was to examine the relationship between firm characteristics and abnormal returns surrounding the shares repurchase. The results concluded that there is a relationship between debt ratio and abnormal returns. Besides that, there is a relationship between prior announcement returns and abnormal returns.

The third objective of the study was to investigate the moderating effect of firm size on relationship between firm characteristics and abnormal returns surrounding the shares repurchase. The results concluded that firm size is a moderating variable, but firm size has no direct relationship with the abnormal returns.

Limitation

First limitation of the study is short sample period. In this study, the sample was limited to public listed companies in Bursa Malaysia that made share repurchase announcements during a sample period from 1 ^st January 2013 to 31 ^st December 2017. Thus, the analysis only covered 5 years of share repurchases. Many stock market situations are not included in the study due to short sample period, such as economic recession, interest rate hike, etc.

Second limitation of the study is unavailability of data. The data and information from the years 2013 to 2017 are collected from secondary sources, such as Bursa Malaysia’s website, Investing.com’s website, Thomson Reuters Datastream and Yahoo Finance’s website. However, some company data cannot be found, such as share price data and firm characteristics data. The companies that without the required information were excluded from the data analysis.

Third limitation of the study is the sample included all industries in Malaysia and all markets of Bursa Malaysia. Different industries and different markets of Bursa Malaysia would have different price movements and different firm characteristics. Therefore, some variables might not suitable to use in the study for certain industries and markets.

Suggestion for future research

There are several recommendations for the future research. First recommendation is use longer sample period, this would include more stock market situations and stock market events. Besides, longer sample period would have larger sample size. Results of larger sample will be more precise compared to small sample.

Second recommendation for the future research is focus on specific industry and specific market of Bursa Malaysia. For example, focus on Main Market or ACE Market of Bursa Malaysia. Future researchers can use suitable variables to explain that particular industry or market. In addition, future researchers can include more factors into the model for more comprehensive understanding about the firm characteristics of share repurchases. Then, find the best fitted regression model by using the stepwise regression.

Furthermore, future research also can consider other internal factors or external factors that have moderating effect on the relationship between the firm characteristics and abnormal returns. For example, liquidity ratios, profitability ratios, solvency ratios, efficiency ratios, market condition, market growth, inflation rate, interest rate, and disposable income.