Finance research
CHAPTER III - Research Methodology
Introduction
In the present section, researcher explains the research process adopted in the study with proper explanation of the research approach, strategy, data collection and analysis methods used for the study along with appropriate justification. Silverman defined the pure research methodology (as cited in Hussey & Hussey, 1997) as “Methodologies refer to the overall approach to the research process, from the theoretical underpinning to the collection and analysis of data. Like theories, methodologies cannot be true or false, only more or less useful”. (p.54)
3.1 Research Designs
For the solution of managerial problems information is provided by a systematic enquiry known as business research. This was mentioned by Cooper and Schindler (2003). Exploratory, descriptive and explanatory are the three different kinds of research design. The definitions are given below.
Exploratory: This design aims at assessing phenomena, questioning and seeking new insights to the happenings in the business.
Descriptive: The situation or events and persons are described by this design. Sekaran U (2000) argues that ‘descriptive study is undertaken in order to ascertain and be able to describe the characteristics of the variables of interest in a situation’.
Explanatory: To explain the relationship between variables the emphasis is on the study of the situation or a problem (Saunders et al, 2007) The present study adopted exploratory study design in order to assess cause and effect. In the present study, researcher wanted to test the hypothesis on whether the merger and acquisition of the companies increases the share holder effect.
3.1.1 Research Approach
“The research approach indicates whether the use of... theory is explicit within the research design” (Saunders, et al., 2000, p.87). Mason (2002, p.179) describes the research approach as “deciding what theory does for your arguments”. This enables the researcher to”
- “Take a more informed decision on the research design
- Support the researcher in the decision-making process as to what will work and what not and
- Adapt the research design to cater for constraints, for example insufficient understanding of the topic to form a hypothesis” (Sauders et al, 2008, p.89)
Inductive versus deductive approach
Inductive approach:
This approach intends to build a theory or theory building process. Further where there is predefined objective set up for the research, it is an organized procedure for analysing data. It is emerged from the frequent, foremost or significant themes inherent in raw data, without the restraint imposed by prepared research methodologies is the primary objective of the inductive approach. According to Saunders et al (2000), rather than perceived thinking the findings from this research are based on unbiased observations. Further, Hyde (2000) reports that inductive approach start with specific instance observation and seeks to establish generalization about the phenomena under investigation.
Deductive approach
According to Saunders et al (2000), researcher develops a hypothesis and theory and tests it through empirical data collection method. Since it begins with an established theory or hypothesis and tested through empirical data collection, generalizability of the research would be reliable and valid. According to Hussey and Hussey (1997), ‘it is dominant research approach in the natural sciences where the laws provide the basis for explanation, permit the anticipation of phenomena, predict their occurrence and therefore allowed them to control. Deduction involves the gathering of facts to confirm or disprove hypothesized relationship among variables that have been deduced from existing knowledge’ (Ghauri and Gronhaung, 2005).’
The present study used deductive approach rather than inductive approach. First, researcher could not get an opportunity to interview company shareholders about the performance and profit of the company. Secondly, it is beyond the expertise of the researcher to develop hypothesis. Thirdly, already studies have proved that merger and acquisition improved the company’s shareholder wealth. Hence with the above following points, researcher adopted deductive approach for the present study.
Qualitative versus quantitative study
In the research approach there are two different ways which include quantitative and qualitative approach. According to Malhotra and Birks (2003, p.133) ‘the nature of the primary data can be quantitative or qualitative’. Cooper and Schindler (2003) stated that ‘quantity depicts the amount and the essential nature or characteristic of anything is the quality’. Merriam (1998, p.27) defined a qualitative case study as “A qualitative case study is an intensive holistic description and analysis of a single instance, phenomenon, or social, unit”. In the present study researcher used quantitative study as the data are expressed in numerical rather than text, as researcher did not get any opportunity to interview investors or shareholders of the companies
Research Design
In order to complete the study quantitative approach was used. The present thesis aimed to investigate the cause and effect between merger and acquisitions and value of shareholder in the short run with particular reference to selected Indian companies. In the preset research, the analyses considered for merger and acquisitions are
- Tata Steel
- Tata Motors
- Hindalco
Sample used for the study
The merger and acquisition transaction considered for the study was announced for all the three companies are listed below. The companies for the study selected based on the following criteria
The private sector companies
Mergers are domestic based
Availability of data for the three companies
Merger and Acquisition dates
1. Tata Steel (Company A)
- a. Estimation period -7th August 2006 to 13th December 2006
- b. Pre- Merger – 14th December 2006 to 29th January 2007
- c. Post-Merger – 31st January 2007 to 13th June 2007.
2. Tata Motors (Company B)
- a. Estimation period -3rd October 2007 to 25th March 2008
- b. Pre- Merger – 8th February 2008 to 25th March 2008
- c. Post-Merger – 26th March 2008 to 5th August 2008.
3. Hindalco (Company C)
- a. Estimation period -18th August 2006 to 26th December 2006
- b. Pre- Merger – 27th December 2006 to 9th February 2007
- c. Post-Merger – 12th February 2007 to 25th June 2007.
Description of samples
Based on appropriate characteristics of the sample members data was selected for the present study which means using judgement sampling method the data was selected. In the present study researcher was interested only on three companies and hence, these companies are selected for the merger and acquisitions transactions
Sample size selected for the present study was only three companies, thus, there would be sample bias. Hence for all the event windows of interest it was therefore important to have all the required share price data. The companies selected had all share price data for various event windows. In addition researcher selected only large acquisitions which involved in cross border transactions. A high inclination to engage in acquisition in emerging markets by large firms which are characterisied by high market to book ratios, high stock price performance and strong liquidity was found by Graham, Martey and Yawson (2008).
Further irrespective of how the acquisition is financed, the large firms experience significant shareholder wealth losses when they announce acquisitions of public firms was found in the study conducted by Moeller, Schlingemann, Stulz (2004). Regardless of whether cash or shared are used as a payment method, this is due to the fact that large firms are likely to overpay. In large acquisitions, Moller et al (2004) also suggest that the hubris hypothesis is big factor where by overpaying managers tend to reduce value from mergers and acquisitions.
Data Collection
3.1.6.1. Data collection Methods
A crucial aspect in any research is the method of data collection because any imprecision would in the collection methods would adversely affect the study results and interpretation provided would be invalid. There are two types of data collection methods are available.
Primary data
The data collection done by the researcher through observations, experiments and surveys are called primary data (Ghauri and Grounhaug, 2005). However, for the preset study researcher not intent to collect any primary data as explained above.
Secondary data analysis
The data collected through desk-based approach was called as secondary data. According to Jackson (1994) ‘the value of a research is related to its data collection methods and importantly, whether or not it includes both secondary and primary data’. Further validity and reliability of data collection methods depends on the location and chosen methods. In the present study, researcher collected data from company website (electronic data base).
There are two types of secondary data available which include internal and external data. The data which can be used readily which has already available in ready format is internal data, while external researcher has to depends various sources like books, directories, internet, and journals. The data obtained for the present study was mainly internal (generated by the firms which is being researched), which has been available in a ready format.
Data analysis
In the present study to measure the impact of an unanticipated event on stock prices researcher used event study methodology (McWilliams and Siegel, 1997). The event study methodology was deemed to be appropriate for the presents study as the aim of the study was to investigate the share price reaction of large organization.
Since the data was not normally distributed, Mann Whitney U test was used to test the hypotheses. Further Man Whitney U test was used to explore whether the price performance mean abnormal due to acquisitions was different statistically from zero at the error level of 5%.
Measurement of Share price performance
The market model was used to measure the share price performance. Mackinlay (1997, p.18) states ‘the market model as a statistical model which related the return of any given security to the return of the market portfolio”.
Using the log returns, daily equity weighted indices were constructed using the formula below
R = log [Pit/Pit-1]
R – For day t, the equal weighted share return
P- At the end of day, the equal weighted share value
The assumed joint normality of asset returns are followed by the model linear specification. The market model for any security is
Rpr = α + βRmr + Єit
E (Єit =0), var (eit) =σ ei
Rpr = t returns period
The parameters of marker model are σ, β, and α are estimated from the regression approach.
Rmr = market portfolio
Єit = error term
Over the constant mean return model, the market model represents a potential improvement. The variance of the abnormal return is reduced by removing the portion of the return is related to variation in the market return. This in turn helps to detect event effects by increasing the credibility. The market model is benefited through market model regression of R2 and the variance reduction of the abnormal return is greater when the R2 is higher and the gain would be larger.
Abnormal return calculation
T can be expressed using the formula ARit=Rit – E (Rit), the abnormal return ARit for the company i on day.
Average abnormal returns calculation
The average abnormal returns are calculated to eliminate stock shocks for the event windows. The formula used to calculate the abnormal returns
AARt = 1/N ƩNt=1 Art
The abnormal return for day was indicated by AARt and number of securities was represented by N.
Cumulative Average abnormal returns
By aggregating AAR over the event window, the cumulative average return was calculated.
CAARt = 1 / N ƩN t = 1 AARt
The cumulative Average Abnormal return for the event period t is represented by the CAAR.
Selection of event window
Studies have showed that it is always better to keep the events very short in order to minimize the impact of confounding events (McWilliams and Siegel, 1997). The reduction in the power of t-statistics perhaps would be reduced using a long event window was showed by several authors (Brown & Warner, 1980; 1983; Cited in McWilliams and Siegel, 1997). The significance of an event would be false due to this reduction in the power. Due to the following above reasons, study classified the acquisitions into three, eleven, twentyone, forty one day and hundred and fourty one day window and assessed the impact of large acquisitions.
The following event windows were chosen for the study
- An Eleven day window [-5, +5]
- A twenty one day window [-10, +10]
- A forty one day window [-20, +20]
- An one twenty one day window [-30, +90]
In order to capture as many transactions as possible during the economic downturn, the 90 day pre event was chosen.
Statistical Analysis
The statistical significance was tested for the CAAR and AAR which was calculated above. SPSS software was used to analyse the data. At the 5% significance levels the null hypothesis could be rejected and similarly if the p value was greater than α = 5% which at the 5% significance levels, the alternative hypothesis was rejected.
