In this section, we provide prior evidence related to our study. Prior evidence presents review of international as well as Indian studies which look at the impact of mergers on the operating performance of firms along with other related issues.

[5] analyzes 22 U.S. mergers during 1960-65 and finds evidence for a high degree of failures, especially for conglomerate mergers but horizontal and vertical integration are found to be successful. [6] assesses 478 U.S. mergers in 1951 and concludes that not all mergers are successful. However, there are many successful merged firms but it cannot be said that the firms would not have been successful if they had not gone through with a merger. In other words, there is no strong evidence to suggest that mergers add value. Also in the U.S., [7] , after analyzing 43 mergers in 1953 finds that investment performance of firms engaged heavily in merger activity was found to be worse than the average investment performance of their respective industries. The study concludes that mergers are risky investments. In the U.K., [8] analyzes a sample of 2,126 mergers during 1948-60 and finds that the performance of 57% of firms declined after merger―the study concludes that mergers are not profitable. In a sample of 60 companies drawn from the list of Fortune 500 companies, [9] finds that on average the financial performance of conglomerates after merger was no better than the related industry averages. Nevertheless, great variation in the performance of conglomerates is found. [10] examine a sample of 69 U.S. mergers which took place between 1952 and 1963 and find that the profitability of the firms after merger is higher than the non-merging firms, but they do not provide any clear evidence of the effects of mergers on the riskiness, growth, financial structure, income taxes paid and liquidity of the acquiring firm. In line with these results [11] - [20] also do not find any significant improvement in the operating performance of firms in the post-merger period. However, there are very few studies which show that mergers lead to improvement in the financial performance of firms. [21] examines 22 U.K. mergers and finds significant improvement in the profitability of firms after merger. [22] examine 50 big U.S. mergers and observe higher operating cash flow returns due to significant improvement in asset productivity compared to their industries. A strong positive relationship is found between post acquisition operating cash flows and abnormal stock returns at the acquisition announcement. The results of other studies which show improvement in the performance of firms after merger are [23] -[28] .

From international studies mentioned above, we find that majority of studies find that mergers do not significantly improve the operating performance of firms but Indian studies add few new dimensions to the literature. In Indian context [29] -[32] analyze the impact of merger by examining the mergers of sick companies. These researchers conclude that merger is an effective way to rehabilitate the sick companies and performance meets the projected performance. However, studies on the normal mergers such as [33] -[36] show that mergers do not improve the performance of the firms. Only [37] shows that the operating cash flow improves after merger in the form of synergies and his study does not find efficient utilization of assets to generate sales.

On the basis of literature review, it can be said that many studies assess the economic impact of mergers on the performance by analyzing the changes in the profitability or other measures of the combined firm. Studies prior to early 90s are based on the analysis of accrual accounting measures and from 90s onwards, researchers have started using cash flow measures in addition to accrual accounting measures to examine the changes in the operating performance after merger. The first study that uses cash flow measures and has given new direction to this research is [22] . Nevertheless, most accounting/operating based studies, whether based on US, UK, India or other parts of the world do provide evidence and support to the view that mergers do not improve the operating performance of firms (for example, [5] [8] [11] ). However, on the contrary, a handful of studies provide evidence of efficiency gains after mergers (for example [22] -[25] [27] [37] [38] ). From the above discussion, it is clear that most of the studies are of the view that mergers do not improve the operating performance of the firms, so we frame the following hypothesis:

H1: There is no significant improvement in the operating performance of firms after merger.

While examining the impact of mergers on the operating performance of the acquiring firms, it is not sufficient to examine the end level performance parameters such as ROA, ROE and CFO alone. Instead, examination of operating performance ought to be completed only after examining performance at each stage of operation. According to [39] , operating synergies are scale economies and occur when the physical process of the firm is changed so that the same amount of inputs or factors of production, produce more quantity of outputs. Having more production with lesser inputs, the firm is able to lower its average cost curve and enjoy an edge over the competing firms. Reduction in cost can be in the form of savings in material cost, labour cost, overheads cost, interest cost and tax cost. He further states that these savings result from the ability of firms to sell the product at a lower price than that of its competitors and thus, captures market share by reaping high profits. He also argues that although the intuition behind all these synergies is appealing but these have never been proved. So with this idea, we take this issue in our study to examine synergy creation at each stage of operation. Since, we have hypothesized in H1 that there is no significant improvement in the operating performance of firms after merger at the end level. Thus, we also hypothesize that there is no significant improvement in the performance at each stage of operation. Hence, we frame the following hypothesis:

H2: Mergers do not create synergy at any stage of operation.

One of the determinants of successful mergers is the type of merger. [17] suggest that successful mergers which lead to higher profitability are from related mergers having similar operations. This is because diversification by large holding firms through mergers can have problems of loss of control. According to [40] , mergers between group and subsidiary firms are very common in India. Many group mergers in India are attributed to the government policies which played an active role in the progress and growth of firms until 1991. This practice of merging firms within the same group has been a major part of restructuring process to improve the performance. These approaches suggest that the direction of mergers whether related/unrelated or group/non-group could influence post-merger process. This implies that mergers with different motives can have different impact on the performance. So, we examine this in our last objective. Having known the financial performance of these two categories of mergers, we go back and find out the motives behind these categories of mergers. [5] found high degree of failure in conglomerate mergers whereas horizontal and vertical mergers are found to be successful. [9] also found that on an average, the performance of conglomerates after merger was not better than the industry average. However, [20] found that the type of acquisition (conglomerate vs non-conglomerate) do not significantly influence post acquisition performance. [41] found that performance of group firms is better than non- group firms. Since the results regarding the types of mergers are mixed and inconclusive, we frame the following hypothesis:

H3: There are no significant differences in pre-merger and post-merger performance of different categories of mergers-related/unrelated mergers, group/non-group mergers and BIFR/non-BIFR mergers.

The present study deals with the mergers which took place in post liberalization era in Indian manufacturing sector. The data consists of all the mergers which took place in manufacturing sector in India from April 1, 2000 to March 31, 2010. The time period is selected to focus on the mergers which took place after the economic reforms of 1991 and to have sufficient post-merger operating performance data. We restrict our study only to manufacturing companies in order to minimize the potential confounding of extraneous variables [20] [41] [42] . The final sample size used for analyses is 62 pairs of mergers consisting of 128 firms (62 bidders and 66 targets). The sample size of 62 mergers is comparable to the previous studies carried out in significantly larger markets like US, UK, and other parts of the globe.

To examine the performance of different types of mergers, we classify our sample into three categories. On the basis of ownership group, first we classify our sample mergers into group and non-group mergers. We use the ownership classification given by Prowess database of CMIE (Centre for Monitoring Indian Economy) to find the ownership status of acquirer and target firms. Prowess gives a firm an ownership status as Private (Indian) if the firm is an independent firm and does not belong to any family business group whereas if the firm belongs to a particular business group then Prowess gives the name of that particular business group as the ownership status of the firm. Out of 62 mergers in our sample, 39 are group mergers and 23 are non-group mergers. Second, on the basis of relatedness of the nature of the business, we classify our sample mergers into related and unrelated mergers. We use National Industry Classification (NIC) given by Prowess database to classify mergers into related and unrelated mergers. NIC codes used in India are similar to Standard Industry Classification (SIC) codes used in U.S. to classify firms into different industries. The broad category of classification in SIC codes is called “Division” whereas it is called “Section” in NIC codes. In SIC codes, each division is further sub divided into various groups whereas in NIC codes, each section is further sub divided into various divisions. Our sample contains all the mergers of section C that is manufacturing section. We classify a merger as related merger if the five digit NIC codes of acquirer and target are same and unrelated merger if the five digit NIC codes of acquirer and target are different. According to the classification based on the related of the nature of the business, there are 46 related and 16 unrelated mergers in our sample. Most of mergers in our sample are concentrated in division 24 which is “manufacture of basis metals”. Third, we classify our sample into BIFR and non-BIFR mergers which are explained earlier. CMIE classify a merger as BIFR merger if the merger is on the order of BIFR. We have 10 BIFR mergers and 52 non-BIFR mergers in our sample.

It is obvious that the pre-merger (Pre₁) and post-merger (Post₂) operating performance of firms involved in merger could be affected due to economy wide and industry factors, or simply could be continuation of per merger performance. In order to control economy wide and industry factors, the study employs an adjusted performance measure while evaluating the post-merger operating performance. The study follows [22] and [20] and combines the pre-merger financial performance data of acquiring and target firms to obtain yearly aggregate performance indicators of the combined firm for pre-merger period. A comparison of the post-merger indicator with the above calculated pre-merger benchmark facilitates the measurement of the impact of merger on the operating performance of the combined firms. For the pre-merger period, the industry average indicator (Pre₃) is subtracted from the pre-merger combined firm indicator (Pre₁) to obtain pre-merger industry adjusted indicator (X). Similarly, for the post-merger period, the industry average indicator (Post₄) is subtracted from the combined firm indicator (Post₂) to obtain the post-merger industry adjusted indicator (Y). Pre-merger adjusted (X) indicator is compared with post-merger adjusted (Y) indicator to evaluate the post-merger performance of the combined acquiring and target firm. Industry average performance indicators are calculated for each firm. The industry of the acquiring firm at the time of merger is used to identify the industry. Average performance indicators of all the firms in that particular industry are calculated, which are then used as benchmark. Comparison of five year pre-merger industry adjusted performance measures is made with five year post-merger industry adjusted performance measures.

Financial data for five years prior and five years post-merger for each firm (62 acquiring firms, 66 target firms) and for each industry average is extracted from Prowess database of CMIE. The extracted financial data encompasses the period from 1999-2000 to 2009-2010. The end level operating performance parameters which are used in the study are: Return on Assets (ROA), Return on Equity (ROE) and Cash flow from operations to total assets (CFO/TA). All these measures are calculated for pre-merger and post-merger period. Pre-merger performance measure is the weighted average of acquirer and target firm(s), weights being assets for calculating weighted ROA and CFO/TA and equity for calculating weighted ROE. Post-merger parameter is the parameter of the acquirer (since target got merged into acquirer). Both pre-merger and post-merger parameters are adjusted against the industry averages each year. ROA is computed as ratio of profit after tax plus interest expense to total assets, which is given as ROA = (PAT + Interest expense)/Total Assets. ROE is computed as ratio of profit after tax minus preference dividend to net worth minus preference share capital, which is given as ROE = (PAT − Preference Dividend)/(Net Worth − Preference share capital). Cash flow from operations to total assets is net cash flow from operations divided by total assets. Net cash flow from operating activities is directly available in Prowess. It can also be calculated by using the following formula given by [20] :

Cash flow from operations (CFO) = WCFO ± changes in (Trade receivables + Prepayments + Inventories + other receivables) ± changes in (Trade creditors + Interest received + Provision for employee entitlements + other creditors).

Working capital from operations (WCFO) = (Operating profit before tax + Interest expense +/− Extraordinary items + Depreciation + Loss on sale of assets and investment + other write off) − (Profit on sale of assets and investment).

In addition to end level parameters, we also examine synergy at each stage of operation by decomposing ROA into six sub level parameters: material to sales, labor to sales, overheads to sales, tax to sales, interest to sales and sales to total assets. These are the sources of economic gain/synergy on merger. All these parameters are calculated for pre-merger and post-merger period. Pre-merger parameter is the weighted average of acquirer and target(s), weights being combined sales of acquirer and target(s). Post-merger parameter is the parameter of the acquirer (since target gets merged into acquirer). Pre-merger parameters and post-merger parameters are adjusted against industry average each year.

1) Match paired t test: The paper examines the proposed hypotheses using statistical data based on mean. Matched paired t test is used to compare five pre-merger industry adjusted performance parameters and five post-merger industry adjusted performance parameters. The year “0” i.e., the year of event is excluded from the analysis. The year of event (year “0”) figures are affected by onetime merger cost incurred during the year. So, it is difficult to compare the results of year “0” with the other years. The test determines whether there is a significant change in the “before/after merger” performance and allows us to attribute the result to the merger. Match paired t test compares pre-merger and post-merger industry adjusted parameters for overall sample and also for different categories of mergers.

2) Cross Section Regression: While using the matched pair t test on mean basis, we assume that the pre- merger performance will continue in future. But it is unreasonable to assume that pre-merger performance will continue in post-merger period at a constant rate [20] . Therefore, following [22] and Manson et al (1994) we relax this assumption and we investigate the impact of merger on the post-merger performance through a cross- sectional regression. The change in industry adjusted operating performance is estimated using the following two regression models for all the nine parameters:

where, Post_xi is the average post-merger industry adjusted performance parameter x for company i. Pre_xi is average pre-merger industry adjusted performance parameter x for company i. β₁ represents the association between pre-merger and post-merger industry adjusted performance. A significant β₁ indicates the continuance of pre-merger performance in the post-merger period. α₁ is the intercept which is independent of the pre-merger performance and indicates the extent to which post-merger performance is a function of merger [20] ). e_xi is error term. The benefit of using regression is that it does not make assumptions regarding the relative change in performance from pre-merger to post-merger.

We use the following regression the find the factor which explain post-merger performance:

where, is the post-merger performance parameter x which are ROA, ROE, CFO/TA and Sales/TA for company i. is the pre-merger performance parameter x which are ROA, ROE, CFO/TA and Sales/TA for

company i. is pre-merger material to sales ratio of company i, is the pre-merger labor to sales ratio of company i, is the pre-merger overheads to sales ratio of company i, is the pre-merger tax to sales ratio of company i, is the pre-merger interest to sales ratio of company i,

Group is a dummy variable which takes the value 1 if the merger is group merger and 0 if the merger is non- group merger. Related is a dummy variable which takes the value 1 if the merger is related merger and 0 if the merger is unrelated merger. BIFR is a dummy variable which takes the value 1 if the merger is BIFR merger and 0 if the merger is non-BIFR merger. PublicAcquirer is also a dummy variable which takes the value 1 if the acquirer is public company and 0 if the acquirer is private company.

The multivariate regression is used in order to know the factors including type of merger which explain the post-merger profitability, efficiency and cash flows.

Table 1 shows the summary statistics of our sample of 62 mergers. In a sample of 62 mergers, there are 62 acquirers and 66 targets. There are 4 such mergers in which the acquirer acquired two targets. We compare acquirer companies with the target companies in terms of size, profitability and efficiency. All these parameters are measured as five year average prior to merger i.e. from t − 5 to t − 1 years, t = 0 being the year of merger.