In Product Market, What Does an Upstream Horizontal
Merger Mean for The Customer of The Merged Firms?
Master Thesis
Amsterdam Business School
MSc Finance (Quantitative Finance Track)
Student: Yi Zhang
Supervisor: Dr. Florian Peters
Date: 30. June. 2018
Statement of Originality
This document is written by Student YI ZHANG who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of contents:
Sections
Page number
1. Introduction --- 4
2. Literature review --- 6
3. Hypotheses --- 10
4. Data preparation --- 11
5. Methodology --- 17
6. Robustness checks --- 24
7. Conclusion --- 27
1. Introduction
Mergers and Acquisitions (M&A) are usually considered high value-adding transactions that are entered into for a number of reasons. The majority of the literature examines whether these transactions have achieved their stated objectives. In examining the value-adding nature of these transactions, many papers scrutinize the M&A impact by looking into the post-M&A performance of the merged firm or the stock price performance of the firms involved in the merger.
As Sheen (2014) says, industries do not exist in isolation but rather are connected through the complex customer-supplier linkages. For the purposes of achieving a holistic view of the impact of M&A, it would seem appropriate to investigate how these transformational transactions impact stakeholders such as customers with inherent interest in the outcome of M&A. There is a growing interest and body of literature since 2010 examining the outcome of M&A by looking into how they impact stakeholders such as suppliers, competitors and customers.
Ahern (2011) finds that horizontal mergers can negatively impact the suppliers by increasing the bargaining power of merged firms. Sheen (2014) finds that horizontal mergers adversely affect the competitors by achieving economies of scale. Fee and Thomas (2004) look into the impact of horizontal mergers on the customers firms by examining the stock price and the post-performance of the merged firm’s customers. Their results are not conclusive. Absent in scholarly literature is how buyers are impacted by the supplier consolidation.
It is beneficial for stakeholders to be aware of and evaluate the potential impact of such transnational transaction. The confirmation of the impact of the horizontal merger on customers of merged firms is still absent. Therefore, my thesis attempts to address the theoretical gap by applying an alternative methodology to assess the impact of horizontal mergers on customers and consolidating it to extant knowledge. This awareness of the potential impact of the mergers will provide the customers with strategic insight and versatility into how to respond to the supplier consolidation, since the impact of the merger takes two to three years to be realized. Customers may, as a strategic response, react by switching to other suppliers or entering into a transformational acquisition to adapt to the new market conditions. The strategic response however, can only be effectuated when the impact of the merger has been evaluated. The eventual impact of the adjustment to the merger will also only be realized two to three years after the transaction. This awareness acts as an early-warning signal because of the lag time. Therefore, from the perspective of customer firms and investors, this thesis can help them to forward plan and be better prepared by considering the impact of the horizontal mergers in supplier industries. For financial firms involved in advisory services, this thesis offers them a new entry point of giving financial advice.
In the existing literature which examines the impact of horizontal mergers on the customers of merged firms, stock price reactions to the M&A announcements and the post-M&A long-term performances are tested (Fee and Thomas 2004). However, the direct connection between suppliers and buyers is the output that suppliers provide to buyers and the price charged by suppliers, which is the Cost of Goods Sold (COGS) of buyers. If there is any impact of M&A on the buyer’s level at all, the impact will be reflected by either the price that the buyer pays or the quantity of goods that the buyer purchases. Neither the stock price nor the long-term performance is an explicit indicator of changes in prices or quantities. Therefore, this thesis includes Cost of Goods Sold (COGS) as a mechanism to imply how supplier consolidation impacts buyer firms.
In order to assess the impact of supplier consolidation, this thesis applies two groups exhibiting specific characteristics. The first group (treatment group) comprises the buyers of the merged firms. The second group (control group) consists of buyers of unmerged firms that are actively competing with the buyers in the first group. The difference between these two groups is that suppliers of the treatment group are merged and the suppliers of the control group are not merged but that the buyers are in competition. The distinction is to detect the impact of supplier mergers on buyers, so the changes which matter are the change caused by the treatment – supplier horizontal merger. There are other factors that can affect the input price of buyers. Therefore, it is necessary to know what the circumstances of the buyer would be if the supplier does not merger. While this counterfactual result is unobservable, a control group having similar characteristics as the treatment group while not receiving the treatment (supplier horizontal merger) should be comprised. A special database created by Hoberg and Phillips (2010) which measures the firm similarity by product description is used to identify the control group. An algorithm in SAS is applied during the data preparation process. Because the impact of M&A will be seen two to three years after the transaction, the three-year change of COGS after supplier merger is estimated for both the treatment and control group, and then the relative change between the treatment group and the control group is calculated as the impact of the horizontal merger on the buyer. Three hypotheses are proposed based on the potential outcomes of the horizontal merger. The first hypothesis is that the treatment group will have a lower COGS than the control group, which implies that economies of scale are an impact of horizontal mergers; the second hypothesis is that the treatment group will have a higher COGS than the control group, which allies with the collusion outcome of horizontal mergers; the third hypothesis is that the treatment group and control group have the same change in COGS which is consistent with the countervailing acquisition hypothesis (Galbraith, 1952) that the buyer’s previous industry concentration induces suppliers to grow larger in order to neutralize the market power of the buyer. By applying the COGS mechanism, both the collusion hypothesis and countervailing acquisitions hypothesis are verified based on the sample. The suppliers of buyers which have larger firm sizes tend to consolidate to
weaken the market power of buyers, and horizontal mergers increase the market power of merged firms, resulting in charging more from the buyers.
This thesis will start by introducing the background of the topic by reviewing related literature; secondly, the process of data preparation and the methodology will be explained in detail; thirdly, the statistical results will be interpreted in terms of economic meanings; then the results will be critically assessed by performing robustness checks; finally, a comprehensive conclusion will be made.
2. Literature review
Industries do not exist in isolation, but rather are connected through the complex network of customer-supplier linkages. Especially in product markets where the finished goods and services are bought and sold, the entities are actively interacting through the customer-supplier or competitor relationships. For example, Banerjee, Dasgupta, and Kim (2008) find that the interaction between a firm and its suppliers and customers can affect its overall strategy. Rhodes and Robinson (2008) discover that “high buys less high”- horizontal mergers are more likely to happen between closer competitors. These dynamic relationships among firms not only impact the corporate decision making, but also affect the overall outcome of corporate decisions. For example, Ahern (2011) shows that in product markets, the distribution of gains from vertical mergers between acquirers and targets can be partially explained by the relationship between them. Less dependence of the target on the acquirer leads to a larger gain of transaction distributed to the target. Hoberg and Phillips (2010) also find that in competitive product markets, the less similar the acquirer and target firms are, the larger the M&A gains are. The reason is that product differences between mergers bring asset complementarities to the merged firm. Besides, according to the countervailing acquisitions theory (Galbraith, 1952), the consolidation in downstream industries stimulates upstream industries to also consolidate in order to naturalise the increasing market power of the customer firms. The impacts of corporate activities are also spread within and across industries via these complex industry networks. For example, Ahern and Harford (2014) find that in product markets, the merger wave in one industry drives the merger waves in other industries. The closer the relation these industries have, the quicker the merger wave shifts; the more central the merger wave is located in the product market, the wider the merger wave spreads in the economy.
As the literature has already shown how dynamic the interactions between industry linkages and corporate decisions can be in product markets, and since M&A is one of the most important activities in the corporate world, it is interesting to see what the M&A outcomes would be after they travel through these crisscrossing industry relationships. Certainly, transactions do not only impact those direct deal participants, but also affect other stakeholders, such as suppliers, customers, and
competitors of the merged firm. For example, Bhattacharyya and Nain (2010) find that in product markets, the suppliers of merged firms will be negatively affected by the increasing bargaining power of merged firms. Also, Sheen (2014) shows that in product markets, horizontally merged firms are able to reduce the output price which adversely impacts its competitors. However, there is not much evidence at the level of the customers of merged firms yet. T
his thesis will focus on the latter to
provide firms with insight into the kinds of impact they can expect from horizontal mergers
on their suppliers.
There are mainly two approaches to testing the effects of M&A. The first approach is to observe the stock price reaction to the M&A announcement. In this event-study based approach, an assumption has been made that stock prices are efficient and able to fully reflect the predicted effects of M&A on a firm’s value. This approach can only be applied to listed firms which have a stock price. The second approach is to test the post-M&A long-term performance of firms. The post-M&A performance can either be compared to the pre-M&A performance or to a control group which has similar characteristics to the merged firm, while not conducting the transaction. An example of assigning a control group to merged firms is the work by Ulrike, Enrico, and Florian (2012). They categorized the bidders who lost the contests into the control group and compared the three-year stock abnormal returns between winners and losers.
There are difference types of M&A and the motivation behind each deal varies. In terms of horizontal transactions, the motivations behind them have been the subject of a long-standing debate in both finance and economic literature. There are two main arguments. The first one is that firms execute horizontal transactions to achieve economies of scale, which can reduce their overall costs. This is consistent with the finding of Sheen (2014), in which the merged firms are able to set a relatively lower output price than their competitors. The economies of scale effect can sometimes also benefit the merged firm’s customers by reducing their cost of purchasing inputs. The second argument is that horizontal acquisition is motivated by collusion. Consolidation helps the merged firm to gain more market power which can be expressed as increasing output price (Eckbo, 1983). This impact which reduces the customer’s welfare is undesirable. US Federal Trade Commission uses consumer surplus as the criterion to decide whether or not to approve a merger proposal (Pittman, 2007).
When it comes to horizontal M&A effect testing, there are two major angles. The first perspective is to find horizontal M&A impacts by testing the merged firms. For example, Eckbo (1983) uses the stock prices of merged firms to test the outcomes of horizontal mergers. He assumed potential collusion is good news which will increase the stock price of both the bidder and target. But there was not much evidence found. The second perspective is to find evidence of horizontal M&A impact from
stakeholders, such as the suppliers, customers, and competitors of the merged firms. There are some examples. Fee and Thomas (2004) test the impact of horizontal mergers (1980 ~ 1997) on the merged firms by looking for evidence from customers, suppliers, and rivals of merged firms in terms of stock price reaction and the change of operation performance. Little evidence was found in the customer firm’s level. Adverse effects on the competitor and supplier are found, complying with the argument that horizontal mergers increase the efficiency and buying power of the merged firm. Besides, to test the impact of horizontal takeovers (1987 ~ 1999), Shahrur (2004) calculates the stock abnormal return of the merged firm’s rivals, suppliers, and corporate customers during the period of 250 to 300 days after the transactions. Similar to the study of Fee and Thomas (2004), not much evidence was found from the corporate customers, while the statistical results of the rivals and suppliers align with the economies of scale argument.
However, change in general firm performance or stock price actually does not tell the ‘why’. What is the fundamental reason of this change? A change in a firm’s profitability could be caused either by a change in sales or in costs. Instead of roughly testing the general firm’s performance, some literature specifically looks into the merged firm’s and stakeholder’s main linkages, such as product price. Because of the diversification of products across firms, there is an obstacle of price and quantity data collections. Therefore, most of these studies are only based on a few deals or focus on a specific consumer product industry, such as airline, banking, hospital, and petroleum. For example, Simpson and Christopher (2008) study the price impact of a merger case between two Michigan petroleum companies by comparing the petroleum price movement in Michigan with the movement in two nearby states that were not affected by this transaction. They do not find that evidence of this deal harms costumer surplus. Besides, Kim and Singal (1993) test all airline merger cases from 1985 to 1988 since governments did not intervene in these mergers. The prices of routes served by merged airlines were relatively expensive compared to the prices of routes which remained unaffected by the mergers. Although they also find that the airline horizontal acquisitions lead to more efficient operations, the market power gains are more than the efficiency gains. In addition, Ashenfelter and Daniel (2008) test the price impact of horizontal mergers based on five big merger cases in consumer product markets. They find an increase in output prices after transactions in four out of the five cases. In recent studies, the obstacle of obtaining price and quantity information has been gradually overcome. Sheen (2014) uses the data from ‘Consumer Reports magazine’, which disclosures the reviews and buying advice on a wide range of consumer products sold in the United States. 719 product-brand combinations within the 20 product categories are examined. He finds that horizontally merged firms are able to set a lower output price than their competitors. Furthermore, Bhattacharyya and Nain (2010) obtain the Producer Price Index (PPI) from the Bureau of Labor Statistics to measure the change over time in the selling prices. Based on the product marker horizontal deals from 1984 to
2003, they find that the more the downstream industry depends on the upstream industry, the larger the buying power gain is in the downstream industry after consolidation.
Although the price impact of horizontal mergers has been examined from the perspective of the suppliers, competitors, and customers of merged firms, there is still an area that is not covered by the existing literature. In both supplier and competitor levels, the price impact of the horizontal M&A has been tested in terms of the main connections between these interest parties and merged firms rather than overall firm performance. In detail, regarding the suppliers of merged firms, Bhattacharyya and Nain (2010) use the output prices of the suppliers of merged firms as an indicator for the price impact of horizontal mergers; regarding the competitors of merged firms, Sheen (2014) uses the product price of the competitors of merged firms to illustrate the price impact of horizontal mergers. However, there is not any research testing the price impact of horizontal mergers in the customer firm’s level by using a more direct connection than general firm performance. As Sheen (2014) finds, there is an output price change in the merged firm relative to its competitors, but how well is this price impact of horizontal merger realized and reflected in the downstream industry? The fact that even customer firms of the same supplier can receive different unit prices should not be ignored. This thesis tries to explore this undiscovered area by using a more direct mechanism to test the price impact of horizontal M&A from the downstream industry’s point of view.
There are two main arguments regarding the price impact of horizontal mergers. The first one is that horizontal mergers decrease the output price because of the increase of production efficiency in the merged firms. The other argument is that horizontal mergers increase the output prices of merged firms, due to the collusion between mergers. The existing literature examining the price effects of M&A on the merged firm’s customers mainly tests the post-performance of customers of merged firms. Those studies expected the price impacts of transactions to show up at the firm performance level. But this is not always the case. The change of firm performance can also be caused by other factors which are unrelated to merged firms. Even when there are performance changes in customer firms which are caused by supplier M&A, they can also be balanced out by performance changes which resulted from other firm activities. Hence, instead of firm performance, a more direct indicator is needed to better detect the price impact of horizontal mergers on the customer of merged firms. The output of upstream industries which will be the input of downstream industries is a direct and tangible link between the upstream and downstream industries, and Cost of Goods Sold (COGS) measures the amount of money that a firm spends on purchasing inputs. Besides, the different prices that one supplier charges from different customers can also be captured by the COGS of different customer firms. Hence, to some extent, COGS can be an indicator for the price impact of supplier consolidation that is more direct than firm performance. Since an increase in sales will also result in
higher COGS, COGS will be scaled by sales before being used as the price impact indicator. Since the price decline and quantity convergence will be realized two to three years after the transaction, it is hypothesised that a three-year change of COGS/Sales in the downstream industry can be treated as an indicator of price impact from the upstream industry.
Besides, to investigate the impact of the supplier mergers (the treatment) on the customer firms (the treatment group), it seems necessary to know what the situation would be if the transaction did not take place, and then compare this situation with the circumstance that the supplier merger did happen. This counterfactual result is unobservable; therefore, a control group is needed which can represent the situation in which no transaction occurred in the supplier industry. In this thesis, the top three competitors of each firm in the treatment group will be distributed into the control group, and the three-year changes of COGS/Sales of firms in the control group will also be calculated and treated as the three-year changes of COGS/Sales of the downstream industry firms under normal conditions – if suppliers did not participate in any transaction (More details in the data preparation and methodology parts).
3. Hypotheses
In this thesis, there are three hypotheses regarding the price impact of horizontal mergers on the corporate customers.
Hypothesis 1: Economies of scale hypothesis
Customers of the merged firms will have a lower COGS/Sales relative to competitors of the customers.
Suppliers who gain economies of scale through horizontal mergers will be able to set a lower output price. Therefore, customers of the merged firms can enjoy lower COGS/Sales. While the decrease of customer firm’s COGS/Sales which is consistent with efficiency enhancement is not absolute evidence for economies of scale in supplier firms, changes in customer firms can also affect the COGS/Sales. For example, customer firms that increase their production can negotiate a better unit price with the supplier by increasing the amount of purchasing. The reason for COGS/sales reduction could also be that the customer firm optimises its production process which reduces the probability of producing inferior goods, therefore decreasing the overall COGS. Or, there is collusion in downstream industries which enhances the bargaining power of customer firms; as a result, the customer firms are able to negotiate a better price with supplier. In addition, an increase in sale price in downstream industries can also dilute the COGS/Sales.
The COGS/Sales of the merged firm’s customers is relatively higher than the COGS/Sales of competitors of the customers.
The horizontal consolidation in supplier industries facilitates merged firms to collude, which increases the market power of merged firms, enabling the merged firm to increase the price, which results in higher COGS/Sales in customers of merged firms. However, the increase in COGS/Sales in the merged firm’s customers is not necessarily evidence for market power enhancement of merged firms. Apart from getting consolidated, there are other sources which can contribute to the output price increasing in the supplier industry, such as the restrictions on the scarce raw materials that are used by the supplier industry. Besides the increased output price, there are also different expressions of the market power enhancement. Suppliers can bind other products with the desired output and force the customer firm to purchase the whole package. In addition, the increase of COGS/Sales could also be a result of decreasing sales in customer firms.
Hypothesis 3: The countervailing acquisition hypothesis
There is not much difference between the COGS/Sales of the merged firm’s customers and the COGS/Sales of the competitors of the customers.
It is difficult to distinguish other possible cause of COGS/Sales changes in the customer firms. The economies of scale and market power enhancements can also coexist when the horizontal merger occurs. Galbraith’s (1952) theory of countervailing acquisitions argues that seller industry concentration induces buyers to grow larger in order to neutralize the market power of sellers. Countervailing acquisitions theory also implies that the consolidation in seller industries may be triggered by the earlier collusion of buyers. The sellers who execute horizontal mergers may have experienced huge market power loss. The countervailing acquisition hypothesis will be used to see if the COGS/Sales changes in downstream industry firms can be attributed, at least in part, to the supplier horizontal merger.
4. Data preparation
Firstly, a dataset which contains horizontal merger deals is constructed to identify the consolidated suppliers and the time period of pre/post transaction. The deals are gathered from Thomson One. Only the deals which satisfy the following criteria are included:
i. All transactions are announced between 1997 and 2014 (the reason for choosing this time period will be explained later in the thesis)
(SIC)
iii. Deals were completed eventually
iv. Regulated utilities (SIC codes 4900 ~ 4949) and financial firms (SIC codes 6000 ~ 6999) are excluded
v. Acquirer bought at least an 80% stake of the target
Secondly, the major corporate customers of each merged firm which will be categorized into the treatment group are identified through the ‘Compustat Segments - Customer’ database, which discloses the information of any customer that represents more than 10% of the total sales of a firm as well as the segment that was primarily responsible for these sales. To link the merged firms with the customers of the merged firm, the ‘CRSP/Compustat merged table’ is used. Because ‘Thomson One’ uses 6-digit CUSIP as the firm identifier while ‘Compustat Segments – Customer’ database uses GVKEY to label firms, the CUSIP-GVKEY linking table which provided by ‘CRSP/Compustat merged table’ will be needed to link those two datasets.
Further on, the financial details of the merged firm’s customers will be needed from ‘Compustat Daily Updates - Fundamentals Annual.’ However, only the customer name is displayed in‘Compustat Segments Customer’ without any other firm identifiers. The customer names displayed in ‘Compustat Segments – Customer’ database are not identical to the company names that were disclosed by ‘Compustat Daily Updates - Fundamentals Annual,’ therefore, similar to the methods of Banerjee, Dasgupta, and Kim (2008) and Fee and Thomas (2004), a fuzzy string matching programming technique in SAS called SPEDIS is used to match the customer names in ‘Compustat Segments – Customer’ with the company names in ‘Compustat Daily Updates - Fundamentals Annual.’ Hence, the generated table contains both merged firms and the customers of merged firms which are all identified by GVKEY.
Thirdly, to identify the control group, which includes the firms from the same industry as the merged firm’s customers whose suppliers did not participate in any transaction, the ‘10-K Text-based Network Industry Classifications (TNIC),’ which is a database created by Hoberg and Phillips, is used. This database provides a similarity score of firms, which measures company similarities based on the product descriptions. Those product descriptions are obtained directly from ‘10-K filings’ on the ‘SEC Edgar website.’ The higher the similarity score a pair of firms has, the more similar their goods are. It is assumed that higher product similarity implies higher competition.
This ‘10-K Text-based Network Industry Classifications (TNIC)’ dataset also uses the GVKEY as the firm identifiers. The firm similarity dataset is merged to the ‘Merged Firm & Customer of Merged Firms’ via the customer GVKEY of the merged firms. For each customer of the merged firms, only
the top three competitors which have the top three similarity scores are kept. The ‘Compustat Segments – Customer’ database is also adapted here to identify the supplier for these three competitors. Those competitors whose supplier is also involved in transactions will be excluded. Hence, eventually the number of firms in the control group is less than three times the number of firms in the treatment group. Therefore, the ‘Merged Firm & Customer of Merged Firm & Competitors of Customer Firms’ linkage table is generated.
In addition, this ‘10-K Text-based Network Industry Classifications (TNIC)’ database starts from 1996, while the database of Edgar website is more complete after 1997 when the electronic filing is required. Therefore, the sample of this thesis will start from 01 Jan 1997. Because the three-year post-transaction change will be calculated and the fiscal year financial information is only available until 2018, only transactions which were announced between 01 Jan 1997 and 31 Dec 2014 are included in the sample. Graph 1 shows the deal distributions across years, and table 1 shows the summary statistics of the sample
Graph 1: The distribution of deals (horizontal M&A in product market) over time (from 1997 to 2014)
As graph 1 shows, from 1999 the number of deals starts to decrease and becomes more stable until 2010. The reason may be the enhancement of transaction regulations, which includes horizontal transactions antitrust enforcement. On the other hand, since this final dataset is generated by merging several datasets, the deal distribution pattern could also be contributed by the data availability in different databases and the matching capacity between different datasets.
Table 1: Summary statistics of the sample
0 20 40 60 80 100 120 140 160 180 200 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Number of deals
Column (1) shows the macro-industry of the merged firms. There are 10 supplier macro-industries, including Consumer Products and Services, Consumer Staples, Energy and Power, Healthcare, High Technology, Industrials, Materials, Media and Entertainment, Retail, and Telecommunications. Column (2) shows the numbers of merged firms. In total, there are 1077 merged firms. The majority of the deals happened in the High Technology industry. Column (3) reports the number of firms in the treatment group which are the major customers of the merged firms. Since one supplier can have more than one customer, the number of merged firm’s customers is 2363 which is higher than the number of suppliers. Column (4) indicates the number of firms in the control group which are formed by the main competitors of the customers of merged firms. Although the top three competitors for each merged firm’s customers are tagged, the competitors which also have merged suppliers are eliminated. Therefore, the number of firms in the control group is 5148, which is slightly less than three times the number of firms in the treatment group. In total, there are 7511 (2363+5148) observations.
(1) (2) (3) (4)
Supplier macro-industry Number of merged firms Number of firms in the treatment group (Customers of the merger firms) Number of firms in the control group (Competitors of the
merged firm’s customers)
Consumer Products and Services 59 121 329
Consumer Staples 80 171 546
Energy and Power 112 248 626
Healthcare 155 397 644
High Technology 397 826 1757
Industrials 142 353 690
Materials 48 88 222
Media and Entertainment 19 36 107
Retail 9 13 23
Telecommunications 56 110 204
N 1077 2363 5148
Fourthly, financial details for both customers of merged firms (the treatment group) and competitors of customer firms (the control group) are obtained from the ‘Compustat Daily Updates - Fundamentals Annual’ database. The summary statistics of the means of the three-year characteristics for both sample groups before the supplier consolidations is shown in table 2.
Table 2: Firm characteristics of the downstream industry firms before the consolidation of the supplier industry
This table reports summary statistics of firm characteristics in the three years before the supplier’s consolidation. The mean and standard deviation (in parentheses) for each variable are reported separately for both the treatment group and control group. Column (1) shows the variables which are defined below. Column (2) reports estimates for firms whose supplier will take horizontal acquisition in the coming year. Column (3) reports estimates from the same year and industry for firms whose suppliers will not participate in horizontal acquisition in the coming year. Column (4) shows the difference between the mean of the treatment group and the control group. ***denotes significance at the 1% level and **denotes significance at the 5% level according to the t-test.
COGS/Sales: The Cost of Goods Sold (COGS) is scaled by Sales
EBITDA/Sales: The Earnings Before Interest, Taxes, Depreciation and Amortization (EBITDA) is scaled by Sales, measuring the profitability of a firm
Return on Assets: The Net Income is divided by Total Assets, seen as a firm profitability indicator Debt/Assets: The Total Debt is divided by the Total Assets, showing the leverage ratio of a firm ln(Sales): The natural logarithm of Sales, measuring size of a firm
ln(Assets): The natural logarithm of Total Assets, measuring size of a firm
ln(Employees): The natural logarithm of number of employee, which also can measure the size of a firm
As table 2 shows, the overall differences of firm characteristics between two sample groups are significant at the 5% level, meaning firms in the treatment group and the firms in the control group are quite different before the supplier consolidations. The customers of the future merged firms have higher ln(Assets), ln(Sales) and ln(Employees) than the competitors of future merged firm’s customers (control group). It is obvious that before the supplier industry consolidation, the firms in the treatment group have larger firm sizes than the firms in the control group. The customers of the future merged firms appear to have larger firm sizes which could be a result of previous consolidation among the customers of the future merged firms. This phenomenon seems to comply with the countervailing acquisitions argument that consolidations in seller industry are driven by the consolidations in buyer industry to weaken the market power of the downstream industry. Besides, this table also shows that the customers of merged firms have higher profitability and COGS than the competitors of the merged firm’s customers. The reason could be that the extra sales that the treatment group made are able to make up for their relatively high COGS and still maintain a higher profitability (EBITDA/Sales, Return on Assets). In terms of the capital structure, both sample groups
have a similar debt to assets ratio. Table 3 below shows the correlations of the variables listed in table 2.
(1) (2) (3) (4)
Treatment group (Customers of the future
merger firms)
Control group (Competitors of the future merged firm’s
customers) The difference COGS/Sales 0.672 0.645 0.027*** (0.225) (0.209) EBITDA/Sales 0.150 0.134 0.016*** (0.132) (0.122) ln(Sales) 10.268 8.946 1.322*** (1.546) (1.588) ln(Assets) 10.136 8.833 1.303*** (1.639) (1.601) Debt/Assets 0.242 0.224 0.017*** (1.152) (1.154) Return on Assets 0.045 0.038 0.006*** (0.060) (0.069) ln(Employees) 4.028 2.991 1.037*** (1.776) (1.699) N 2363 5148
Table 3: Correlations between firm characteristics before the supplier consolidation
The correlations of variables listed in table 2 are shown below. * denotes significance at the 5% level
COGS/Sales EBITDA/Sales ln(Sales) ln(Asset) Debt/Assets Return on Assets ln(employee)
COGS/Sales 1 EBITDA/Sales -0.651* 1 ln(Sales) 0.288* 0.032* 1 ln(Asset) 0.027* 0.275* 0.899* 1 Debt/Assets 0.079* 0.126* 0.153* 0.272* 1 Return on Assets -0.176* 0.459* 0.228* 0.175* -0.221* 1 ln(employee) 0.154* 0.020 0.848* 0.782* 0.202* 0.225* 1
As table 3 shows, there are strong correlations among ln(Assets), ln(Sales), and ln(Employee). Despite the fact that not all increases in assets and employees are contributed to the production activity, a one unit increase in Assets or Employees results in more than 80 percent increase in the total Sales. In this case, it is reasonable to use these three estimates to measure the size of a firm. Table 3 also shows a negative correlation between COGS/Sales and firm profitability, as higher cost is expected to harm profitability, while the COGS/Sales increased with the increase of sales. The reason could be that larger firms tend to be more dependent on their supplier. According to Ahern, K. (2011), the higher the customer dependency is, the larger the market power of the supplier is. Therefore, suppliers are able to charge a higher price from larger customer firms as a result of a higher COGS in larger customers. The reason can also be that the extra output produced by firm extension is not always sold right away. It is time consuming to explore the customer base. Therefore, although the Sales and COGS both increase as the firm extends the production, the increase of COGS will occur earlier and faster than the enhancement of sales. Besides, table 3 indicates that the firm size is positively associated with firm profitability. This relationship can be interpreted differently. For example, larger firms are able to achieve the economics of scale, resulting in a higher profitability. Or, larger firms have larger bargaining powers which can enforce the supplier to lower the price. Furthermore, larger firms have larger market powers which enable them to increase the price and result in higher revenues.
5. Methodology
M&A activities do not happen to firms in an entirely random way. Most of the time, there are certain motivations behind them, for example realization of synergy or exploration of a new geographical market. Hence, in these ‘as if’ randomness experiments, the difference-in-difference (DID) method is used. Instead of identifying all the control variables which can affect the COGS of the merged firm’s customers over time, a control group is defined for firms with similar characteristics and market conditions as the merged firm’s customers. The intuition of identifying the treatment group and the control group is shown in figure 1 below.
Figure 1: The distribution of the control group and the treatment group
In figure 1, a product market with an upstream industry and a downstream industry is assumed. The process of the treatment group and the control group identification is demonstrated.
Si (i=1,2,3,4,5): Firms from the supplier industry, with Si (i=1,2,3,4,5) as competitors of each other Ci (i=1,2,3,4,5,6,7,8): Firms from the downstream industry, with Ci (i=1,2,3,4,5,6,7,8) as competitors of each other
The firms, Si (i=1,2,3,4,5), in the upstream industry supply the firms, Ci (i=1,2,3,4,5,6,7,8), in the downstream industry. Some of the customer firms share the same supplier, e.g. S1 is the supplier of C1 and C2; S3 is the supplier of C4 and C5; C7 and C8 are the customer firms of S5, while some of the customer firms are supplied by different upstream industry firms, e.g. S2 is the supplier of C3, C6 is the customer firm of S4. For same reasons, S1 and S2 are merged. Therefore, C1, C2, and C3 are the customers of merged firm S12. Firms like C1, C2, and C3 which are customer of merged firms are distributed into the treatment group. Other downstream industry firms such as C4, C5, C6, C7 and C8 whose suppliers S3, S4, and S5 did not participate in any transaction are categorized into the control group.
Ideally, the only difference between the treatment group and control group is whether the supplier does horizontal transaction. Since firms from the same industry can still vary in many aspects, in order to minimize these differences, only the three closest competitors of each customer of the merged firm are included in the control group.
In reality, differences exist in firm-panel data both between firms and across time. Therefore, the firm fixed effects via GVKEY and the time fixed effect via years are included in the methodology. The firm fixed effect is used to control those unobserved variables that vary from one company to the other but do not change over time, for example company cultures. The time fixed effect is to control those unobserved variables that evolve over time but are constant across firms, such as economic conditions.
Difference in Difference (DID) method means two differences occur. The first difference in this methodology is the three-year change of firm characteristics after the supplier consolidation. Due to the price decline and quantity convergence in the merged firm only being realized two to three years after the transaction, a three-year post-M&A change of firm characteristics (COGS, profitability, and firm size) will be calculated for firms in both the treatment group and control group. Since COGS and EBITDA will increase as sales increase, they are scaled by Sales. The formulas of variable calculations are listed below. The second difference in this methodology is the difference between the treatment group and control group.
∆(#$%&&'()*),-= (#$%&&'()*), /0- − (#$%&&'()*),/ (1) ∆(345678 &'()* ),-= ( 345678 &'()* ), /0- − ( 345678 &'()* ),/ (2) ∆(Assets),-= ln(Assets), /0- − Ln(Assets),/ (3) ∆(Sales),- = ln(Sales), /0- − Ln(Sales),/ (4) ∆(Employee),-= ln(Employee), /0- − ln(Employee),/ (5) ∆(ROA),-= ln(ROA), /0- − ln(ROA),/ (6) ∆(Debt/Assets),- = ln(Debt/Assets), /0- − ln(Debt/Assets),/ (7) i: firms of the downstream industry, including firms from both sample groups
t: for firm i from the treatment group, t is the time when the supplier of firm i participates in horizontal merger; for firm i from the control group, t is the time when the supplier of the firms in the treatment group corresponding to firm i merged.
As table 2 shows, the firm characteristics of the treatment group and the control group are quite different. However, in the DID method, what matters is that the changes of firm characteristics before the treatment are not different between the treatment group and the control group. Therefore, a test of this common trend assumption had to be carried. The summary statistics of three-year firm characteristic changes before the supplier consolidation is shown in table 4, columns 2, 3, and 4; and the summary statistic of three-year firm characteristic changes after the supplier consolidation is shown in table 4, columns 5, 6, and 7.
Table 4: three-year change of firm characteristics before and after the supplier consolidation
This table reports summary statistic of three-year change of firm characteristics before (Column 2, 3, and 4) and after (Column 5, 6, and 7) seller industry consolidation for both the treatment group and control group. The mean and standard deviation (in parentheses) for each variable are reported separately for both the treatment group and control group. Column (1) lists all the variables. Column (2) reports the three-year change of the firm characteristics of merged firm’s customers before the
consolidation of the seller industry. Column (3) reports estimates from the time period for firms who are the competitors of the merged firm’s customers whose supplier did not participate in any M&A. Column (4) shows the difference between the three-year firm characteristic change of treatment group and control group before the treatment. Column (5) reports the three-year changes of firm characteristics of merged firm’s customers after the consolidation of the seller industry. Column (6) reports estimates from the time period for firms who are the competitors of the merged firm’s customer whose supplier did not participate in any M&A. Column (7) shows the difference between the three-year firm characteristic change of the treatment group and control group after the treatment. T-tests of the difference are run, ***denotes significance at the 1% level; **denotes significance at the 5% level.
(1) (2) (3) (4) (5) (6) (7)
Before treatment (before supplier consolidation)
After treatment (after supplier consolidation) Treatment group (customers of merger firms) Control group (Competitors of the treatment group) The difference Treatment group (customers of merger firms) Control group (Competitors of the treatment group) The difference
3-year change of COGS/Sales -0.004 -0.010 0.005** 0.002 -0.001 0.003**
(0.075) (0.089) (0.061) (0.074)
3-year change of EBITDA/Sales 0.002 0.004 -0.002 -0.006 -0.005 -0.001
(0.069) (0.889) (0.067) (0.082)
3-year change of Sales -0.178 -0.169 -0.009 0.178 0.169 0.009
(0.378) (0.429) (0.378) (0.429)
3-year change of Assets 0.312 0.309 0.003 0.224 0.203 0.021**
(0.401) (0.442) (0.382) (0.398)
3-year change of Employee 0.155 0.158 -0.002 0.079 0.075 0.004
(0.391) (0.445) (0.337) (0.386)
3-year change of Return on Assets -0.007 -0.006 -0.002 -0.012 -0.007 -0.005* (0.075) (0.912) (0.083) (0.109)
3-year change of D/A 0.004 0.009 -0.005 0.010 0.014 -0.004
(0.087) (0.101) (0.086) (0.101)
Table 4 reports the three-year firm characteristic changes of downstream firms before and after the consolidation of the upstream industry. As column (4) shows, the estimates of the treatment group and the control group are similar before the supplier consolidation, except that the difference in COGS/Sales is significant at the 5% level. Both the treatment group and control group have a decline in COGS/Sales in the three years before the supplier consolidation, while the decline of COGS/Sales in the treatment group is smaller than the decline in the control group. As table 2 shows, the average size of firms in the treatment group is larger than the average size of firms in the control group. Larger firms tend to have a higher dependence on the supplier; for example, the cost of switching suppliers is higher for larger firms. As a result, larger firms have less bargaining power than smaller firms. There could be another reason for higher COGS/Sales in the treatment group before the treatment. Since the suppliers of the treatment group and the control group are different, the suppliers of the treatment group may already have a higher market power before they do horizontal mergers. These certain characteristics of the future merged firms may also drive them to do transactions in the future. To a large extent, the common trend assumption of DID is satisfied.
As columns (5) and (6) show, the after-treatment estimates of the two sample groups share the same changing director, except for the change of COGS/Sales. The COGS/Sales of the merged firm’s customers increases after the suppliers consolidation, while the COGS/Sales of the competitors of merged firm’s customers decreases in the same period. As column (7) shows, the differences in COGS/Sales change of two sample groups are significant at the 5% level. This phenomenon is consistent with the market power hypothesis that horizontal mergers facilitate the collusion between merged firms. With increasing market powers, the mergers are able to charge a relatively higher output price. However, according to column (4), even before the supplier consolidation, the firms in the treatment group already have a relatively higher COGS/Sales than the firms in the control group. Therefore, it is hard to tell if the increase of COGS/Sales is caused by the upstream industry consolidation.
Besides, by comparing column (7) with column (4), we can see that the differences in asset change of two sample groups are insignificant before the transactions while becoming more significant after the transactions. The firms in the treatment group seem to have a higher increase in total assets than their competitors. The reason could be that the supplier consolidation triggers the consolidation among the corporate customers of merged firms. Just as what Galbraith’s (1952) theory of countervailing acquisitions argues, seller industry concentration induces buyers to grow large in order to neutralize the market power of sellers. It is also complying with the finding of Ahern and Harford (2011) that merge waves can be passed from the upstream industry to the downstream industry. In addition, the average Return on Assets of the treatment group is also lower than the average Return on Assets of
the control group. There are some potential reasons. For example, the increase in COGS reduces the net income, or the increase in total assets dilutes the returns.
Figure 2: Firm characteristic developing trends from three years before treatment until three years after treatment
The 6 sub-graphs in figure 2 show the comparisons of firm characteristics (COGS/Sales, EBITDA/Sales, Return on Assets, ln(Assets), ln(Sales), ln(Employee)) developing trends between the treatment group (customer of merged firms) and the control group (competitor of merged firm’s customers) over the period from three years before to three years after the treatment (supplier horizontal mergers). The number in horizontal shows the time line. 0 is the time when the suppliers of the treatment group participate in horizontal merger. -3, -2, and -1 represent 3, 2, and 1 years before the transaction; 3, 2, and 1 represent 3, 2, 1 years after the transaction. The vertical shows the ratio amount.
(1) (2) (3) (4) (5) (6) 0.62 0.63 0.64 0.65 0.66 0.67 0.68 0.69 0.7 -3 -2 -1 0 1 2 3
Cost of Goods Sold
Customer of merged firms Competitor of merger' customers 0.138 0.14 0.142 0.144 0.146 0.148 0.15 0.152 0.154 -3 -2 -1 0 1 2 3 EBITDA/Sales Customer of merged firms Competitor of merger' customers 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 -3 -2 -1 0 -1 2 3 Return on Assets Customer of merged firms Competitor of merger' customers 8.5 9 9.5 10 10.5 11 -3 -2 -1 0 -1 2 3 ln(Sales) Customer of merged firms Competitor of merger' customers 8 8.5 9 9.5 10 10.5 11 -3 -2 -1 0 -1 2 3 ln(Assets) Customer of merged firms Competitor of merger' customers 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 -3 -2 -1 0 -1 2 3 ln(Employee) Customer of merged firms Competitor of merger' customers
As the sub-graph (1) shows, the customers of merged firms always have higher COGS/Sales than their competitors. In this (-3, +3) window, the COGS/Sales of the control group and treatment group develop in a similar pattern. While after the treatment, the COGS/Sales of the treatment group evolves in a steeper trend than the control group does. In other words, after the supplier consolidation, the COGS/Sales of the customers of merged firms appears to increase.
According to sub-graph (2) which reports the development of EBITDA/Sales, the treatment group and control group roughly share the same trend. Before the treatment, the treatment group has a slight EBITDA/Sales advantage over the control group. But one year after the treatment, the treatment group had a lower EBITDA/Sales than the control group. In other words, corresponding to sub-graph (2), sub-graph (3) indicates that after the supplier consolidation, the Return on Assets of the treatment group became lower than the Return on Assets of the control group. In terms of firm size, as sub-graph (4), sub-sub-graph (5), and sub-sub-graph (6) show, it is obvious that the firms in the treatment group have larger firm sizes than the firms in the control group during the whole event window. In addition, both the treatment group and the control group tend to increase their firm sizes gradually.
Although the DID method has removed many differences between two sample groups by applying the difference twice, the firms still differ within the same sample group. To minimize these noises, the firm fixed effect and time fixed effect still need to be included. The regression used by this thesis is showed below and the regression results are reported in the table 5.
A DID method with entity and time fixed effects:
∆Y,- = 𝛽N+ 𝛽P𝐷 + 𝛼,+ 𝜑/+ 𝜀,/
∆Y,-: the three-year post-merger change of firm characteristics, including ∆(#$%& &'()*),-, ∆(
345678 &'()* ),-, ∆(Assets),-, ∆(Sales),-, and ∆(Employee),-.
D: a dummy. D = 1 for firms in the treatment group (customer of merged firms); D = 0 for firms in the control group (competitor of merged firm’s customers)
𝛼,: firm fixed effect. 𝜑/: time fixed effect. 𝜀,/: residual.
Table 5: Effects of supplier horizontal acquisition on customer firms
This table reports the coefficients from firm-panel regressions of change of firm characteristics on an indicator of whether the supplier does transaction. Firm fixed effects via GVKEY and time fixed effects via year are included. The dependent variables are the three-year change of COGS/Sales
(column 2), three-year change of EBITDA/Sales (column 3), three-year change of ln(Sales) (column 4), three-year change of ln(Assets) (column 5), and three-year change of ln(Employee) (column 6). *** denotes significance at the 1% level; ∗∗ denotes significance at the 5% level.
(1) (2) (3) (4) (5) (6) 3-year change of COGS/Sales 3-year change of EBITDA/Sales 3-year change of Sales 3-year change of Assets 3-year change of Employee D(Customers of merger firms) 0.005** -0.007*** 0.002 -0.002 0.003
SE (0.002) (0.003) (0.010) (0.012) (0.014)
Firm fixed effect Yes Yes Yes Yes Yes
Time fixed effect Yes Yes Yes Yes Yes
N 6,268 6,268 6,268 6,268 6,268
R^2 0.447 0.461 0.712 0.548 0.495
As column (2) shows, supplier horizontal acquisitions cause the average COGS of merged firm’s customers to increase by 0.5 percentage points relative to the competitors of these customers. This increase in the COGS, which is significant at the 5% level, complies with the market power hypothesis that horizontal mergers increase the market power of suppliers and enable these merged firms to charge more from the customer firms. Correspondingly, column (3) shows that the average profitability declined by about 0.7 percentage points among the merged firm’s customers relative to other competitors. Since the coefficient in column (4) is insignificant, the changes of sales in merged firm’s customers and the changes of sales in the competitors of these costumers are not very different. Therefore, the profitability dropping of merged firm’s customers is most likely caused by the COGS changes. As column (5) and (6) show, the average changes in assets and employees are approximately identical between the treatment group and the control group. This may indicate that there are not big transactions happening in the downstream industry after the upstream industry consolidation.
6. Robustness checks
• Subgroup tests via industry:
The transaction effects can also vary in different industry settings. Therefore, the sample is split into different subgroups in terms of the supplier macro-industries which are defined by Thomson One. Table 6 reports the summary statistics of each subgroup, and table 7 shows the regression results by subgroup.
Table 6: summary statistics of subgroups
This table reports the summary statistics of each subgroup. There are 10 subgroups which include supplier Macro-Industry Consumer Products and Services, Consumer Staples, Energy and Power, Healthcare, High Technology, Industrials, Materials, Media and Entertainment, Retail, and Telecommunications industries. Column (2) shows the number of firms in the treatment group and Column (3) shows the number of firms in the control group.
(1) (2) (3) Supplier macro-industry Number of firms in the treatment group (customers of merger firms) Number of firms in the control group (Competitors of the
merged firm’s customers)
Consumer Products and Services 121 329
Consumer Staples 171 546
Energy and Power 248 626
Healthcare 397 644
High Technology 826 1757
Industrials 353 690
Materials 88 222
Media and Entertainment 36 107
Retail 13 23
Telecommunications 110 204
N 2363 5148
As the table shows, the high technology group has the biggest sample size, and there are the least samples in the retail group.
Table 7: Effects of supplier horizontal acquisition on customer firms by subgroup
The table reports the coefficients from firm-panel regressions of change of firm characteristics on an indicator of whether the supplier does the transaction in terms of the supplier macro-industry. Firm fixed effects via GVKEY and time fixed effects via year are included. The dependent variables are the year change of COGS/Sales (column 2), year change of EBITDA/Sales (column 3), three-year change of ln(Sales) (column 4), three-three-year change of ln(Assets) (column 5), and three-three-year
change of ln(Employee) (column 6), The standard errors are showed in parentheses. *** denotes significance at the 1% level; ∗∗ denotes significance at the 5% level.
(1) (2) (3) (4) (5) (6) 3-year change of COGS/Sales 3-year change of EBITDA/Sales 3-year change of Sales 3-year change of Assets 3-year change of Employee Consumer Products and
Services 0.038 -0.011 0.094* 0.220** 0.109 (0.055) (0.011) (0.682) (0.091) (0.060) R^2 0.686 0.578 0.700 0.589 0.620 Consumer Staples -0.003 0.003 -0.030 -0.019 -0.036 (0.004) (0.004) (0.027) (0.041) (0.040) R^2 0.280 0.249 0.553 0.482 0.389
Energy and Power 0.003 0.000 0.059 0.049* 0.038
(0.009) (0.010) (0.043) (0.037) (0.038) R^2 0.343 0.389 0.730 0.584 0.482 Healthcare -0.007*** 0.005 0.000 -0.004 0.002 (0.004) (0.004) (0.023) (0.027) (0.024) R^2 0.452 0.528 0.566 0.471 0.424 High Technology -0.003 0.003 -0.002 0.018 -0.005 (0.003) (0.004) (0.015) (0.019) (0.018) R^2 0.268 0.256 0.556 0.439 0.502 Industrials 0.005 -0.003 0.009 0.001 0.027 (0.005) (0.004) (0.025) (0.029) (0.023) R^2 0.305 0.397 0.592 0.539 0.562 Materials 0.004 -0.007 -0.032 0.032 -0.090 (0.020) (0.015) (0.063) (0.088) (0.053) R^2 0.538 0.563 0.682 0.644 0.590
Media and Entertainment -0.008 -0.003 -0.050 0.031 -0.002
(0.008) (0.006) (0.038) (0.060) (0.060)
R^2 0.790 0.756 0.800 0.656 0.757
Telecommunications 0.025 -0.015 -0.047 -0.165* -0.085
(0.019) (0.012) (0.069) (0.095) (0.066)
As table 7 reports, the estimates vary across the subgroups. As column (2) shows, the consolidation in the Consumer Products and Services, Consumer Staples, Healthcare, and High Technology industries decreases the COGS of customers of merged firms relative to competitors of these customers. Besides, only for the healthcare industry, the coefficient of the three-year change of COGS/Sales is significant at the 1% level. The horizontal mergers which happened in the Energy and Power, Industrials, Materials, Media and Entertainment, and Retail industries appear to increase the COGS of the merged firm’s customers, but none of the estimates are significant.
In the subgroup of Consumer Products and Services, the coefficients of change in firm assets and change in sales are statistically significant. It seems that the transactions in the Consumer Products and Services industry drive the customer firms to also consolidate, maybe in order to resist the increasing market powers of suppliers. Besides, the customers of merged firms in the Energy and Power, and Telecommunications industries also appear to increase their assets after their suppliers consolidated.
• Industry concentration & product differentiation
The industry concentration degrees and the product differentiation levels vary by industry. Both concentration degree and product differentiation level can not only drive the incidence of horizontal transactions, but also impact the outcome of horizontal transactions. Regarding the Cournot oligopoly model (Gehrig, 1981), industry concentration is positively associated with pricing power. According to Chamberlin theory (Chamberlin E. H. 1949), product similarities of firms are negatively related to their pricing power.
The industry concentration level and product similarities are taken into consideration in the robustness checks. The industry concentration level can be measured by Herfindahl-Hirschman Index (HHI) and product similarities can be measured in terms of the product descriptions. Both HHI data and similarity scores are provided by the database of Hoberg and Phillips. After including the concentration levels and product similarities of the upstream industries as additional fixed effects, the estimates become insignificant while the R^2 increased. These changes suggest that the model is better fit for the sample when the concentration levels and product similarities of the upstream industries are taken into account.
7. Conclusion
Most existing literature emphasizes examining the impact of M&A on the merged firms and tests the impacts of M&A by investigating general firm performance. This thesis differs by assessing the impact of supplier consolidation on the buyer levels by using Cost of Goods Sold (COGS), which is
an explicit indicator of customer-supplier relationships that has not been tested in other literature. In addition, an algorithm and a special database which measured the firm similarities based on product descriptions are used to prepare the dataset.
In contrast with the unpronounced discovery of Fee and Thomas (2004), the results of this thesis appear to support both the collusion hypothesis and countervailing acquisitions hypothesis. The finding that horizontal mergers increase the COGS in the merged firm’s buyers complies with the collusion hypothesis that horizontal mergers increase the market power of merged firms. The finding that the size of the merged firm’s customers is larger than the size of the unmerged firm’s customers is consistent with the countervailing acquisitions hypothesis that supplier horizontal mergers are driven by the downstream industry’s previous consolidation. This countervailing acquisitions finding complies with existing literature such as the discovery of merger wave transmission by Ahern and Harford (2011).
This thesis contributes to the growing research field of the interaction between corporate decisions and industry linkages. After understanding how the impacts of firm activities travel through those dynamic economic networks and eventually land on different entities, a corporate decision which maximizes stakeholder utility is more likely to be made. The finding of this thesis helps firms to think forward and be prepared for the coming impact of activities of the upstream industry. Furthermore, this thesis which views the outcome of transformational transactions from a new perspective and discovers an inherent phenomenon offers financial advisors a new entry point of giving advice. The motivations of horizontal mergers and the achievements of integration are different case-by-case. The settings of the industry and the relationships between customers and suppliers vary per transaction. To some extent, these factors can also determine the impact of transformational transactions on stakeholders. To take a cautionary stance, an immense sample which across industries and regions may not be ideal to test the effects of M&A. For further investigations, the sample can be distributed in terms of industry concentration or product differentiation levels and then tested. Furthermore, since both the customer-supplier and competitor relationship are dynamic, the methodology can be optimized by considering the changes of industry linkages.
Reference
Ahern, K. (2011). Bargaining Power and Industry Dependence in Mergers. Journal of Financial Economics, 103 (2012), pp. 530–550.
Ahern, K. and Harford, J. (2011). The Importance of Industry Links in Merger Waves. Journal of Finance. Vol, LXIX, NO. 2, April 2014.
Allayannis, G. and Ihrig, J. (2001). Exposure and Markups. Review of Financial Studies, Vol. 14, No. 3, pp. 805–835.
Ashenfelter, O. Hosken, D. and Weinberg, M. (2009). Generating Evidence to Guide Merger
Enforcement. National Buereau of Economic Research 1050 Massachusetts Avenue Cambridge, MA
02138.
Bhattacharyya, S. and Nain, A. (2010). Horizontal Acquisitions and Buying Power: A Product Market Analysis. Journal of Financial Economics, 99 (2011), pp. 97–115.
Banerjee, S. Dasgupta, S. and Kim, Y. (2008). Buyer – Supplier Relationships and the
Stakeholder Theory of Capital Structure. The Journal of Finance. Vol. LXIII, NO. 5. October 2008
Chamberlin E. H. (1949). The theory of monopolistic competition: a re-orientation of the theory of value. Oxford University Press. 6th ed 1949
Chouinard, Hayley and Jeffrey M. Perloff, (2009). Gasoline Price Differences: Taxes,
Pollution Regulations, Mergers, Market Power, and Market Conditions. The B.E. Journal of
Economic Analysis and Policy,
Eckbo, B. (1983). Horizontal Mergers, Collusion, and Stockholder Wealth. Journal of Financial Economics, 11 (1983), pp. 241–273.
Fee, E. and Thomas, S. (2004). Sources of Gains in Horizontal Mergers: Evidence from Customer, Supplier, and Rival Firms. Journal of Financial Economics, 74 (2004), pp. 423–460.
Gehrig, W, (1981). On the Complete Solution of the Linear Cournot Oligopoly Model, The Review of Economic Studies, 1 October 1981, Vol.48(4), pp.667-670 [Peer Reviewed Journal]
Hoberg, G. and Phillips, G. (2016). Text-Based Network Industries and Endogenous Product
Differentiation. Journal of Political Economy 124 (5), 1423-1465.
Hoberg, G. and Phillips, G. (2010). Product Market Synergies and Competition in Mergers
and Acquisitions: A Text-Based Analysis. Review of Financial Studies 23 (10), 3773-3811.
Kim. Han. and Singal, V. Mergers and Market Power: Evidence from the U.S. Airline
Industry. American Economic Review, 1993, 83, 549–569.
Moeller, Sara B. Frederik, P. Schlingemann. and Rene M, Stulz. (2004). Firm size and the
gains from acquisitions. Journal of Financial Economics 73, 201–228.
Pittman, R. (2007). Consumer Surplus as the Appropriate Standard for Antitrust Enforcement. Technical Report
Rhodes, M. and Robinson, D. (2008). The Market for Mergers and the Boundaries of the
Firm. The Journal of Finance. Vol. LXIII, NO. 3, June 2008.
Shahrur, H. (2004). Industry Structure and Horizontal Takeovers: Analysis of Wealth Effects
on Rivals, Suppliers, and Corporate Customers. Journal of Financial Economics. 76 (2005), pp.
61–98.
Sheen, A. (2014). The Real Product Market Impact of Mergers. The Journal of Finance. Vol. LXIX, NO. 6, December 2014.
Simpson, J. & Christopher, T. (2008). Do Gasoline Mergers Affect Consumer Prices? The
Marathon Ashland Petroleum and Ultramar Diamond Shamrock Transaction. Journal of Law and
Economics. 51 (1), 135–52 .
Stillman, R., 1983. Examining antitrust policy towards horizontal mergers. Journal of Financial Economics 11, 224–240.
Ulrike. M, Enrico. M, and Florian S. P (2012). Winning by Losing: Evidence on The
Long-Run Effects of Mergers. Nber Working Paper Series. Working Paper 18024.
