Theory Meets Practice in Merger Control Koki Arai Senior Researcher for Economic Analysis Japan Fair Trade Commission 1st ATE Symposium on Competition

Theory Meets Practice in Merger Control

Koki AraiSenior Researcher for Economic Analysis

Japan Fair Trade Commission

1st ATE Symposium onCompetition Policy Issues: Theory Meets Practice

The views in this presentation are my own, not any organizations.

December 13, 2013Massey University

１

Outline

• Theory Meets Practice in Merger Control• Case Study: ASML / Cymer• Ex-post Evaluation of the effects of mergers

２


• There are many studies on a merger.• Many of them include policy implications.

Difficult to use a research that indicate probability of

welfare increase or decrease through pure theory

Easy to explain analysis that describe an incentive

mechanism with a few examples

Who is user? ３


• In merger control discussion, theory meets lawyers and businesses.

For lawyers, it needs to give simple but reasonable economic explanation.

For businesses, it needs to enforce rules based on evidences concerning ex-post assessment.

４


For lawyers, it needs to give simple but reasonable explanation.

– A case study in Japan: Discussion of vertical merger and its remedy

– How to explain anticompetitive effect– New measures to be theorized

５



– Empirical studies of mergers’ effects and implications from competition policy

– Research from widely viewpoints

CR04-11: CPRC Report (The Competition Policy Research Center)

６

Case StudyASML / Cymer

I. Introduction• ASML US, the subsidiary of ASML Holdings N. V.

that runs business of manufacturing and selling lithography systems used in the front-end process of semiconductor manufacturing, is planning to acquire all the shares of Cymer which runs business of manufacturing and selling light sources composing an important part of the lithography system.

７

８

II. Particular field of trade1. Upstream market(light source)

(1) Product Range– KrF light source, and– ArF light source– (EUV lithography systems are under technical

challenge)

(2) Geographic Range– the whole world

• Lithography system manufacturers and distributors, domestic and overseas light source users give non-discriminatory treatment to domestic and overseas light source manufacturers. ９★

II.Particular field of trade1. Upstream market(light source)

(1) Product range• Lightsource, a device that generates laser beams, is one of the essential and

important parts of lithography systems as mentioned in II-2,below,and is used to print electronic circuits on wafers. The light source in which the parties currently have transaction is DUV (Deep Ultraviolet Light) light source.DUV light source can be divided into two major types: KrF light source and ArF light source.

• Telling of a general nature of light sources, the shorter wavelength it generates, the higher resolution performance it achieves that enables print circuits to be done in more microscopic bandwidth. With regard to the wave lengthof the light source, KrF light sources have wavelength light of about 248nano meter (hereinafter “nm”) and ArF light sources have wavelength light of about 193nm. Light sources with longer wavelength light are used to print circuits with broad bandwidth. Light sources with shorter wavelength light are used to print circuits with narrow bandwidth.


(1) Product range• Although there is another type of light source besides DUV

light source called EUV(Extreme Ultraviolet Light)light source which has wavelength light of about 13.5nm, EUV light sources and EUV lithography systems are under technical challenge. Therefore, current sales of EUV light sources are marginal and made only for research and development purposes.

• As mentioned above, due to the differences of resolution performances and price ranges between KrF light sources and ArF light sources, users which are manufacturers of lithography systems do not recognize KrF light sources and ArF light sources as substitutable. Therefore, the JFTC defined one product range as “KrF light sources” and another product range as “ArF light sources” both are separately subject to its review.


(2) Geographic range• Light source manufacturers and retailers(hereinafter “light

source manufacturers”) sell their light sources at a substantially same price all over the world. Moreover, lithography system manufacturers and distributors, domestic and overseas light source users, give non-discriminatory treatment to domestic and overseas light source manufacturers.

• Therefore, for each of the light sources as defined in (1), above, the geographic range is, respectively for each, defined as “the whole world”.

II. Particular field of trade2. Downstream market(lithography system)

(1) Product Range– “KrF lithography systems,” – “ArF lithography systems,” and – “ArF immersion lithography systems”

• Geographic Range– the whole world

１０

★


(1) Product range• Lithography system is a device that makes an image of

electronic circuit patterns(circuit original plate)in reduced size, projected through its lens and prints the image on a wafer which is the basic structure of semiconductor integrated circuits. – In case of light sources, one with shorter wavelength light has

higher resolution performance that enables the light sources to print circuits with narrow bandwidth, a lithography system with ArF light source called “immersion lithography system” exists.

– This lithography is designed to enhance high resolution via application of refraction index of water created when the area between the lens and wafer is immersed with water.


(1) Product range• With respect to resolution performances of

lithography systems by the light source, the lithography system attached with KrF light source is capable of resolution performance of approximately 100-250 nm, the lithography system attached with ArF light source is capable of approximately 65-90 nm and the immersion lithography system attached with ArF light source is capable of approximately 45-65nm.

• Therefore, the resolution performance of ArF immersion lithography system is the highest among the lithography systems.


(1) Product range• With respect to KrF lithography systems, ArF lithography

systems and ArF immersion lithography systems, since there are differences between resolution performances and price ranges among them, the substitutability for semiconductor manufacturers and distributors and semiconductor manufactures to produce by order which are customers of lithography systems does not exist.

• Therefore, the JFTC defined the product ranges as “KrF lithography systems”, “ArF lithography systems” and “ArF immersion lithography systems” individually for each.

　★ １１


(1) Product range• Nonetheless, chipmakers which are customers

of lithography systems can freely choose any light sources manufactured by each of light source manufacturers when they purchase lithography systems.


(2) Geographic range• Lithography system manufacturers sell

lithography systems at substantially same price all over the world. Chipmakers which are domestic and overseas users give non-discriminatory treatment to domestic and overseas lithography system manufacturers. Therefore “the whole world” is individually defined as a geographic range.

１２

III. Review concerning substantial restraint of competition 1. The status of the parties and the competitive situation

(1) Upstream market(light source)• KrF: HHI - 5,300. • ArF: HHI - 6,300. (2) Downstream market (lithography system)• KrF lithography systems: HHI - 8,300. • ArF lithography systems: HHI - 5,100. • ArF immersion lithography systems: HHI - 7,500.

★ １３


(1) Upstream market(light source)• In the market for KrF light sources, the market share of

Cymer would be approximately 60% (ranked in the first in the market) and the HHI would be approximately 5,300.

• In the market for ArF light sources, the market share of the parties would be approximately 75% (ranked in the first in the market) and the HHI would be approximately6,300.

• Therefore, both products do not meet the safe harbor standards for vertical business combinations.

• Company A (a domestic manufacturer) is the only competitor of Cymer.

１４

★


(2) Downstream market (lithography system)• In the market for KrF lithography systems, the market share of ASML

would be approximately90% (ranked in the first in the market) and the HHI would be approximately 8,300.

• In the market for ArF lithography systems, the market share of the parties would be approximately45% (ranked in the second in the market) and the HHI would be about 5,100.

• In the market for ArF immersion lithography systems, the market share of the parties would be approximately85% (ranked in the first in the market) and the HHI would be approximately 7,500.

• Therefore, all products do not meet the safe harbor standards for vertical business combinations.

• With respect to KrF lithography systems, Company X and Company Y (both of them are domestic manufacturers) are the only competitors of ASML. With respect to ArF lithography systems and ArF immersion lithography systems, Company X is the only competitor of ASML.

１５

III. Review concerning substantial restraint of competition

Three discussion points:i) Refusal of sale, etc. of light sources

transactionii) Refusal of purchase, etc. of lithography

systems transactioniii) Access to confidential information

１６

III. Review concerning substantial restraint of competition2. Refusal of sale, etc. of light sources transaction

(1) Impact of refusal of sale• In the downstream market, Company X and Company Y

which manufacture and distribute KrF lithography systems, ArF lithography systems or ArF immersion lithography systems procure an appreciable extent of KrF light sources or ArF light sources from Cymer of the upstream market.

• As a result of the Acquisition, in case where Company X or Company Y are deprived of an opportunity to deal with Cymer or in case where Company X or Company Y is disadvantageously treated in transactions compared with ASML, Company X or Company Y are placed in a disadvantageous situation and there are some possibilities of resulting in market foreclosure or exclusivity. １

７


(1) Impact of refusal of sale• Cymer occupies a high market share of the

upstream market and there are few competitors in the upstream market.

• Therefore, if Cymer substantially sells light sources exclusively to ASML, and thus the competitors in the downstream market lose the primary procurement sources of light sources and result in market foreclosure or exclusivity, it is considered that such situation has a large impact on competition in the downstream market. １

８

１９

Possibility of Refusal of Sales

１９


(2) Allegations of the parties and assessments thereofA. Allegations of the parties• According to the parties’ claim, upon selling lithography

systems, as to a light source which constitutes an important part of lithography systems, whereas it is chipmakers who decide to choose which light source of which light source manufacturer, if the parties engaged in input foreclosure, the parties lose not only their light source profit causes but also lose trust from chipmakers and that leads to have impact on ASML’s lithography sales.

• Therefore, the parties claimed that input foreclosures provide no incentive for them.

２０


(2) Allegations of the parties and assessments thereof

B. Review and assessment of the allegations of the parties

• It is chipmakers that purchase lithography systems and choose light sources attached to lithography systems.

• According to the following facts, it is considered that chipmakers have countervailing power to a certain degree against the input foreclosure by the parties: ★ ２

１


– (i) chipmakers state that, should the parties exercise an input foreclosure after the Acquisition, chipmakers are still able to give their opinions regarding the choice of light source manufactures to the parties since the state where multiple choices of light sources are retained contributes to price and performance competition;

– (ii) most of the sales of the parties are occupied by several major chipmakers and

– (iii) the development of lithography systems and light sources are carried out according to the roadmap of the whole semiconductor industry that includes such as chipmakers.


(3)Measures proposed by ASML US• After the JFTC explained to ASML US as saying

that such input foreclosure might be a point potentially to argue in the review of the Acquisition, ASML US has proposed that it would take the following measures against the concern of the input foreclosure.

２２


(3)Measures proposed by ASML US– (i) With respect to DUV light sources, Cymer will continuously

do business with Company X and Company Y under fair, reasonable and non-discriminatory terms of trade as well as in the manner of paying regard to and being consistent with the existing agreements. Moreover, with respect to EUV light sources, after the Acquisition, Cymer will do business with Company X and Company Y under fair, reasonable and non-discriminatory terms of trade as well as in the manner of paying regard to and being consistent with the industry standard.

– (ii) Cymer will implement joint development activities with Company X and with Company Y under the reasonable terms of trade. With respect to DUV light sources, Cymer will implement it in the manner consistent with the existing agreements.

★ ２３


(3)Measures proposed by ASML US– (iii) For five years from the execution of the

Acquisition, the parties will report the status of compliance with the measures mentioned above to the JFTC once a year.

– (iv) The report mentioned (iii) is to be created by an audit team independent from parties, which will be appointed subject to a prior approval of the JFTC.


(4) Assessment under the AMA• The measures proposed by ASML US mentioned (3), above, are as

follows: Cymer will continuously deal with Company X and Company Y in a manner consistent with the terms of trade equivalent to that of prior to the Acquisition. Moreover, an audit team independent of the parties’, which will be appointed subject to a prior approval of the JFTC, conducts an audit and Cymer will report to the JFTC regarding the result of audit for a certain period of time after the Acquisition, thus the effectiveness of the measures will be ensured. Moreover, as mentioned in (2) B above, there is competitive pressure from chipmakers to a certain degree.

• Therefore, taking the measures proposed by ASML US, etc. into consideration, the Acquisition will not cause the input foreclosure.

２４

III. Review concerning substantial restraint of competition3. Refusal of purchase, etc. of lithography systems transaction

(1) Impact of refusal of purchase on competition• In the upstream market, Company A which runs

business of manufacturing and selling KrF light sources and ArF light sources sells an appreciable extent of KrF light sources or ArF light sources to ASML of the downstream market.

• As a result of the Acquisition, there is a possibility of placing Company A in a disadvantageous situation and resulting in market foreclosure or exclusivity, in case where Company A is deprived of an opportunity to deal with ASML or Company A is treated disadvantageously in transactions compared to that of Cymer. ２

５


(1) Impact of refusal of purchase on competition• ASML occupies a high market share of the

downstream market and there are few competitors in the downstream market.

• Therefore, if ASML virtually procure light sources exclusively from Cymer, and thus the competitors in the upstream market lose sale destinations and excluded from the upstream market, it is considered that such situation has a large impact on competition in the upstream and downstream markets. ２

６

２７

Possibility of refusal of purchase

２７


(2) Allegations of the parties and assessments thereofA. Allegations of the parties• As mentioned in 2(2) A above, the parties alleged that

there was no incentive for the parties to engage in the customer foreclosure because if the parties engaged in it, there would be competitive pressure from the chipmakers due to the fact that the choice of the light source is dependent on the decision of chipmakers.

B. Review and assessment of the allegations of the parties• As mentioned in 2(2)B above, chipmakers have

countervailing power to a certain degree against the customer foreclosure by the parties. ２

８


(3) Measures proposed by ASML US• After the JFTC explained to ASML US that such

customer foreclosure might be a possible issue in the review of the Acquisition, ASML US has proposed that it would take the following measures against the concern of the customer foreclosure.

２９

★


(3) Measures proposed by ASML US– (i) When ASML develops in partnership with Cymer

or Company A and places orders for products, parts and services of light sources to them, ASML will determine the supplier based on objective and non-discriminatory criteria, such as quality, logistics, technology, cost and chipmakers’ preferences etc.

– (ii) ASML will continuously permit chipmakers to choose light sources of their choice, and not unduly exert influence on the decision of chipmakers with respect to the choice of light sources.


(3) Measures proposed by ASML US– (iii) ASML will substantially simultaneously provide both

Cymer and Company A with information which is necessary in research and development of light sources and order placements for light source products, parts and services.

– (iv) For five years from the execution of the Acquisition, the parties will report the status of compliance with the measures mentioned above to the JFTC once a year.

– (v) The report mentioned (iv) is to be created by an audit team independent from parties, which will be appointed subject to a prior approval of the JFTC.


(4) Assessment under the AMA• The measures proposed by ASML US mentioned (3) above

represent its promise that after the Acquisition, ASML will continuously deal with Company A in a manner consistent with the terms of trade equivalent to that of prior to the Acquisition. Moreover, an audit team independent of the parties’, which will be appointed subject to a prior approval of the JFTC, conducts an audit and ASML will report to the JFTC regarding the result of audit for a certain period of time after the Acquisition, thus the effectiveness of the measures will be ensured. Moreover, as mentioned in (2) B above, there is competitive pressure to a certain degree from chipmakers.

• Therefore, taking the measures proposed by ASML US etc. into consideration, the Acquisition will not cause the customer foreclosure. ３

０

III. Review concerning substantial restraint of competition4. Access to confidential information

(1) Impact of access to confidential information on competition

• Light source manufacturers and lithography systems manufacturers share various confidential information, such as product development, product specification, their customers, etc. with each other in terms of developing, manufacturing, and selling products.

• Thus, after the Acquisition, there is a possibility that, Cymer accesses to Company A’s confidential information shared between ASML and Company A through ASML, or ASML accesses to Company X or Company Y’s confidential information shared between Cymer and Company X or Company Y through Cymer. ３

１



• It is recognized that there is less possibility the parties and competitors take coordinated conduct because technological innovation is frequent in upstream and downstream markets and there is competitive pressure to a certain degree from chipmakers.

• However, there is a possibility that the parties may use the confidential information for their advantages, and thereby their competitors may be placed in a disadvantageous situation and foreclosure or exclusivity in market may be occurred. ３

２



• The parties occupy high market shares of the both upstream and downstream markets and there are few competitors in these markets respectively.

• Therefore, if the confidential information of competitors is shared between the parties and market foreclosure or exclusivity are resulted in, it is considered that such situation has a large impact on competition in the upstream and downstream markets.

３３

３４

Possibility of access to confidential information

３４


(2) Measures proposed by ASML US• After the JFTC explained to ASML US that handling confidential information of

competitors might be a possible issue in the review of the Acquisition, ASML US has proposed that it would take the following measures against the handling of confidential information.– (i)Directors/Employees of Cymer who are responsible for the confidential information

of Company X or Company Y will be prohibited from providing the confidential information to directors/employees of ASML and enter into a non-disclosure agreement.

– (ii)Directors/Employees of ASML who are responsible for the confidential information of Company A will be prohibited from providing the confidential information to directors/employees of Cymer and enter into a non-disclosure agreement.

– (iii)To comply with (i)and (ii)above, the parties will create a protocol of information blackout for its employees.

– (iv)For five years from the execution of the Acquisition, the parties will report the status of compliance with the measures mentioned above to the JFTC once a year.

– (v) The report mentioned (iv) is to be created by an audit team independent from parties, which will be appointed subject to a prior approval of the JFTC. ３

５


• (3) Assessment under the AMA• The measures proposed by ASML US as mentioned in (2) above

represent its promise that after the Acquisition, the parties implement measures to prevent disclosure of confidential information which includes their directors/employees to enter into a non-disclosure agreement.

• Moreover, an audit team independent of the parties’, which will be appointed subject to a prior approval of the JFTC, conducts an audit and ASML will report to the JFTC regarding the result of audit for a certain period of time after the Acquisition, thus the effectiveness of the measures will be ensured.

• Therefore, taking the measures proposed by ASML US, etc. into consideration, the Acquisition will not raise an issue of access to confidential information of competitors. ３

６

IV Conclusion

• The JFTC concluded that, taking the measures proposed by ASML US, etc. into consideration, the Acquisition would not substantially restrain competition in any particular fields of trade.

３７

Contribution of Economics

• Theory Meets Practice :• The foreclosure doctrine states that the main

purpose of an acquisition of an upstream or downstream firm is to weaken the extent of competition in either market by foreclosing competitors from that part of the market taken by the acquired firm.– A theory of extending market power from one stage

to another must be explicit about what vertical integration allows the firm to do that control over the upstream price does not. (Salinger, 1988 QJE)

３８

Contribution of Economics

• Theory Meets Practice :• New measures to be theorized

– “There is a possibility that the parties may use the confidential information for their advantages.”

– The authority considers remedial measures with case-by-case approaches.

– There is little knowledge of this type of measure like a firewall shared in the authority from the theoretical point of view.

３９

Behavioral Remedies

• The USDOJ’s Policy Guide to Merger Remedies deleted the structural remedies principles:– “A. Structural Remedies Are Preferred” in III. A. of

the Policy Guide of October 2004 is deleted in the Policy Guide of June 2011.

– Several behavioral remedies are taken in recent cases such as US v. GrafTech/Seadrift, and US v. George’s Foods.

４０

Behavioral Remedies

• We have developed several useful merger analyzing tools such as UPP, GUPPI, IPR and First-Order Approach.

• Recent merger control establish two pillars: applying these tools (UPP in 2010 US Merger Guidelines) and taking thorough measures including the behavioral remedies.

４１

Need to be theorized

• The thorough measures including behavioral measures are needed to be theorized from the economics viewpoints.

• The practitioners are seeking to appropriate measures, and the wide range of theoretical analysis and suggestion is required in the cutting-edge competition policy.

４２

Theory meets practice in merger control

Ex-post Evaluation of the effects of mergers


– Empirical studies of the effects of mergers and the implications for competition policy

– Research from widely viewpoints– CR04-11: CPRC Report (The Competition Policy

Research Center)• Hiroyuki Odagiri, Koki Arai, Noriyuki Doi, Yasushi Kudo,

Takuji Saito, Kuninobu Takeda, and Chiharu Yanagita

４３

独占禁止法第 15 条Anti-Monopoly Law in Japan, Section 15

• 会社は、次の各号のいずれかに該当する場合には、合併をしてはならない。– 当該合併によって一定の取引分野における競争を実質的に制限するこ

ととなる場合。

• No company shall effect a merger if any of the following items applies:(i)When the effect of the merger may be

substantially to restrain competition in a particular field of trade.

• This regulation applies to all types of business combination.

４３

★

The Industrial Structure Vision 2010 by Industrial Ministry in Japan (METI ）

• Competition policy: ensured transparency of procedures, conversion to corporate merger review that considers the medium- to long-term and the global market.

Business integration of Nippon Steel and Sumitomo Metal (Press release, Feb. 3, 2011) ）

• Objectives– The companies would, through the business integration,

accelerate their global strategies and realize a level of competitiveness which is globally outstanding in all aspects, including technology, quality, and cost, by combining their respective resources that each has built up, and generate synergies through consolidation of the superiority area in their respective businesses.

– Through the business integration, the companies aim to better respond to the needs of customers both in Japan and overseas and desire to contribute to further development of the Japanese and global economy and improvement of global society.

４４

★

Business integration of Nippon Steel and Sumitomo Metal （ continued ）

• Goals of the integrated company• 5. Maximizing corporate value and the improving evaluation

from shareholders and capital markets– By implementing the foregoing measures, the integrated

company would seek to maximize its corporate value and would use its utmost efforts to obtain a high evaluation from shareholders and capital markets by further improving its profitability, promoting the strategic utilization of funds and assets, and building up a stronger financial base.

Analyzing Effects of Mergers with an Oligopoly Model

• The oligopoly model suggests that a merger raises the market price, reduces the output and profits of merging firms (but may increase the profit rate), and increases the profits of non-merging firms. The social welfare is hurt.– The results may change when the merger

contributes to efficiency increase and thereby reduces the merging firm’s marginal cost.

４５

Merger Guidelines (JFTC, March 2007)

• “When improvements in efficiency, whether through economies of scale, integration of production facilities, specialization of factories, reduction in transportation costs or efficiency in research and development, is deemed likely to make the company group take competitive action after the business combination, this factor will also be considered to determine the impact of the business combination on competition.”

４６

Considerations of Efficiency in the Guideline

• “Efficiencies to be considered in this case are determined from three aspects: – (i) efficiencies should be improved as effects specific to the

business combination; – (ii) improvements in efficiencies should be materialized;

and – (iii) improvements in efficiency contribute to the interests

of users.”

• => Condition (iii) requires that the interests of users (i.e., consumer welfare) to increase. Therefore, the efficiency increase must be very large.

４７

Needs for Empirical Studies

• Have mergers in the past really achieved such aims?– “Improve its profitability”: Did profit rates rise?– “High evaluation from shareholders”: Did stock

prices rise?– “Competitiveness in technology, quality, and

cost”: Did R&D expenditures increase? Did inventions increase?

– “Better respond to the needs of customers”: Did prices not rise?

４８

CPRC Joint Research Project on Post-Merger Evaluation

• We evaluate post-merger profitability, stock prices, R&D, and selling prices, using the data from mergers that occurred after 2000.– For a similar study on mergers that

occurred during the 1990s, see an earlier CPRC research report (CR02-03, 2003).

４９

Many studies for ex-post evaluation recently

• Retrospective Analysis of Agency Determinations in Merger Transactions Symposium, June 28-29, 2013, ABA

• International Journal of the Economics of Business – Special Issue: Hospital Mergers and Antitrust

Policy, Volume 18, Issue 1, 2011 – Special Issue: Issues in Empirical Merger

Analysis, Volume 13, Issue 2, 2006,

５０

★

Limitations of the Study• Due to limitations of the study, the results remain

tentative.– Dependence on published reports (except the price

study): financial reports and share price data of non-public firms are unavailable, which unfortunately are actually many.

– Short post-merger period: However, business experts tend to say that mergers need to have effects within 3 years.

– Diversity of merger motives and types: Quite a few seem to have been made to rescue failing companies.

Mergers and Profitability

• The purpose of the project is to estimate the impact of mergers on corporate performance. To do so, we need to distinguish two types of the impact: merger itself and merged firms’ characteristics.

５１

Company character-istics

Company perform-ance

Merger

We need to analyze this effect only.

５２

Selecting Control Firms against Merged Firms

• A simple comparison of pre- and post-merger changes in financial indicators (e.g. ROE) between merged firms and non-merged firms may not capture merger effects alone. In this study, we apply a propensity score matching method to select control firms, and then compare financial indicators of merged firms to those of control firms.

５３

Selecting Control Firms witha Propensity Score Matching Method

• We begin by estimating firms’ propensity scores for undertaking mergers, using firm characteristics as the explanatory variables. Then, based on the estimated propensity scores, we match a non-merged control firm to each merged firm. Finally, we compare changes in performance between them (average treatment effects).

５４

The Industry to be Analyzed in this Study

• We analyze mergers among first regional banks and second regional banks.– Years: 2000 - 2006 (Fiscal Year).– Number of merger cases: 6.

５５

★

Estimation Result of the Propensity ScoreResult of Logit Estimation (Sample Size : 773)

Explanatory Variables Estimated Coefficients

Significance

Constant -407.693 ***ROE 0.188Square of ROE -0.020 ***Operating Cost Ratio 4.769 *Capital Asset Ratio -1.899 ***Bad loans Ratio 0.535 ***Log of Assets 52.613 **Square of Log of Assets -1.723 *Growth Rate of Regional Production

0.541 ***

The explanatory variables except constant are lagged ones.Standard errors are adjusted for banks, and ***, ** and * mean statistically significant at 1% , 5%, and 10%, respectively.

Merged Banks and selected Control Banks

Merged Banks (Month / Year of Merger) Control Banks

The Kinki Osaka Bank (April / 2000) The Towa Bank

The Shinwa Bank (April / 2003) The Chukyo Bank

Tsukuba Bank (April / 2003) Suruga Bank

Momiji Bank (May / 2004) The Yachiyo Bank

The Nishi-Nippon City Bank (October / 2004) The Yachiyo Bank

The Kiyo Bank (October / 2006) The Hokuto Bank

Measurement of Corporate Performance

• We employ ROE (Current Profit / Net Worth) as a measure of corporate performance, and look at its changes from one year before merger as defined below. For merged banks before the merger, we create hypothetical merged banks by summing up the two firms’ data, to compute ROEi,base. To control the effects of business fluctuation, we subtract the annual average from ROE to make year-adjusted ROE.

.mergerbeforeyearoneatifirmofROEbaseiROE,ttimeatifirmofROEtiROE

baseititi ROEROEROE

:,:,

,,,

５６

Year-adjusted ROE in a Year before Mergers

Merged Banks and Control BanksMerged BanksROE Year-adjusted ROE

Control BanksROE Year-adjusted ROE

The Shinwa Bank 9.696% The Chukyo Bank 11.928%

The Kinki Osaka Bank 34.664% The Towa Bank 25.079%

Tsukuba Bank 4.784% Suruga Bank 16.084%

Momiji Bank 11.682% The Yachiyo Bank 2.017%

The Nishi-Nippon City Bank 3.334% The Yachiyo Bank 2.017%

The Kiyo Bank -14.035% The Hokuto Bank 3.282%

５７

Comparison of Corporate Performance (ROE)

The Shinwa Bank and The Chukyo Bank

The Kinki Osaka Bank and The Towa Bank

５８

★


Tsukuba Bank and Suruga Bank

Momiji Bank and The Yachiyo Bank


Nishi-Nippon City Bank and The Yachiyo Bank

The Kiyo Bank and Hokuto Bank


• Number of merger cases in which ROE decreased relatively to non-merging banks– 1st year after the merger: 3 (out of 4)– 2nd year after the merger: 4 (out of 6)– 3rd year after the merger: 3 (out of 6)– Note that statistical significance is lacking in

the changes in performance (both means and medians).

５９

Conclusion

• We used the sample of regional banks in Japan, applied the propensity score method to select matching firms, and estimated the impact of mergers on corporate performance. Our results suggest that mergers rarely contribute to increases in profitability.

６０

Mergers and Stock Prices (An Event Study)

Methodology

・ Applying an event study method, we analyze the extent that the announcement of a merger influences the stock price.・ The effect of merger is calculated as the

excess of actual post-announcement rate of return over the expected rate of return calculated with a market model.

６１

Model

Calculate estimated parameters ,

　　　　　　

Stock rate of return of firm i, period t

Market rate of return, period t

　　　　　　　　

.,,, titmiiti uRbaR

tiR ,

Rm,t

ia ib

６２

(2) Calculate abnormal rate of returns

(3) Calculate cumulative abnormal rate of returns,

　

.ˆˆ ,,, tmititi RbaRAR tiAR ,

2

1

., ,21

T

Tttii ARTTCAR

CARi

６３

Cases of Business Combination Studied : 15

・ Mergers (6)・ Establishment of holding company (8)・ Establishment of a consolidated subsidiary (1)

６４

An example:Establishment of a joint holding company by Sankyo Co., Ltd. and Daiichi Pharmaceutical Co., Ltd.(2005)

・ Event day ： February 21, 2005

６５

　 Case: Mergers of Sankyo and Daiichi Pharmaceutical

-0.100

-0.050

0.000

0.050

0.100

0.150

Dayly transition of the CAR （ Event day ; February 21,2005 ）

SANKYO & DAIICHI PHARMACEUTICAL

三共㈱：SANKYO CO.,LTD.

第一製薬㈱：DAIICHI PHARMACEUTI-CAL CO.,LTD.

エーザイ： EI-SAI CO.,LTD.

中外製薬：CHUGAI PHAR-MACEUTICAL CO.,LTD.

ＣＡ

Ｒ

６６

★

• The CAR of merger companies (Sankyo & Daiichi) was

positive immediately after the merger announcement but

eventually became negative (though statistically

insignificant), while that of the rival companies remained

positive.

–Suggests that the stock market expected the long-run

profitability of the merged firm would rather decrease, while

that of the rival firms would increase.

Results of the 15 Case Studies: Summary

イベント日event day

一日目one day

二日目two day

三日目three day

四日目four day

五日目five day

六日目six day

七日目seven day

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

上昇rise11 10

9

76 6 6 6

下落fall2

12

12 2 2 2

効果なしEffect none

2 4 4

7 7 7 7 7

Transition of the CAR

案件数 number of matters

６７

• In 11 cases out of 15 (i.e., 73%), CAR was positive immediately after the merger announcement; however, after a week, it was positive only in less than a half of the cases, that is, 6 cases or 40% of the cases. In the majority of the cases, CAR was not significantly different from zero.

６８

Conclusion• In the capital market, the investors will make an

investment decision based on the predicted present value of the firm’s long-run profits. Hence, if the merger is expected to raise efficiency of the firm and thereby its long-run profitability, the merger announcement should raise its stock price and result in a positive CAR.– According to our study of 15 cases, such a positive

effect was observed immediately after the announcement in many of the cases but, after a week, the effects were obscure in the majority of the cases. ６

９

Mergers and R&D

Purpose

• To analyze the impact of mergers on innovation– To examine the post-merger R&D activities of

Japanese manufacturing firms, after 2000 :

• 　 Indices to be studied are ：– R&D expenditures (as an intensity, i.e., a ratio to sales) – The number of published patent applications

７０

Review of Theoretical & Empirical Studies

• Theoretical predictions– Quiet life hypothesis vs. Schumpeterian hypothesis

• In controversy– “Innovation aversion” (Farrell & Shapiro[2010])

• New products may compete against the merger partner’s products, lessening incentives for innovation.

– The effect of a merger depends on two relations; 　 • Market & technology relatedness within the

merging firms• Relations between merging and non-merging

firms７１

Empirical Analyses: Diversity of Evidences• More evidences suggest the negative

effects–Colombo & Garrone [2006] suggest

the following:Mergers, particularly horizontal

mergers, tend to reduce R&DTend to reduce R&D when the merger

partners’ technologies are substitutable

７２

• Conclusion from the survey of theoretical and empirical analyses

• Today, competition authorities tend to give more weight to the investigation of the impact of mergers on innovation in their enforcements. Unfortunately, economic studies suggest that this impact is neither simple nor universal.

７３

Empirical Analysis

Methodology

• 39 mergers after 2000• Comparison of R&D activities before and after

mergers• R&D measures

– R&D intensity (R&D expenses/sales) (consolidated)

– Number of published patent applications (published after 1.5 years from application) (non-consolidated)

７４

Results: 39 Cases after Year 2000

　　　　　　　　　　　　　　　 B3 ・ A3 　　 B3 ・ A5 　　　 B3 ・ A3-5 　　

Δ ＞ 0 　　 Δ ＜ 0 　 All 　　　 Δ ＞ 0 　 Δ ＜ 0 　　 All 　　　 Δ＞ 0 　　 Δ ＜ 0 All

R&D intensity 17 　　 22 　　 39 　 17 　　 17 　　 34 16 　 19 　 34 　

Published application 11 　　 28 　　 39 　　 9 　　 25 　　 34 　　

11 　　 23 　 34

Δ ： post-merger value less pre-merger value (Δ ＞ 0 　 or 　＜ 0 )

　　 B3 ： 3-years mean, pre-merger; A3(5) ： 3 (5)-years mean, post-merger; A3-5 ： 3-years mean during the 3rd to 5th year after merger

７５

Results concerning R&D Intensity

• R&D intensity increased in less than half of the cases

• In R&D-intensive industries, it increased in the majority of cases (9 out of 15 among firms with R&D intensity ≥ 3%, 7/10 with R&D intensity ≥ 4%)

• It increased in 5 of the 6 cases that were subject to JFTC’s review

７６

★

Results concerning published patent applications• Decreased in the majority of cases, similarly to

R&D intensity• Even in R&D-intensive industries, it increased in less

than half of the cases (5 out of 14 among firms with R&D intensity ≥ 3%, 4/10 with R&D intensity ≥ 4%)

• Also, it increased in just 2 of the 5 cases that were subject to JFTC’s review

• However, the effect of time lag between R&D and patent applications need be examined.

Result: Published Applications

　　　Δ ＞ 0 Δ ＜ 0

All 　

2-year lag (BA3 and A3-5 compared) 12 22 34 　3-year lag (A3 and A4-5 compared) 12 22 34　　 BA3: 3 years-mean during pre-merger 1st year to post-merger 2nd year

A3: 3-years mean during post-merger 1st to 3rd year A3(4)-5 ： 3 (2)-years mean during post-merger 3rd (4th) to 5th year

Conclusion

• No evidence that mergers always promote R&D.

• In R&D-intensive industries, there were more cases in which mergers promoted R&D: however, they rarely resulted in more patents.– However, the study is constrained by imperfect measures

of innovation activities and the lack of detailed information. An examination of such detailed information will be needed in actual merger investigations.

７７

Mergers and Prices

The Purpose• In this study, we examine the price

effect of mergers by analyzing how retail prices of goods in the market changed before and after the mergers.

７８

Methodology and Model

• Using retailer scanner data and following the methodology of Ashenfelter and Hosken (2008), we compare pre- and post-merger monthly changes in retail prices of the products of merged firms (or firms in a combined group) with those of others – We will also apply the method for the estimation

of the effects on sales and market shares.

７９

– i : product, j : region, t :month– pi,j,t : sales-weighted price index

– αi,j : product- and region–specific factor

– posti,j,t : = 0 before the merger; = 1 after the merger

– MPPi : Merging Party Product

(= 1 if merging firms’ product; = 0 otherwise)– εi,j,t : error term

pi, j ,t i, j 1 post i, j,t 2 post i, j ,t MPPi i, j ,t

８０

Cases for the analysis• Household flavor seasonings (3000

items)• Sugar (850 items)

• Instant noodles (2200 items)

８１

Data

• The data were compiled from the POS (point-of-sales) data of more than 600 stores in 8 regions and consist of monthly sales and quantity.– 8 regions: Hokkaido, Tohoku, North

Kanto, Tokyo Metropolitan, Hokuriku, Tokai, Kinki, Chugoku, Shikoku, and Kyusyu

８２

★

Estimated Coefficients

β1 : Post β2 : Post×MPP

Flavor Seasoning 0.0016*** 0.0039***

Sugar 0.0528*** 0.0384***

Instant noodles -0.0023*** 0.0013***

Results

• β1 　 (the coefficient of “post”) is positive → an increase in the average market price after the merger (red column)

• β2 (the coefficient of post×MPP)

is positive → an increase in the prices of the merger firms’ products (blue column)

Flavor Seasoning Sugar Instant noodles

-2

0

2

4

6

8

10

retail price after M&A

β2 : Post×MPP β1 : Post

８３

In depth study: Flavor seasoning (Ajinomoto & Yamaki)

• The cross term with Ajinomoto was positive and significant while that with Yamaki was negative and significant– Relative to the market, Ajinomoto raised its

prices while Yamaki lowered its prices after the acquisition.

– Possibility that merging parties changed their product positioning and differentiated their products in order not only to avoid cannibalization but also to gain new customers. ８

４

Ajinomoto vs. Market Yamaki vs. Market

0.1

.2M

onth

ly s

ales

sha

re

.91

1.1

Lasp

eyre

s in

dex

2007m3month

0.1

.2M

onth

ly s

ales

sha

re

.91

1.1

Lasp

eyre

s in

dex

2007m3month

Black line: maket price level

Gray line: Ajinomoto’s price level

Black line: market price level

Gray line: Yamaki’s price level

８５

Estimation Results on Sales and Market Shares

• After the merger, sales of merger firms and their market shares decreased (flavored seasoning and sugar) while sales of the average firm in the industry increased (flavored seasoning) or decreased (sugar).

• Also a case in which the merger firms increased their sales (instant noodles).

８６

Estimated effects on sales and market shares

Market Sales β1 : Post β2 : Post×MPP

Flavor Seasoning 1.18*** -3.29***

Sugar -0.74*** -21.79***

Instant noodles -23534*** 68902***

Market Shares β1 : Post β2 : Post×MPP

Flavor Seasoning 0.00000*** -0.00002***

Sugar 0.00003*** -0.00021***

Instant noodles -0.00000*** 0.00000***

８７

Conclusion• After the merger, in all the three markets

studied, – The average market price increased (except

instant noodles).– The price of the merger firms’ product

(weighted average of the prices of the firms’ entire products) increased more than the average market price.

８８

Conclusion (continued)• Effects on product positioning

– In the flavored seasoning market, one of the merger partners raised its price while the other lowered its price, suggesting a change in the product positioning strategy.

• => Suggested is the need to examine the effects on product composition and positioning that a merger may cause. Some of the consumers may benefit whereas the others may be hurt.

８９

Summary of Empirical Analyses

• Effects on Profitability and Stock Prices– No significant effect on profit rate was observed.

Though statistically insignificant, there were more cases of declining profit rates.

– The stock prices rose in many cases on the day of announcement but declined within a few days, making the cumulative abnormal returns not significantly different from zero in the majority of cases.

=> Thus suggested is that efficiency increase was rarely large enough to raise profitability and stock performance.

９０

• Effects on R&D expenditures and number of patents– Both R&D expenditures and the number of

patents decreased in the majority of cases.– However, there were also cases in which R&D

expenditures increased, usually in R&D intensive industries.

– The measures may not have captured the innovation activity accurately. More detailed case-specific investigations are needed.

９１

• Effects on prices– The study is confined to 3 industries.– The average market prices increased (except

instant noodles), and the average prices of the merger firms’ products increased further.

– In an industry with product differentiation, the merger firm may raise the prices of some products while lowering the prices of others, suggesting product re-positioning.

９２

Thus, in a nutshell,• Decreased consumer welfare (in the three

cases studied).• Unlikely to have improved efficiency enough to

increase the firms’ profitability and to raise consumer welfare.

• Questionable if the mergers contributed to “long-term competitiveness” as the Industrial Structure Vision hopes.

• However, differences across cases need be noted.

９３

Implications for Merger Investigations

• The priority for competition policy offices should be to watch the impact on prices and the consumer welfare.

• On efficiency increase and the increase in ‘long-term competitiveness’, the firm needs to show the detailed plans, together with the scope for implementation.

９４

Use of Economic and Quantitative Evidences

• Merger evaluations in Japan have tended to be based on qualitative evaluation based on interviews, etc.

• A wider use is needed of more economic evaluations based on quantitative analyses, including SSNIP test, UPP test, and merger simulations.

９５


• In case of merger control discussion, theory meets practice in two aspects:

Case Study: ASML/Cymer Competition Authorities enforce the law case-by-

case basis, with attention to the theory.Ex-post assessment from various perspectives

Not only enforcement but also theory is needed to be assessed after the fact.

９６

Thank you for your attention.

• Questions and Comments are welcomed.

• E-mail: [email protected]

mailto:[email protected]

Documents

Theory Meets Practice in Merger Control Koki Arai Senior Researcher for Economic Analysis Japan Fair Trade Commission 1st ATE Symposium on Competition