The long list of semiconductor transactions around the globe in the past few years underlines China’s continuing role as one of the key jurisdictions for global semiconductor transactions and their clearance prospects under applicable merger control laws. On June 24, 2022, China’s top legislature formally published the amended Anti-Monopoly Law, with sweeping changes in force from August 2022. While it does not specifically target the semiconductor industry, it may impact the merger control review practices applicable to semiconductor deals, particularly considering that the semiconductor industry is one of the most heavily scrutinized sectors in China. This article assesses what global semiconductor players and investors should expect when pursuing a clearance in this new antitrust era in China.

By Hazel Yin, Haoyang Zhuo & Laurent Bougard[1]

 

I. INTRODUCTION

The numerous semiconductor transactions around the globe in the past few years have included transformative multi-billion deals with potentially far-reaching impact on the global semiconductor landscape, as well as many smaller deals (some statistics indicate that 80 percent of semiconductor deals were valued below $1bn in the last five years).[2]

Despite volatility in demand for semiconductor products and a projected modest decline in global semiconductor revenue in 2023,[3] robust demand remains in a few segments, shifting from consumer electronics to emerging areas like the internet of things (“IoT”), 5G connectivity and, most notably, the automotive industry, which has suffered from a chip shortage. This chip shortage has also spurred considerations of how to secure a stable access to these key inputs.

As industry consolidation continues, navigating regulatory challenges associated with competition authorities in different jurisdictions has become critical for successful deal-making. China is the top market for semiconductor products across different layers of the supply chain so a Chinese merger control filing is very often required. In 2021 alone, transactions reviewed by China’s antitrust regulator, the State Administration of Market Regulation (“SAMR”) for merger control in the semiconductor sector amounted to a total transaction value of RMB 736.1bn.[4]

In addition to geopolitical complexities, especially following China-U.S. trade tensions, August 2022 marked a new era for antitrust in China, with the amended Anti-Monopoly Law (the “New AML”) coming into force. The New AML brought about sweeping changes, including potentially heightened scrutiny of mergers and acquisitions concerning critical areas such as the national economy and people’s livelihood.[5] while it does not specifically target the semiconductor industry, it may impact the merger control review practices applicable to semiconductor deals, particularly considering that the semiconductor industry is one of the most heavily scrutinised sectors in China.

 

II. A GLANCE AT MERGER REVIEW STATISTICS FOR SEMICONDUCTOR DEALS IN CHINA

In 2021, SAMR completed reviews of 727 transactions, including 26 semiconductor transactions. While semiconductor deals only account for less than 4 percent of transaction volume, they contribute approximately 20 percent by value in all merger control reviews SAMR completed in 2021.

Chart 1 Completed merger review cases in 2021

On the other hand, semiconductor-related filings have a much larger share in remedy cases – of the 57 conditionally approved transactions by end of 2022, 14 cases concerned semiconductor products or services, accounting for approximately 25 percent. When it comes to the applicable procedures (i.e. normal vs simplified procedure), of the 78 semiconductor deals SAMR reviewed from 2016 to 2020, more than 50 percent of transactions were reviewed under the normal procedure,[6] which typically results in a longer review timeframe of between three and six months. In 2021, the proportion was about 42 percent.[7] By contrast, as a reference, the ratios of filings under normal procedures for the digital sectors, chemicals, and the pharmaceuticals/medical industries are approximately 25 percent, 19 percent and 23 percent in 2021, respectively.[8]

 

III. POTENTIAL TRENDS AND CHALLENGES FOLLOWING THE NEW AML

A. Timetable Management: Will the New Stop-the-clock Mechanism Help Manage Review Timelines?

Managing regulatory timetables for global transactions is always challenging. In China, once the merger review process is formally initiated (through case acceptance), SAMR has up to 180 calendar days to review and conclude a case.

Before the new AML was adopted, there was no mechanism for SAMR to suspend its review process, even for valid reasons. However, for complicated transactions with competition concerns, it was often challenging to complete reviews within the statutory time limit given the complexity of products and markets concerned, and the time required to identify the concerns and to agree on the proper remedies to address these. Also worthy of note is that SAMR has been under-staffed for many years, with only a handful of case handlers, who concluded more than 700 merger filings in 2021, marking a 54 percent increase compared with 2020.[9]

In practice, if 180 days are not sufficient to complete the China merger review process, the parties would have had to pull and refile the transaction to buy more time. For instance, in 2022, all three conditionally approved transactions in the semiconductor industry, namely Siltronic/GlobalWafers, AMD/Xilinx, and II-VI/Coherent, went through a pull-and-refile process and took over a year to receive approval (see Table 1 below).

Table 1 Review timetable for conditionally approved semiconductor transactions in China

No.

Case Name

Total review period (days)

Duration of case acceptance (days)

Any pull & refile

Extra days used after the first round of review (days)

1

MediaTek/Mstar (2013)

417

60

Yes (Once)

177

2

Merck/AZ Electronic Materials (2014)

106

14

No

/

3

NXP/Freescale (2015)

237

42

Yes (Once)

15

4

Broadcom/Brocade (2017)

222

52

No

/

5

ASE/SPIL (2017)

457

111

Yes (Once)

166

6

KLA/Orbotech (2019)

292

59

Yes (Once)

53

7

II-VI/Finisar (2019)

264

53

Yes (Once)

31

8

Infineon/Cypress (2020)

239

62

No

/

9

Nvidia/Mellanox (2020)

359

113

Yes (Once)

66

10

Cisco/Acacia (2021)

451

59

Yes (Twice)

212

11

SK hynix/Intel (2021)

370

97

Yes (Once)

93

12

Siltronic/GlobalWafers (2022)

392

94

Yes (Once)

118

13

AMD/Xilinx (2022)

368

78

Yes (Once)

110

14

II-VI/Coherent (2022)

372

88

Yes (Once)

104

In an effort to streamline the merger review process, the New AML has introduced a “stop-the-clock” mechanism, enabling SAMR to suspend its review period (with written notice) in any of the following circumstances: (1) the notifying party fails to submit requested information in a timely manner, preventing the review from progressing; (2) new facts emerge that materially impact the review and prevent the review from progressing without verifying the facts first; or (3) the notifying party applies to suspend the review to allow sufficient time for remedy discussions. It is anticipated that the “stop-the-clock” regime will replace the current practice of having to “pull and refile” a transaction in China.

In fact, it is not uncommon around the globe that both regulators and notifying parties are constantly under time pressure to resolve all concerns and complete the review process within the prescribed timeframe, especially when for a high-profile, transformational deal in strategic or sensitive segments. Competition authorities in other jurisdictions have been utilising the “stop-the-clock” mechanism to provide more flexibility and motivation for the parties to properly manage the regulatory timetable. For instance, under the EU merger control regime, the initially set deadline for the European Commission (“EC”) to decide on a deal can be extended if the parties fail to provide complete information requested by the EC,[10] and the EC has become more willing to suspend their review clock during the past decade, usually with informal communications and negotiations with the parties for a mutually acceptable re-start date. This practice is limited to in-depth (“Phase II”) investigations of transactions.

Failure to provide information required by SAMR in time will probably be the most likely triggering event for SAMR to proactively suspend the review clock. This is also consistent with the stop-the-clock mechanism of the EU, which ties the suspension of review clock with scenarios where the parties have not responded to a formal information request or a part of an information request.

Semiconductor deals can often give rise to lengthy fact-finding processes, with multiple rounds of comprehensive enquiries for notifying parties before SAMR can conclude on a proper scope of market definition (especially in niche markets). These are the starting point for SAMR to establish any theory of harm. A wide scope of market testing and engagement with relevant third parties can also expand the scope of information required, in order to bottom out any potential concerns on defined markets.

The second triggering event for SAMR to proactively stop the clock relates to emergence of new facts that prevent SAMR from proceeding with the review. The amended AML does not provide guidance on what types of changes will qualify as “new facts” that will materially impact the review process. It also remains to be seen how such a suspension mechanism is distinguished from a withdrawal of notifications triggered by significant changes, which is an existing mechanism whereby notifying parties can withdraw their notification in the case of major changes to a transaction or the market competition landscape that warrants a new filing. Without further clarification on the nature of such “new facts,” the decision seems largely at SAMR’s discretion. The notifying parties in a semiconductor transaction may have to constantly monitor the competition and industry related changes that may impact the China market, to mitigate uncertainty or address potential issues as early as possible.

Despite these uncertainties, it is also expected that SAMR will apply the stop-the-clock mechanism very cautiously as part of the efforts to modernise its merger review process, on top of the efforts already made over the years to accelerate the review process. Above all, the vast majority of the transactions cleared by SAMR have benefited from the simplified procedure, where SAMR will be unlikely resort to suspension of the review clock.

SAMR has also been trying to provide extra comfort and confidence to notifying parties in the application of the new mechanism. For instance, pursuant to the Draft Provisions on the Examination of Concentration of Undertakings (Draft Implementing Rules) published by SAMR alongside the new AML, failure to provide required information would not automatically lead to suspension of the review clock, and the parties can request an additional extension before SAMR eventually decides to stop the clock.[11]

Existing EU practices may shed light on how such a mechanism could work. Similarly, the EU has no ultimate limit on for how long and how many times the review clock can be stopped. But in practice, the EC is usually open to negotiate the submission deadline for pending information requests and the duration of the suspension of review clock. Although such communications are informal and do not legally restrict the EC from extending the suspension, these informal contacts and the agreement on a mutually acceptable re-start date help set forth clearer expectations for next steps and work plans, providing clarity to notifying parties. Such suspensions are observed in the context of requests for a significant volume of internal documents during in-depth investigations – the EC imposes a deadline which cannot be realistically achieved, in turn triggering the informally discussed (if not agreed) suspension of the timeline.

B. Ex Officio Power: Would SAMR Call in a Semiconductor Transaction Below the Turnover Thresholds?

Another highlight of the New AML is the codification of SAMR’s power to “call in” a transaction that falls below the filing thresholds and to require the parties to notify if the transaction is likely to eliminate or restrict competition. The call-in power is not new in China: prior to the new AML, the implementing rules of the AML already make it clear that SAMR could call in a transaction for review if evidence indicated that it was likely to be anticompetitive. That said, SAMR has rarely, if ever, exercised its call-in power in the past decade. In fact, there are no procedural rules to follow, even if SAMR intends to call in a transaction. 

More proactive enforcement may now become a reality with the Draft Implementing Rules providing for such procedural rules. Pursuant to the Draft Implementing Rules, SAMR has the power to call in a transaction either before or after a transaction has closed (without any time limit).[12] It is also expected that SAMR will explore facts first, during which time the parties may have the opportunity to explain and argue why no Chinese filing is triggered or necessary before SAMR formally determines to call in the transaction and initiate the merger review. Being formally called in can lead to immediate implications on the timeframe of a pending transaction – as Chinese merger control review is suspensory the parties cannot close the deal before receiving SAMR clearance.[13] In the case of a closed transaction, SAMR also has the power to suspend implementation of the transaction until a clearance is granted.[14]

SAMR’s call-in power is believed to be aimed at tackling so-called “killer acquisitions,” which is a trend that had already been picked up by competition authorities in many jurisdictions. Killer acquisitions refer to transactions where, to avoid future competition, a large incumbent acquires a start-up in order to shut down the latter’s products, pipelines or projects before it can grow into a viable competitive threat.[15] Killer acquisitions have become a particular concern in the digital and life science sectors, where the target company may already have substantial market value and growth potential and yet insufficient turnover to trigger a filing requirement. Although SAMR’s call-in power can capture killer acquisitions, it enables the authority to review a broader range of potentially problematic transactions such as where the target company is already a well-established player globally and yet has not had significant China presence.

Such flexibility seems particularly relevant to the semiconductor industry – the geographic market of semiconductor products is very often global. This is because modern microchips have billions of components and are ubiquitous, and these hugely complicated products have resulted in an equally complex supply chain composing of thousands of specialised companies across the world.[16] Even if a party or parties do not currently have sufficient turnover or a strong market position in China, the global nature of the market might draw concerns as to whether the market power of the parties outside of China may be easily translated into a similar position in China down the line. Similarly, a semiconductor deal that focuses on niche products, while not a stereotype killer-type acquisition, may still be captured by the call-in power. With the proposed increase of Chinese merger control filing thresholds, where China turnover threshold will double from RMB 400mn to RMB 800mn;[17] more transactions will likely fall below the new thresholds, and may open the door for potential call-in.

This may be particularly relevant to high-profile transactions undergoing close scrutiny in other jurisdictions, which heightens the risk of Chinese stakeholders to spot the transaction and voice their concerns to the Chinese regulators. Vocal third-party complainants knowing the in’s and outs of Chinese merger control may be a particular threat to overseas semiconductor transactions. Their complaints can be driven by genuine competition concerns, or industry-related concerns, but also commercial considerations. These can all be raised with SAMR, who would then verify the evidence and take a view on whether a call-in is warranted. 

This call-in risk serves as a reminder on the importance of managing regulatory uncertainties and identifying China-specific (competition or non-competition) issues upfront – regardless of notifiability – especially in light of the suspensory effect once a transaction is called in.

C. Substantive negotiation: will there be additional complexity on remedy design?

The new AML also imposed additional uncertainty when it comes to discussions on remedies. As mentioned above, the review clock can be stopped if the notifying party applies to suspend the review to allow sufficient time for remedy discussions. Ultimately the notifying parties must bear the timing risk, as SAMR can eventually resort to blocking the deal if remedy proposals are found ineffective or insufficient.

On the other hand, designing effective remedies for a global semiconductor deal looks to become increasingly challenging. Globally, there is a trend of renewed focus on non-horizontal mergers by competition authorities, such as in the U.S. and EU.[18] In China, the authority’s attention to such concerns can be traced back to the early era of AML enforcement.[19] Since SAMR’s establishment in 2018, approximately 60 percent of SAMR remedy cases involved vertical or conglomerate concerns (sometimes together with horizontal concerns), and SAMR has been largely resorting to behavioural remedies to resolve potential vertical foreclosure or conglomerate concerns in China. Behavioural remedies can also be employed to solve China-specific concerns in addition to cross-jurisdiction divestitures.[20]

Remedies adopted in recent decisions typically include a combination of: (i) commitment of continued supply to Chinese customers on FRAND terms and maintaining the quality of product/aftersales support; (ii) commitments to ensure interoperability with third-party products; (iii) commitment to not engage in tying/bundling practices; and (iv) commitment to continue R&D investments (in China). While one may argue the effectiveness of behavioural remedies in general, during the past few years, these remedies have been frequently used to mitigate a range of competition concerns and industry issues that are commonly seen in the semiconductor sector.

However, whether such behavioural commitments will continue to work has come more in doubt against the backdrop of tightened export control restrictions, especially the U.S.’ upgraded chip bans on China. The upgraded U.S. export control rules released in 2022 cover a wider range of restrictions that limit China’s access to advanced computing chips, its ability to develop and maintain supercomputers, and to manufacture advanced semiconductors. It not only covers the chips as well as relevant technology and software, but also includes manufacturing equipment.

It remains unclear whether transacting parties can simultaneously satisfy their supply commitments to SAMR and the U.S. export control requirements, which may be in tension, or even contradiction with each other. Going forward, companies in a similar situation may also have to show that they have the willingness and, most importantly, ability to navigate conflicting regimes to receive the go ahead for their transaction.

 

IV. CONCLUSION

As the world’s largest consumer of semiconductor products, China will inevitably play a key role in the global regulatory processes, and merger control processes in particular. While there is at this time no clear empirical evidence that SAMR’s attitudes towards semiconductor or other transactions are changing, uncertainty both in terms of procedure and substance can be expected to increase in light of the New AML.

Semiconductor players or investors should ensure they understand how to manage uncertain review timelines in China, which are likely to manifest in case of complaints or substantive issues. Concerns unrelated to competition also need to be carefully mapped out upfront, particularly those involving any critical inputs, sectors or stakeholders in China, and/or threatening supply to China market in the context of U.S. export controls. These can lead to additional layers of complexity when determining whether a transaction would be notifiable in China, or be deemed problematic by SAMR, in which case the question shifts to whether a commercially acceptable remedy may be found to address China-specific concerns, on top of a globally coordinated remedy package.

Click here for the full article.

[1] Hazel Yin is an antitrust partner at RuiMin Law Firm, China. Haoyang Zhuo is an associate with the Antitrust, Competition & Trade group at Freshfields Beijing office.  Laurent Bougard is a senior associate with the Antitrust, Competition & Trade group at Freshfields Hong Kong office. The authors would like to thank Ethan Qu and Xiaopeng Lai at Freshfields Beijing office for their research assistance. The views expressed herein are those of the authors only and do not represent those of the entities that the authors are affiliated with.

[2] See Accenture’s report at https://www.accenture.com/_acnmedia/PDF-176/Accenture-Semi-MA-Value-Extraction-From-Smaller-Deals.pdf.

[3] For example, Gartner projected a 3.6 percent decline in global semiconductor revenue in 2023. See Gartner’s projection available at https://www.gartner.com/en/newsroom/press-releases/2022-11-28-gartner-forecasts-worldwide-semiconductor-revenue-growth-to-decline-3-6-percent-in-2023#:~:text=Global%20semiconductor%20revenue%20is%20forecast,billion%20(see%20Table%201).&text=Currently%2C%20the%20semiconductor%20market%20is,markets%20and%20enterprise%2Ddriven%20markets.

[4] See SAMR’s Annual Report on Antitrust Enforcement in China (2021), available at https://www.samr.gov.cn/xw/zj/202206/t20220608_347582.html.

[5] See Article 37 of the New AML, which calls for strengthening merger control of transactions in critical sectors impacting national economy and people’s livelihood.

[6] See SAMR’s Annual Report on Antitrust Enforcement in China (2020), available at http://www.gov.cn/xinwen/2021-09/24/content_5639102.htm.

[7] See SAMR’s Annual Report on Antitrust Enforcement in China (2021), available at https://www.samr.gov.cn/xw/zj/202206/t20220608_347582.html.

[8] The split between normal cases and simple cases is impacted by a series of factors, and most notably the approach to market definition. Under the Chinese merger control rules, if the parties’ combined market share exceeds 15 percent for any horizontal overlap or each of the parties’ market shares exceed 25 percent in any vertical relationships, the transaction will have to go through the normal procedure. In semiconductor deals, if the parties have a strong market position in a niche market or a potential segment, the authority may be more prudent about whether a broad market definition remains appropriate and may prefer to have a deep dive into any plausible narrower segments to rule out potential anti-competitive effects. This may also explain some of the semiconductor cases where the parties were required to switch their application under the simplified procedure to the normal procedure.

[9] See SAMR’s Annual Report on Antitrust Enforcement in China (2021), available at https://www.samr.gov.cn/xw/zj/202206/t20220608_347582.html; and SAMR’s Annual Report on Antitrust Enforcement in China (2020), available at http://www.gov.cn/xinwen/2021-09/24/5639102/files/77006c5bccc04555aa05f30c9a296267.pdf.

[10] See Article 10 (Time limits for initiating proceedings and for decisions) and Article 11 (Requests for information) of Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (the EC Merger Regulation) (EUMR).

[11] Article 23 of Draft Provisions on the Examination of Concentration of Undertakings for public consultation (2022), available at https://www.samr.gov.cn/jzxts/tzgg/zqyj/202206/t20220624_348145.html.

[12] See Article 26 of the New AML.

[13] See Article 26 of the New AML and Article 7 of Draft Implementing Rules.

[14] See Article 7 of Draft Implementing Rules.

[15] See OECD report Start-ups, Killer Acquisitions and Merger Control, available at https://www.oecd.org/daf/competition/start-ups-killer-acquisitions-and-merger-control-2020.pdf.

[16] See The Economist’s briefing, The semiconductor industry and the power of globalization, available at https://www.economist.com/briefing/2018/12/01/the-semiconductor-industry-and-the-power-of-globalisation.

[17] See Article 3 of Draft Provisions of the State Council on the Notification Thresholds for Concentration of Undertakings for public consultation (2022), available at https://www.samr.gov.cn/jzxts/tzgg/zqyj/202206/t20220625_348150.html.  

[18] Both the DOJ and FTC had tried to block vertical mergers in 2022 (e.g. the DOJ unsuccessfully attempted to block UnitedHealth/Change Healthcare and FTC failed in challenging Illumina/Grail, both involving vertical transactions). Both agencies are also reviewing the Vertical Merger Guidelines over the concern that the old guidelines (2020 Vertical Merger Guidelines) overstate the potential efficiencies of vertical mergers. The EC has recently also stepped up its scrutiny of vertical mergers by prohibiting Illumina/Grail.

[19] China’s first prohibition decision against Coca-Cola/Huiyuan (2009) identified conglomerate concerns where Coca-Cola was found to have the ability to leverage its market power in the carbonated soft drinks market to enhance its presence in the neighboring juice beverages market.

[20] The conditional approval of Dow/DuPont (2017) in China is such an example. In addition to the global divesture, the parties also committed to keep supplying certain crop protection products to willing Chinese customers and not to enter exclusive arrangements with Chinese distributors.