The COVID-19 pandemic did not slow M&A in the semiconductor industry. On the contrary, 2020 saw the total value of semiconductor M&A deals setting an all-time high at $118 billion. According to a report by IC Insights,[1] a semiconductor market research firm, the five mega-deals announced in the second half of last year had a combined value of $94 billion, accounting for about 80 percent of the total for the entire year. Among these M&A transactions, NVIDIA’s $40 billion acquisition of processor-design technology supplier ARM from Softbank is the most high-profile deal. Other spotlight deals also include AMD’s all-stock deal worth $35 billion to buy Xilinx, Intel’s $9 billion sale of its NAND flash memory business and 300mm wafer fabrication plant for producing microchips in China to SK Hynix, ADI’s acquisition of Maxim for $21 billion, and Marvell Technology’s acquisition of Inphi[2] for $ 10 billion.

The wave of industry consolidation in 2020 was driven by the integrated circuit (“IC") giants’ strategy to build their future competitiveness in emerging and high-growth market segments, such as embedded machine-learning and AI, automatic driving, data centers, and the Internet of Things. Accordingly, the M&As are mainly characterized by a strong alliance or complementary advantages between leading companies in each segment. However, those transactions are far from the final landing-the last but also the most critical and difficult hurdle will be to obtain the applicable antitrust approval(s). Undoubtedly, the front page news for 2021 will be which of those deals will get the green light?

From the perspective of potential competitive concerns that may be caused by these transactions, NVIDIA’s intended acquisition of ARM, which would be the biggest deal in the semiconductor industry if completed, seems to be most problematic.[3] According to public information, NVIDIA’s deal is subject to antitrust clearance from competition authorities of the U.K., the EU, the U.S., and China. It is commonly believed that one of the biggest challenges may come from China. Some even indicate that it is unlikely that China will approve the transaction. What has led to such speculation? Why is the antitrust review process of cases in the semiconductor industry in China is more complicated?

In recent years, U.S. companies have maintained the highest share of M&As in the semiconductor industry. Japanese and Korean companies are also revitalizing their semiconductor industry through M&As. However, for reasons known to all, Chinese semiconductor enterprises find it difficult to engage in overseas M&As and are struggling to acquire supplies and expand their business landscape. It takes very little imagination to foresee the impact on China’s semiconductor industry when these industry giants build a diversified chip industry chain and enter China’s market with competitive product portfolios. In this sense, this round of consolidation in the semiconductor industry indeed triggers competitive concerns for China’s market, with tougher merger control enforcement in the semiconductor industry looming around the corner.

This article provides an overview of how China’s competition authority has enforced merger control rules in the semiconductor sector. It also analyzes the major challenges faced by enterprises in the semiconductor industry in the merger control enforcement and proposes strategies to solve these challenges.

The semiconductor industry comes with rich valuation and is critical to many products widely used in our daily life. Countries therefore tend to give tactical significance to the semiconductor industry. Accordingly, any monopoly in this industry will be detrimental not only to consumer welfare in a general sense, but also to a country’s overall economic development. It is notable that the review standards specified by the Anti-Monopoly Law of China (the “AML") include not merely competition-related factors of a transaction, but public interests and the impact on the development of the national economy as well,[4] which explains why semiconductor mergers are constantly under strict antitrust scrutiny in China.

Since the AML came into force, China’s merger control enforcement authority, SAMR (and formerly MOFCOM), has announced conditional clearance for 48 cases up to the end of 2020, with ten involving the semiconductor industry, higher than the number of remedy cases in other industries.

（点击图片查看大图）

The above table shows that in throughout the recent ten years, remedy cases in the semiconductor industry account for a relatively high share of the total in each year. Notably, the big deals in the semiconductor industry that have been challenged due to competitive concerns identified in the merger review process are in fact not limited to those listed in the above table. In 2018, the famous Qualcomm/NXP deal collapsed for failing to secure the approval of China’s antitrust authority. In 2016, another two high-profile merger deals, the Applied Materials/Tokyo Electron deal and the Lam Research/KLA-Tencor deal were dropped by the merging parties for the same reason.

1. The Time Frame

Under the AML, in general, the maximum statutory review period for the notification of a transaction is as long as 180 days after SAMR formally accepts the case.[5] However, in practice, it may take more than 180 days, if the notifying party is required to withdraw and re-file the notification, where SAMR will run out of the time to complete its review owing to complex remedy negotiations.

Remedy cases in the semiconductor industry even saw a longer time frame than the average. Among all the 48 remedy cases, 21 cases went through the withdraw-and-refile process, with seven of them in the semiconductor industry. The average review period[6] for the conditionally approved cases is 259 days, while cases in the semiconductor industry take 303 days, indicating that cases in the semiconductor industry may prompt increased competitive concerns, thus requiring a more intricate remedy negotiation.

（点击图片查看大图）

2. Solicitation of Stakeholders’ Opinions

As explained above, SAMR also evaluates the impact of a transaction on public interests and the development of the national economy. Therefore, SAMR considers and gives weight to the feedback it receives from key stakeholders in its consultation process. Moreover, opinions solicited from stakeholders also play an important role in assessing the remedy proposal. If the feedback is positive from all stakeholders, SAMR will clear the transaction with the proposed remedies imposed.

SAMR usually solicits opinions from industry regulators, trade associations, key suppliers and customers, competitors, and sometimes industry experts.[7] It is particularly important to solicit the opinions of stakeholders in the merger case involving sensitive sectors such as the semiconductor industry when identifying competitive concerns and assessing the effectiveness of the proposed remedies.

In reviewing merger cases in the semiconductor industry, SAMR usually seeks opinions from the Ministry of Industry and Information Technology (“MIIT"), the Ministry of Science and Technology (“MST"), and the National Development and Reform Commission (“NDRC"). The most influential trade association is China Semiconductor Industry Association (“CSIA").

3. Engagement of Independent Third-Party Consulting Agencies

Based on our review of the decisions of the cases approved with conditions, the authority mentioned in ten decisions that it engaged independent third-party consulting agencies to conduct an economic analysis of the competitive concerns. Overall, it is not a very high proportion against the total 48 remedy cases. However, it is notable that four semiconductor cases engage independent third-party consulting agencies, including Media Tek/MStar, Ase Semiconductor/Siliconware Precision Industry, Ⅱ-Ⅵ/Finisar, and Nvidia/Mellanox, maintaining a high presence compared to cases in other industries.

The economic analysis conducted by third-party consulting agencies plays an important role in identifying and quantifying the harm to competition. For instance, in the Ase Semiconductor/Siliconware Precision Industry deal, a horizontal merger, the economic analysis showed that the profit margins of Ase Semiconductor and Siliconware Precision Industry to Chinese customers were relatively close, and the correlation coefficient of profit margins in China was 0.72 (1 is exactly the same), a strong correlation between the profit margins of both sides over time, indicating that they were close competitors in China’s market. Based on such observations, MOFCOM concluded that the concentration would eliminate the close competition between the merging parties, thus causing damage to China’s market even though the relevant geographical market should be defined as global. In the Ⅱ-Ⅵ/Finisar case, SAMR concluded that there was fierce price competition between the merging parties in the relevant market based on the economic analysis on the relevant bidding data, suggesting that with Ⅱ-Ⅵ to participate in the bidding, Finisar’s is more willing to cut prices. Furthermore, as with the practice of the competition authorities in other antitrust jurisdictions, in non-horizontal mergers, the economic analysis could also serve as a useful tool, for example to identify the intent of the merged firm to exercise foreclosure behaviors.

1. Competitive Concerns Identified in the Remedy Cases

Theories of Harm, which are generally divided into unilateral effects and coordinated effects, provide an analytical framework to assess whether and if so, how a merger would eliminate or restrict the effective competition in the relevant market. However, the specific competitive concerns that may be caused by a horizontal merger are usually different from those of a non-horizontal merger. The table below summarizes the competitive concerns underlying the SAMR/MOFCOM’s enforcement decisions involving the semiconductor industry.

（点击图片查看大图）

2. Competitive Concerns Identified by SAMR/MOFCOM in Horizontal Mergers

In its review of horizontal mergers, SAMR/MOFCOM basically adopted an approach similar to the theories of harm to assess both unilateral effects and coordinated effects of a merger as used by its counterparts in other antitrust jurisdictions.

With regard to unilateral effects, the competitive concerns that are frequently expressed are elimination of close competition between merging parties; reduction of alternative choices for customers resulting in a higher procurement risk; and loss of innovation. The market structure has always been an important factor in assessing whether the merged firm has the ability to engage in anti-competitive practices. Another key indicator to evaluate the unilateral effect is the extent of close competition between merging parties. As noted above, SAMR/MOFCO may use various economic tools to test if merging parties closely compete with each other.

Coordinated effects are not commonly discussed in the enforcement decisions. Among seven cases involving horizontal overlaps, only the recent case of Ⅱ-Ⅵ/Finisar, and two much more earlier cases of Western Digital/Hitachi Storage and Seagate /Samsung’s HDD Business mentioned that the merger might lead to collusive practices. According to our observation, the relevant markets in those cases show clear signs of vulnerability to coordinated conducts such as oligopoly structure, a high degree of market transparency and the absence of buyer’s countervailing power.

3. Competitive Concerns Identified by SAMR/MOFCOM in Non-Horizontal Mergers

In line with the general trend of industry integration, most remedy cases in the semiconductor industry in recent years are non-horizontal mergers. In our view, the recent cases, on the one hand, reflect the latest development of the enforcement practices in assessing competition impact of non-horizontal mergers in major antitrust jurisdictions, and on the other hand, show certain special competitive concerns of China’s competition authority.

In the remedy cases involving vertical or complementary integration, competitive concerns typically identified in the unilateral effect analysis include vertical foreclosure, degradation of interoperability, and gaining access to competitively sensitive information of upstream or downstream rivals. These competitive concerns are also highlighted in the relevant vertical merger guidelines in the U.S. and the EU.

In addition to the common theories of harm used by other jurisdictions, China’s competition authority tends to assume that the merged firm would leverage its market dominance in one market to gain unfair competitive advantages in another neighboring market by abusive behaviors of tying and bundling. SAMR/MOFCOM raised this concern in all the remedy cases of a vertical or neighboring nature in the semiconductor industry. It is reasonable to infer that the concern of potential abusive behaviors may come from the feedback it receives in its consultation with key stakeholders. With a view to  preventing abusive behaviors in advance, identifying such competitive concern helps to impose concrete behavioral remedies such as “no tying and bundling" commitment. An antitrust investigation against abusing dominance conducts can be much more challenging for the competition authority as tying and bundling sale is not a per se violation, even for dominant enterprises.

As noted above, it is widely recognized that China’s competition authority appears to have a stronger preference for behavioral remedies compared with any other competition authority. Such preference manifests itself in remedy cases in the semiconductor industry. Among the ten semiconductor related remedy cases, nine cases, including both horizontal and non-horizontal mergers, involve behavioral remedies. The only one case concluded with a structural remedy of business divesture is the NXP/Freescale deal. The behavioral remedies imposed in semiconductor cases above mainly include commitments:

（点击图片查看大图）

Through those conditional clearances, conclusions can be reached regarding the general trends in merger remedy cases in the semiconductor industry in China. Enterprises in this industry considering M&A deals should get well prepared for the challenges they may encounter in the merger review process.

First, it is reasonable to foresee a tougher merger control enforcement in the semiconductor industry due to its strategic significance to the national economy. However, it does not necessarily mean that SAMR will definitely give more weight to industrial policies than to competition ones when reviewing transactions in the semiconductor industry. A more reasonable interpretation is that the competition authority tends to pay close attention to the potential competitive concerns that the proposed transaction may bring about to China’s market, even though the relevant geographical market for semiconductor-related products should normally be defined as global. Therefore, the special competition environment faced by the enterprises in China’s semiconductor industry chain, for instance, the restrictions on the U.S. export and foreign investment out of national security concerns and heavy dependence of Chinese downstream players on foreign suppliers, should be considered with prudence.

Second, SAMR is more concerned with non-horizontal deals today, which means merging parties should expect and gear up for more rigorous antitrust review of deals especially those involving significant vertical or neighboring relationships and thus it is important for the parties involved to make strategies to proactively address the competitive concerns that SAMR may hold.

Third, SAMR is actively coordinating with other competition authorities with regard to the scope of remedies when reviewing global deals. However, it is not uncommon for SAMR to impose remedies beyond the scope of the remedy packages employed by other antitrust authorities to cope with the competitive concerns specific to China’s market. Notably, the commitment of “no tying or bundling" or any other abusive conducts have become increasingly common in the high-profile mergers involving complementary product portfolios. The rationale behind such remedy measure is about preventing abusive behaviors in the aftermath of a merger, and to a certain extent, it also reflects the consistency of Chinese law enforcers in attaching importance to proactive regulation and supervision.

Last but not the least, statistics show that the time taken for SAMR to review semiconductor-related cases is relatively longer. Parties to a transaction with competitive concerns should therefore prepare themselves for uncertainty when it comes to the merger review time frame in China. The rights and obligations related to obtaining antitrust approval under the transaction agreement should be designed with care. Furthermore, it is highly recommended that parties to a transaction identify the potential competitive concerns, to build a defense strategy with the assistance of antitrust counsels and economic consulting firms, and to plan the proposal for remedies at an early stage.

* This article was first published in the CPI Antitrust Chronicle: China Issue (March 2021), available at: https://www.competitionpolicyinternational.com/category/antitrust-chronicle/antitrust-chronicle-2021/winter-2021-volume-3-number-2/.