CONFIDENTIAL — SUBJECT TO FEDERAL RULE OF EVIDENCE 408
WHITE PAPER ON COMPETITIVE EFFECTS
PROPOSED ACQUISITION OF VECTRIS PHOTONICS, INC. BY HALCYON SEMICONDUCTOR AKTIENGESELLSCHAFT
Submitted to the Antitrust Division United States Department of Justice 950 Pennsylvania Avenue NW Washington, DC 20530
Submitted by
Calderwood & Harkness LLP On Behalf of Halcyon Semiconductor Aktiengesellschaft
October 10, 2023
HSR Filing: August 18, 2023 Second Request (Civil Investigative Demand): September 15, 2023
Primary Contact: Howard Bellamy, Partner Antitrust & Competition Practice Group Calderwood & Harkness LLP 1333 New Hampshire Avenue NW, Suite 700 Washington, DC 20036 Telephone: (202) 555-6100
Email: hbellamy@calderwoodharkness.com
Engagement No. 1032-00005
This submission and the information contained herein are provided subject to Federal Rule of Evidence 408 and are intended solely for use in connection with the Antitrust Division's review of the proposed transaction described herein. This document constitutes a confidential commercial submission and is exempt from disclosure under Exemptions 4 and 7(A) of the Freedom of Information Act, 5 U.S.C. § 552(b)(4) and (b)(7)(A). No portion of this submission may be disclosed to any third party without the prior written consent of Calderwood & Harkness LLP.
Right-click to update Table of Contents
I. EXECUTIVE SUMMARY
This white paper is submitted on behalf of Halcyon Semiconductor Aktiengesellschaft ("Halcyon" or "HSAG"), a publicly traded German semiconductor company listed on the Frankfurt Stock Exchange (FSE: HSAG) and on the Nasdaq Global Select Market through its American Depositary Receipt program (Nasdaq: HCYN), in connection with the proposed acquisition of Vectris Photonics, Inc. ("Vectris"), a Delaware-incorporated fabless semiconductor design company. The proposed transaction is structured as an all-cash reverse triangular merger pursuant to the Agreement and Plan of Merger dated August 10, 2023 (the "Merger Agreement"), under which Halcyon Acquisition Sub, Inc., a newly formed, wholly owned subsidiary of Halcyon, will merge with and into Vectris, with Vectris surviving the merger as a wholly owned subsidiary of Halcyon. The aggregate equity consideration for the transaction is approximately $800,000,000, representing approximately $13.70 per fully diluted share of Vectris common stock.
Halcyon submits this white paper in response to the Division's Second Request, issued on September 15, 2023, and in light of preliminary concerns communicated by the Division's staff during the parties' initial meeting on September 28, 2023. Halcyon appreciates the opportunity to present the Division with a comprehensive analysis of the competitive effects of the proposed transaction and respectfully submits that the acquisition will not substantially lessen competition in any relevant market within the meaning of Section 7 of the Clayton Act, 15 U.S.C. § 18.
The Division's preliminary concerns appear to center on the combination of Halcyon's and Vectris's respective positions in the market for silicon photonics transceiver modules sold in the United States. Halcyon recognizes and takes seriously the Division's mandate to preserve competitive markets, but respectfully submits that those preliminary concerns are based on an unduly narrow market definition that fails to account for the competitive realities of the optical interconnect industry. In particular, the Division's preliminary market definition does not adequately reflect the substantial demand-side and supply-side substitutability among competing optical transceiver and interconnect technologies, the vigorously competitive landscape in which multiple well-funded competitors operate, the low barriers to entry and expansion that characterize the fabless semiconductor design model, the significant countervailing buyer power exercised by the hyperscale cloud operators who are the primary purchasers of these products, or the substantial pro-competitive efficiencies that this transaction will generate.
The core arguments presented in this white paper may be summarized as follows:
First, the relevant product market is broader than "silicon photonics transceiver modules for hyperscale data center interconnects." The appropriate market encompasses the full range of optical transceiver and interconnect modules for data center applications, including traditional pluggable optical transceiver modules based on indium phosphide and other III-V compound semiconductor technologies, co-packaged optics solutions from non-silicon-photonics platforms, active optical cables, and VCSEL-based solutions. Under this broader and more economically accurate market definition, the combined entity's share would be approximately 18–22%, well below the levels that ordinarily raise competitive concerns.
Second, even under the Division's narrower preliminary market definition, current market shares significantly overstate the combined entity's competitive significance. The silicon photonics industry is characterized by rapid technological change, intense research-and-development competition, and frequent product cycle turnover. In such dynamic, innovation-driven markets, current market shares are poor predictors of future competitive significance, a principle well recognized by the Horizontal Merger Guidelines and the case law.
Third, barriers to entry and expansion in silicon photonics transceiver design are low. The fabless business model—exemplified by Vectris itself—dramatically reduces the capital requirements for entry. Multiple foundries offer silicon photonics fabrication services, design talent is increasingly available, and no single firm controls a blocking patent position. Several new entrants have announced silicon photonics programs in the past 24 months, and at least three firms are expected to begin volume shipments of competing products within the next 12–18 months.
Fourth, the customers who purchase silicon photonics transceiver modules are among the largest and most sophisticated technology companies in the world. These hyperscale data center operators maintain multi-sourcing strategies as a matter of corporate policy, qualify and purchase from multiple transceiver suppliers simultaneously, and possess the financial resources and technical capability to develop their own in-house silicon photonics solutions—a threat that provides a powerful competitive constraint on all transceiver suppliers.
Fifth, the proposed transaction generates substantial, verifiable, merger-specific efficiencies that will benefit consumers. The combination of Halcyon's fabrication capabilities, analog integration expertise, and power management intellectual property with Vectris's silicon photonics design leadership and customer relationships is fundamentally complementary. The integration will enable monolithic integration of photonic and electronic functions, faster time-to-market for next-generation products, improved manufacturing yields, and an estimated 15–20% reduction in per-unit production costs—efficiencies that cannot be achieved through arm's-length commercial relationships.
Sixth, the acquisition does not eliminate a maverick or disruptive competitor. While Vectris is an innovative and successful company, its pricing behavior has been consistent with industry norms, and the silicon photonics transceiver market is characterized by multiple innovative competitors, none of which plays a unique disciplinary role.
For all of these reasons, and as discussed in detail below, the proposed acquisition of Vectris by Halcyon does not violate Section 7 of the Clayton Act, and Halcyon respectfully urges the Division to close its investigation and permit the transaction to proceed.
II. BACKGROUND ON THE PARTIES
A. Halcyon Semiconductor AG
Halcyon Semiconductor Aktiengesellschaft is a publicly traded semiconductor company organized and existing under the laws of the Federal Republic of Germany. The company is headquartered at Halcyon-Platz 1, 81739 München, Germany, and is registered with the Munich commercial register under registration number HRB 178254. Halcyon's shares are listed on the Frankfurt Stock Exchange under the ticker symbol HSAG, and the company also maintains an American Depositary Receipt program listed on the Nasdaq Global Select Market under the ticker symbol HCYN. As of the signing of the Merger Agreement on August 10, 2023, Halcyon's approximate market capitalization was €22.4 billion.
Halcyon's core business is the design and manufacture of advanced analog and mixed-signal semiconductors, power management integrated circuits, and radio-frequency components serving a diverse range of end markets, including automotive electronics, industrial automation, telecommunications infrastructure, and data center equipment. The company reported fiscal year 2022 revenues of approximately €6.8 billion and employs approximately 28,000 people worldwide.
Halcyon operates three semiconductor fabrication facilities: a 200mm/300mm fab in Munich, Germany, which serves as the company's primary analog and mixed-signal manufacturing center; a 300mm advanced process fab in Dresden, Germany, which specializes in power management and radio-frequency device fabrication; and a 200mm fab in Singapore, which primarily serves the company's automotive and industrial product lines. These facilities represent Halcyon's substantial investment in vertically integrated manufacturing capabilities, which are central to the company's quality control, process innovation, and cost structure.
Halcyon's activities in the silicon photonics space are a relatively small product line within its broader semiconductor portfolio. The company entered the silicon photonics market approximately four years ago through internal R&D investment and a small tuck-in acquisition of a photonics design team based in Glenrock Assurance, Switzerland. Halcyon currently offers a limited range of silicon photonics transceiver modules targeted at 400G data center interconnect applications. The Division's preliminary estimates attribute approximately 19% of U.S. silicon photonics transceiver module sales to Halcyon—a figure that, while meaningful, reflects a modest competitive position within a rapidly evolving market segment where multiple competitors are investing heavily and where leadership positions shift frequently.
Halcyon's senior management includes Dr. Katarina Breuer, who serves as Chief Executive Officer; Markus Ellinger, who serves as Chief Financial Officer; and Dr. Luisa Hartmann, who serves as General Counsel.
B. Vectris Photonics, Inc.
Vectris Photonics, Inc. is a Delaware C-corporation (Delaware Secretary of State File No. 5918437) headquartered at 4750 Ridgepoint Drive, Suite 300, San Jose, California 95134. Vectris is a fabless designer of silicon photonics transceivers and co-packaged optics chiplets for hyperscale data center interconnects. The company was founded in 2013 by Dr. Anand Raghavan, a former professor of electrical engineering at Stanford University, and has grown to become a recognized innovator in the silicon photonics transceiver space.
For fiscal year 2022, Vectris reported revenues of approximately $312 million and EBITDA of approximately $74 million. The company employs approximately 1,450 people, the majority of whom are engineers and technical personnel located at its San Jose headquarters and at a design center in Austin, Texas. Vectris's engineering team includes a number of leading experts in silicon photonics design, photonic integrated circuit architecture, and high-speed optical signaling.
As a fabless company, Vectris does not own or operate semiconductor fabrication facilities. Instead, Vectris designs its silicon photonics chips and contracts with third-party foundries and outsourced semiconductor assembly and test ("OSAT") providers for fabrication, packaging, and testing. This fabless model has allowed Vectris to focus its resources on design innovation while leveraging the manufacturing capabilities of established fabrication partners.
Vectris holds a substantial intellectual property portfolio, including 187 issued United States patents and 43 pending patent applications, primarily in the areas of silicon photonics device design, transceiver module architecture, co-packaged optics integration, and high-speed optical modulation techniques. While this portfolio is valuable, it represents only a fraction of the thousands of patents held by dozens of companies and research institutions in the broader silicon photonics space, as discussed further in Section V below.
Vectris's key customers include three of the top five U.S. hyperscale cloud operators, measured by revenue. These customer relationships, while significant, are non-exclusive, and each of Vectris's major customers also sources optical transceiver products from multiple competing suppliers, as discussed in detail in Section VI.
Vectris's senior management includes Dr. Anand Raghavan, who serves as Founder and Chief Executive Officer, and Margaret Sørensen, who serves as President and Chief Operating Officer. Vectris's significant shareholders include Dr. Raghavan, who holds approximately 18.2% of the outstanding shares; Granite Peak Ventures Fund IV, L.P., which holds approximately 22.7%; Nordhaven Capital Partners III, L.P., which holds approximately 14.5%; Stonebridge Institutional Investors, L.P., which holds approximately 9.1%; with the remaining approximately 35.5% constituting the public float.
C. Transaction Overview
On August 10, 2023, Halcyon and Vectris executed the Merger Agreement, pursuant to which Halcyon Acquisition Sub, Inc., a newly formed Delaware corporation and wholly owned subsidiary of Halcyon, will merge with and into Vectris in a reverse triangular merger. Upon consummation of the merger, Halcyon Acquisition Sub will cease to exist, and Vectris will survive as a wholly owned subsidiary of Halcyon. The transaction is structured as an all-cash acquisition at an aggregate equity value of approximately $800 million.
Under the terms of the Merger Agreement, each outstanding share of Vectris common stock will be converted into the right to receive approximately $13.70 in cash, without interest. This per-share merger consideration represents an approximately 39.5% premium over Vectris's unaffected closing stock price of $9.82 on May 19, 2023, which was the last trading day prior to market speculation regarding a potential transaction. As of the date of the Merger Agreement, Vectris had approximately 51,200,000 shares of common stock outstanding on a basic basis and approximately 58,394,160 shares outstanding on a fully diluted basis, including approximately 7,194,160 shares underlying in-the-money stock options and restricted stock units. The implied valuation multiples are approximately 2.6× enterprise value to revenue and approximately 10.8× enterprise value to EBITDA, based on Vectris's fiscal year 2022 financial results.
The transaction is being financed through a combination of a $500 million senior secured term loan commitment from Meridian Atlantic Bank, N.A. (the commitment letter was executed on August 8, 2023) and approximately $300 million from Halcyon's existing cash reserves. The financing is not subject to any material conditions beyond those customary for transactions of this type.
Halcyon and Vectris filed their respective Hart-Scott-Rodino Act premerger notification forms with the Federal Trade Commission and the Department of Justice on August 18, 2023. On September 15, 2023, the Antitrust Division issued a Second Request in the form of a Civil Investigative Demand directed to both parties. Halcyon is cooperating fully with the Division's investigation and has been engaged in the process of collecting, reviewing, and producing responsive documents and data. Halcyon submits this white paper to provide the Division with a complete and accurate picture of the competitive landscape and to demonstrate that the proposed transaction does not raise competitive concerns warranting further investigation or enforcement action.
III. THE RELEVANT MARKET IS BROADER THAN THE DIVISION'S PRELIMINARY DEFINITION
Market definition is a critical threshold question in any merger analysis because it provides the framework within which competitive effects are assessed. As the Supreme Court has long recognized, a properly defined relevant market must reflect the commercial realities of the industry and the actual competitive alternatives available to customers. See Brown Shoe Co. v. United States, 370 U.S. 294, 336–37 (1962) (relevant market must correspond to "the commercial realities of the industry"); see also U.S. Dep't of Justice & FTC, Horizontal Merger Guidelines § 4.0 (2010) ("The Agencies' analysis need not start with market definition."), § 4.1.1 (defining relevant markets based on the hypothetical monopolist test). Halcyon respectfully submits that the Division's preliminary market definition is too narrow and does not accurately reflect the competitive dynamics that constrain the pricing and innovation behavior of silicon photonics transceiver suppliers.
A. The Division's Preliminary Market Definition
Based on communications with the Division's staff and the scope of the Second Request, it appears that the Division is considering a relevant product market defined as "silicon photonics transceiver modules for hyperscale data center interconnects," limited to sales in the United States. Under this preliminary market definition, the Division's staff have attributed approximately 28% of market sales to Vectris and approximately 19% of market sales to Halcyon, yielding a combined post-merger share of approximately 47%.
Halcyon recognizes that these preliminary share figures, taken at face value and without context, might appear to suggest a degree of market concentration warranting scrutiny. However, as demonstrated below, this preliminary market definition is based on an artificially narrow delineation of the relevant product market that fails to account for the full range of competitive alternatives available to data center operators. The Division's focus on silicon photonics as a distinct product market treats a particular underlying technology platform as the defining characteristic of the relevant market, when in fact data center operators make purchasing decisions based on the functional performance, cost, and reliability of optical interconnect solutions irrespective of the underlying semiconductor technology employed.
Halcyon respectfully submits that the Division's preliminary definition fails to satisfy the hypothetical monopolist test set forth in the Horizontal Merger Guidelines. A hypothetical monopolist of silicon photonics transceiver modules alone could not profitably impose a small but significant and non-transitory increase in price ("SSNIP") because customers would respond by substituting to competing optical transceiver and interconnect technologies that serve the identical functional purpose. The evidence supporting this conclusion is set forth in the following subsections.
B. Demand-Side Substitutability
The fundamental function served by silicon photonics transceiver modules is the transmission of data at high speeds within and between data center facilities. Data center operators do not purchase silicon photonics transceivers because of an inherent preference for the silicon photonics technology platform itself; rather, they select optical transceiver and interconnect solutions based on a set of performance, cost, and operational criteria—including data rate, reach, power consumption, form factor, reliability, and per-port cost—that can be satisfied by multiple competing technologies.
The relevant competitive alternatives that serve as demand-side substitutes for silicon photonics transceiver modules include, at a minimum, the following:
(1) Traditional Pluggable Optical Transceiver Modules Based on Indium Phosphide and Other III-V Compound Semiconductor Technologies. These modules have been the dominant technology in data center optical interconnects for more than a decade and continue to represent the majority of transceiver module shipments by volume. Indium phosphide (InP)-based transceivers are available in the same form factors (e.g., QSFP-DD, OSFP) and at the same data rates (400G, 800G) as silicon photonics transceivers. While silicon photonics offers certain potential advantages in integration density and manufacturability, InP-based transceivers offer competitive performance and are available from numerous established suppliers. Data center operators routinely evaluate and switch between silicon photonics and InP-based transceivers based on price, availability, and performance considerations.
(2) Co-Packaged Optics (CPO) Solutions from Non-Silicon-Photonics Platforms. The emerging co-packaged optics segment encompasses a range of technologies that integrate optical interconnect functions directly with network switch ASICs or other data center processors. While some CPO solutions utilize silicon photonics technology, others are based on alternative photonic integration platforms, including thin-film lithium niobate, polymer waveguide, and hybrid III-V/silicon approaches. These non-silicon-photonics CPO solutions compete directly with silicon photonics-based CPO offerings for the same hyperscale data center design wins.
(3) Active Optical Cables (AOCs). Active optical cables provide a pre-terminated, plug-and-play optical interconnect solution that competes with transceiver modules for shorter-reach data center interconnect applications. AOCs incorporate optical transceivers within the cable assembly itself and are widely used in top-of-rack and intra-rack data center connections. For a significant portion of data center interconnect applications, AOCs represent a direct functional substitute for pluggable transceiver modules.
(4) Vertical-Cavity Surface-Emitting Laser (VCSEL)-Based Solutions. VCSEL-based multimode transceivers and AOCs remain a widely deployed and cost-effective solution for shorter-reach data center interconnects, particularly at distances up to approximately 100 meters. While the industry has been trending toward single-mode solutions (including silicon photonics) for longer-reach and higher-data-rate applications, VCSEL-based solutions continue to compete for a substantial share of data center interconnect ports, and ongoing advances in VCSEL technology are extending the reach and data rate capabilities of these products.
The demand-side substitutability of these technologies is supported by substantial pricing evidence. Halcyon has provided the Division with transactional pricing data and customer correspondence demonstrating that when silicon photonics transceiver prices have increased—whether due to supply constraints, component cost increases, or other factors—hyperscale customers have shifted procurement volumes to competing technologies, including InP-based pluggable transceivers and active optical cables. This observed customer switching behavior is a strong indicator of high cross-elasticity of demand and confirms that a hypothetical monopolist of silicon photonics transceivers alone could not profitably impose a SSNIP.
The pricing evidence is corroborated by customer Request for Quotation ("RFQ") and Request for Proposal ("RFP") documents produced in the ordinary course of business, which routinely solicit bids from transceiver suppliers across multiple technology platforms simultaneously. These procurement documents reflect the commercial reality that data center operators view silicon photonics transceivers as one option among several competing technologies, not as a distinct product market.
Accordingly, the appropriate relevant product market should be defined as "optical transceiver and interconnect modules for data center applications," encompassing the full range of pluggable transceiver modules, co-packaged optics solutions, active optical cables, and VCSEL-based solutions used in hyperscale data center environments. Under this broader—and more economically sound—market definition, the combined entity's share of the relevant market would be significantly lower than under the Division's narrow definition, estimated at approximately 18–22%. This share level is well below the thresholds that ordinarily raise competitive concerns under the Horizontal Merger Guidelines.
C. Supply-Side Substitutability
Even setting aside the extensive demand-side substitution evidence, the relevant market should be broadened to account for supply-side substitutability. The Horizontal Merger Guidelines recognize that market definition may be expanded where firms that do not currently produce the relevant product could rapidly and economically shift production to compete in the relevant market in response to a SSNIP, without incurring significant sunk costs. See Horizontal Merger Guidelines § 5.1 (2010).
Numerous semiconductor firms with established design and fabrication capabilities in adjacent optical technologies—including InP-based transceiver manufacturers, photonic integrated circuit designers, and diversified semiconductor companies with optical communications product lines—possess the technical expertise, fabrication relationships, and customer channels necessary to reposition their production to silicon photonics transceivers within a commercially relevant timeframe. The technical skills required for silicon photonics design overlap substantially with those required for other optical semiconductor design disciplines, and the availability of third-party foundry services specializing in silicon photonics fabrication further reduces the barriers to supply-side repositioning.
The supply-side substitution evidence is compelling. Within the past 24 months alone, at least four major semiconductor companies have publicly announced silicon photonics product development programs. Three additional companies are known to have functional silicon photonics transceiver prototypes currently under customer evaluation at hyperscale data center operators. These announcements and development programs represent concrete evidence that supply-side repositioning into silicon photonics transceivers is not merely hypothetical—it is actively occurring.
Furthermore, foundry and OSAT capacity for silicon photonics is commercially available from multiple sources. Several leading semiconductor foundries now offer dedicated silicon photonics process nodes with mature process design kits ("PDKs"), providing fabless designers with turnkey access to silicon photonics fabrication. The availability of these foundry services means that a firm with optical semiconductor design expertise can enter the silicon photonics transceiver market without making the multi-billion-dollar capital investments required to build a fabrication facility—a point underscored by Vectris's own success as a fabless silicon photonics company.
Under the framework of the Horizontal Merger Guidelines, supply-side substitution of this character—where repositioning can occur rapidly (within approximately one year), without the need for significant sunk investments, and where the repositioning firms would likely provide a competitive constraint comparable to that of existing market participants—supports defining the relevant market to include the broader set of optical transceiver and interconnect technologies.
D. Geographic Market
Halcyon submits that the relevant geographic market is worldwide, or at a minimum encompasses the United States, Europe, and the Asia-Pacific region. Data center operators source transceiver modules globally, and transceiver suppliers compete on a worldwide basis for customer design wins and procurement contracts.
The silicon photonics transceiver industry is characterized by global supply chains and global competition. Halcyon and Vectris compete against firms headquartered in the United States, Europe, China, Japan, and Taiwan, all of which actively sell transceiver products into the U.S. market through direct sales, distribution partnerships, or both. Transportation costs for transceiver modules are trivially low relative to product value—a single transceiver module may weigh only a few grams and can be shipped by air from any point on the globe to a U.S. data center within days at negligible cost. There are no significant tariff barriers, import restrictions, or regulatory obstacles that would impede the sale of transceiver modules manufactured outside the United States to U.S.-based data center operators.
Even if the Division were to define the geographic market as the United States alone, the competitive analysis must account for the constraint imposed by imports and by foreign-based competitors who actively sell into the U.S. market. Several of the leading competitors in optical transceiver modules are headquartered outside the United States and derive a significant portion of their revenues from sales to U.S. hyperscale data center operators. These foreign-based competitors maintain U.S. sales offices, U.S.-based technical support teams, and U.S. warehousing and logistics capabilities, and they participate in the same RFQ and qualification processes as U.S.-based suppliers. Excluding these competitors from the competitive analysis would significantly distort the picture of competitive conditions in the U.S. market.
IV. MARKET SHARES OVERSTATE COMPETITIVE SIGNIFICANCE
Even if the Division were to adopt the narrower market definition of "silicon photonics transceiver modules for hyperscale data center interconnects"—a definition that Halcyon believes is economically unsound for the reasons set forth in Section III—the resulting market share figures significantly overstate the competitive significance of the proposed transaction. Current market shares in the silicon photonics transceiver industry are a particularly unreliable indicator of future competitive conditions due to the dynamic, innovation-driven nature of the market and the vigorous pipeline competition that characterizes the industry.
A. Market Shares in Innovation-Driven Markets
The silicon photonics transceiver industry is characterized by rapid technological change, intense R&D competition, and frequent product cycle turnover, with major new product generations typically introduced on 12- to 24-month cycles. In this environment, competitive advantage is driven primarily by innovation—the ability to design and bring to market next-generation products that offer higher data rates, lower power consumption, smaller form factors, and lower per-bit cost. A firm's current market share reflects its success in prior product generations and may bear little relationship to its competitive position in the next generation of products.
The Horizontal Merger Guidelines expressly recognize that "market shares may not fully reflect the competitive significance of firms in the market or the impact of the merger" in markets characterized by rapid technological change. Horizontal Merger Guidelines § 5.3 (2010). Similarly, the courts have consistently held that in dynamic, innovation-driven industries, current market shares are less probative of competitive effects than in mature, stable markets. See, e.g., United States v. Tarvick Hughes Inc., 908 F.2d 981, 986 (D.C. Cir. 1990) (market share statistics must be "supplemented by other evidence" and "may provide an inaccurate account of the merger's probable effects on competition").
Historical market share data in the silicon photonics transceiver market vividly illustrate this point. Firms that held leading positions just three to five years ago have seen their shares decline dramatically as new technologies emerged and as competitors introduced superior products. Conversely, firms with modest shares several years ago have rapidly gained share through successful product innovation. This pattern of volatile and shifting market shares is a strong indicator that current share figures are a poor predictor of the competitive landscape that will exist 12, 24, or 36 months from now—the timeframe most relevant to the Division's competitive effects analysis.
B. Concentration Metrics Under the Broader Market Definition
Under the broader relevant product market definition advocated by Halcyon in Section III—optical transceiver and interconnect modules for data center applications—the pre-merger Herfindahl-Hirschman Index ("HHI") is significantly below the 2,500-point threshold that the Horizontal Merger Guidelines identify as indicative of a "highly concentrated" market. The change in HHI (delta) attributable to the proposed transaction under this broader definition would be well below the 200-point threshold that typically triggers competitive concern under the Guidelines.
Specifically, based on Halcyon's analysis of publicly available revenue data and third-party market research reports, the pre-merger HHI under the broader market definition is estimated at approximately 1,100–1,400, indicating a moderately concentrated market. The delta attributable to the transaction is estimated at approximately 80–120 points. Under the Horizontal Merger Guidelines framework, mergers resulting in a delta of less than 100 points in a moderately concentrated market are "unlikely to have adverse competitive effects and ordinarily require no further analysis." Horizontal Merger Guidelines § 5.3 (2010). Even at the upper end of the estimated range, the delta of approximately 120 points falls well below the 200-point threshold that triggers a presumption of competitive concern.
Appendix A to this white paper sets forth detailed market share estimates under both the narrow and broad market definitions, together with corresponding HHI calculations and supporting sources.
C. Pipeline Competition and Innovation
The competitive significance of current market shares is further diminished by the robust pipeline competition in the silicon photonics transceiver space. Multiple competitors have active silicon photonics product development programs at advanced stages, including prototyping, customer sampling, and formal qualification testing. The customer qualification cycle for new transceiver suppliers at major hyperscale data center operators typically takes 6–12 months, and several competitors are currently in or nearing completion of qualification programs with major hyperscale operators.
This pipeline competition ensures that, even under the Division's narrow market definition, the competitive landscape will look markedly different within the next 12–24 months than it does today. New products from multiple competitors are poised to enter the market, providing additional competitive alternatives for hyperscale customers and further constraining the pricing and output behavior of existing suppliers—including any combined Halcyon-Vectris entity.
Furthermore, Vectris's competitive position, while strong today, is not so unique or durable that its acquisition by Halcyon would materially diminish the pace or scope of innovation in the silicon photonics space. Vectris's innovations have been important, but they have occurred in the context of an industry with multiple well-funded R&D programs, active academic research, and a deep pool of engineering talent. The key technologies underlying silicon photonics transceivers—including silicon waveguide design, germanium photodetectors, hybrid laser integration, and high-speed modulator architectures—are the subject of extensive academic publication and are understood by the broader engineering community. Vectris does not possess any unique, non-replicable technological capability that would be "lost" to the competitive marketplace as a result of this transaction.
V. BARRIERS TO ENTRY AND EXPANSION ARE LOW
The Horizontal Merger Guidelines recognize that the prospect of entry by new competitors or expansion by existing competitors can counteract any potential anticompetitive effects of a merger. Entry is most likely to be effective when it is "timely, likely, and sufficient"—that is, when new entry can occur within approximately two years, when it is economically rational for potential entrants to invest in entering the market, and when the scale of entry would be sufficient to replace the competitive constraint lost through the merger. See Horizontal Merger Guidelines § 9 (2010). The silicon photonics transceiver market satisfies all three of these criteria.
A. Fabless Business Model Reduces Capital Barriers
The most significant barrier to entry in many segments of the semiconductor industry is the enormous capital investment required to build and operate semiconductor fabrication facilities. Modern leading-edge fabs can cost $10 billion or more to construct, and even mature-node fabs require investments of several billion dollars. However, the fabless semiconductor business model—in which companies design chips but outsource fabrication to third-party foundries—has dramatically reduced the capital requirements for entry into semiconductor design markets, including silicon photonics.
Vectris itself is the clearest possible evidence that the fabless model works in silicon photonics. Vectris has built a successful $312 million revenue business and an estimated 28% share of the U.S. silicon photonics transceiver market without owning a single fabrication facility. Vectris designs its photonic integrated circuits and contracts with third-party foundries for fabrication, relying on commercially available silicon photonics process nodes. If Vectris could achieve this scale as a startup, there is no reason why other well-resourced firms cannot do the same.
Access to foundry services for silicon photonics is commercially available and expanding. Multiple leading semiconductor foundries now offer dedicated silicon photonics process nodes, complete with mature process design kits, design rule manuals, and customer support infrastructure. These foundry services enable fabless designers to access state-of-the-art silicon photonics fabrication without bearing the capital cost of building and maintaining a fab. In addition, multiple OSAT providers offer packaging, assembly, and testing services tailored to silicon photonics transceiver modules, further reducing the need for entrants to invest in their own manufacturing infrastructure.
Based on Halcyon's analysis of industry economics, the estimated capital required for a new fabless silicon photonics design company to achieve initial product qualification—including funding for design engineering, prototype fabrication, testing, and initial customer qualification—is approximately $50–100 million. While this is a substantial investment in absolute terms, it is modest by semiconductor industry standards, where individual product development programs at major companies routinely involve investments of hundreds of millions of dollars. Moreover, this level of investment is well within the reach of the established semiconductor companies, well-funded startups, and venture capital-backed firms that constitute the potential entrant pool for silicon photonics.
B. Access to Talent and IP
Silicon photonics design talent is increasingly available from major research universities and industry programs. Programs at leading academic institutions have produced a growing pipeline of engineers with expertise in photonic integrated circuit design, silicon waveguide engineering, and high-speed optical signaling. Industry training programs and the natural mobility of engineers among semiconductor companies further ensure that design expertise is broadly distributed and accessible to potential entrants.
While Vectris holds a significant patent portfolio—187 issued U.S. patents and 43 pending applications—this portfolio must be evaluated in the context of the broader silicon photonics intellectual property landscape. That landscape includes thousands of patents held by dozens of companies and research institutions worldwide. No single firm, including Vectris, controls a blocking patent position that would prevent competitors from designing, manufacturing, and selling silicon photonics transceiver products. Design-around opportunities exist for key Vectris patents, as is common in the semiconductor industry, where multiple technical approaches can achieve equivalent functional results. Furthermore, cross-licensing of semiconductor patents is standard industry practice, and Halcyon is not aware of any instance in which Vectris or any other silicon photonics firm has used its patent portfolio to exclude a competitor from the market.
The absence of a blocking patent position is confirmed by the fact that multiple competitors are currently developing and selling silicon photonics transceiver products without any license from Vectris, and without any apparent concern that Vectris's patents would impede their market participation. This competitive reality undermines any argument that Vectris's intellectual property constitutes a meaningful barrier to entry.
C. Recent and Planned Entry
The theoretical analysis of low barriers to entry is confirmed by the empirical evidence of recent and planned entry into the silicon photonics transceiver market. Within the past 24 months, the following developments illustrate the active and accelerating pace of entry and expansion:
• A major Asian semiconductor conglomerate with established capabilities in compound semiconductor devices publicly announced the launch of a silicon photonics transceiver product line targeting 400G and 800G data center applications. This firm has committed more than $200 million in R&D investment to its silicon photonics program and has begun customer qualification activities with multiple hyperscale operators.
• A European photonics startup, backed by more than €150 million in venture capital funding, has developed a novel silicon photonics transceiver architecture and has shipped engineering samples to prospective customers. The company has announced plans to begin volume production in the first half of 2024.
• A leading U.S.-based diversified semiconductor company with a strong existing position in optical communication components has expanded its product portfolio to include silicon photonics transceiver modules, leveraging its existing customer relationships and global sales infrastructure. This firm's silicon photonics products are currently in the advanced stages of customer qualification at two major hyperscale operators.
• A well-funded silicon photonics startup based in the San Francisco Bay Area, backed by a consortium of strategic and financial investors, emerged from stealth mode in early 2023 and has announced a product roadmap targeting co-packaged optics applications using a proprietary silicon photonics platform. The company has recruited a senior engineering team from several established optical semiconductor companies.
• Multiple additional semiconductor companies are known to have active internal silicon photonics R&D programs that have not yet been publicly announced but are reflected in patent filings, academic publications, and industry conference presentations.
In addition to these specific examples, broader venture capital investment trends underscore the attractiveness and accessibility of the silicon photonics transceiver market. According to publicly available data, venture capital investment in silicon photonics startups has accelerated significantly over the past two years, with total disclosed funding exceeding $1.2 billion across more than 15 companies in the 2022–2023 period. This level of venture capital interest reflects the investor community's assessment that the silicon photonics market offers attractive growth opportunities and that barriers to entry are manageable.
Based on publicly available information and Halcyon's industry knowledge, at least three companies are expected to begin volume shipments of competing silicon photonics transceiver products within the next 12–18 months. These anticipated market entries will further expand the set of competitive alternatives available to hyperscale data center operators and will increase competitive pressure on all existing market participants.
D. Timely, Likely, and Sufficient Entry
Under the framework of the Horizontal Merger Guidelines, entry into the silicon photonics transceiver market satisfies all three requirements for entry to be considered an effective competitive check:
Timely. Entry into silicon photonics transceiver design and sales is achievable within approximately two years or less—well within the timeframe contemplated by the Guidelines. The availability of foundry services, mature PDKs, and experienced design talent enables new entrants to move from initial design to customer qualification in 12–18 months, with volume production following shortly thereafter. The examples of recent entry described above confirm that this timeline is realistic and achievable.
Likely. Entry is economically rational and is demonstrably occurring. The silicon photonics transceiver market is growing rapidly, driven by the explosive growth in hyperscale data center traffic and the industry transition to higher data rates. This growth creates strong economic incentives for entry, and the evidence shows that multiple firms are actively investing to capitalize on these opportunities. The likelihood of entry is further confirmed by the significant venture capital investment flowing into silicon photonics startups.
Sufficient. The scale of expected entry is sufficient to replace any competitive constraint that might theoretically be lost through the merger. The entering firms include major semiconductor companies with substantial financial resources, global sales infrastructure, and established customer relationships, as well as well-funded startups with innovative technologies and aggressive growth plans. These entrants are capable of achieving meaningful market share within a commercially relevant timeframe.
Under the Horizontal Merger Guidelines, entry that is timely, likely, and sufficient would be expected to prevent any anticompetitive price increase or output reduction that might otherwise result from the proposed transaction. The robust entry conditions in the silicon photonics transceiver market provide an additional and independent basis for concluding that the transaction does not raise competitive concerns.
VI. SOPHISTICATED CUSTOMERS WITH SIGNIFICANT COUNTERVAILING BUYER POWER
The Horizontal Merger Guidelines recognize that the competitive effects of a merger may be mitigated where customers are large, sophisticated, and possess significant bargaining leverage over their suppliers. The silicon photonics transceiver market is characterized by precisely this dynamic: the primary purchasers are among the largest and most powerful technology companies in the world, and they exercise substantial countervailing buyer power that constrains the pricing and competitive behavior of transceiver suppliers.
A. Customer Concentration and Sophistication
The primary purchasers of silicon photonics transceiver modules for hyperscale data center interconnects are the largest technology companies in the world, including the leading operators of hyperscale cloud computing platforms. Three of the top five U.S. hyperscale cloud operators, measured by revenue, are Vectris customers. These companies are among the most sophisticated and well-resourced technology buyers globally, with annual capital expenditures in the tens of billions of dollars and dedicated procurement teams with deep technical expertise in optical interconnect technologies.
These hyperscale customers employ engineers and procurement specialists who possess detailed knowledge of transceiver technology, manufacturing processes, cost structures, and competitive alternatives. They conduct rigorous technical evaluations of transceiver products, including performance benchmarking, reliability testing, and detailed cost analysis, before qualifying a supplier for production procurement. This level of customer sophistication means that transceiver suppliers, including Halcyon and Vectris, cannot extract supracompetitive pricing or reduce product quality without risking the loss of customer business to competing suppliers.
The concentrated nature of the customer base further amplifies buyer power. A relatively small number of hyperscale data center operators account for the majority of silicon photonics transceiver module demand. The loss of a single major customer can have a material impact on a transceiver supplier's revenue and profitability. This customer concentration gives hyperscale operators significant leverage in negotiations with transceiver suppliers and ensures that suppliers compete aggressively for customer business.
B. Multi-Sourcing Strategies
Hyperscale data center operators, as a matter of corporate policy and supply chain risk management best practices, maintain multi-sourcing strategies for optical transceiver and interconnect modules. These customers qualify and purchase from multiple transceiver suppliers simultaneously to ensure supply security, maintain competitive pricing pressure, and preserve the flexibility to shift volume among suppliers in response to changing market conditions.
The multi-sourcing approach of hyperscale customers is a deliberate and well-established practice. Each major hyperscale operator maintains a qualified supplier list ("QSL") that typically includes three or more approved transceiver suppliers for each major product category. Procurement volumes are allocated among qualified suppliers through a competitive bidding process, and allocations can be—and frequently are—adjusted in response to pricing, quality, delivery performance, and new product availability.
This multi-sourcing dynamic means that no single supplier—including Vectris—is a "must-have" for any hyperscale customer. All of Vectris's major customers have alternative qualified suppliers from which they can and do source optical transceiver products. If a combined Halcyon-Vectris entity were to attempt to raise prices or reduce quality post-merger, hyperscale customers would have the ability and the incentive to shift procurement volumes to alternative suppliers, thereby defeating any attempt to exercise market power.
The documentary evidence produced in connection with the Second Request confirms this multi-sourcing reality. Internal procurement documents from Vectris's customers reflect regular competitive bidding processes involving multiple transceiver suppliers, frequent volume allocation adjustments among qualified suppliers, and active customer efforts to qualify additional suppliers to further intensify competitive pressure.
C. Vertical Integration Threat
The countervailing buyer power of hyperscale data center operators is further reinforced by the credible threat of vertical integration. Several hyperscale cloud operators have invested in or are actively developing their own in-house silicon photonics and custom transceiver programs. These internal development programs reflect the strategic importance that hyperscale operators attach to optical interconnect technology and their willingness to invest in developing their own capabilities if external suppliers fail to deliver competitive pricing, quality, and innovation.
The threat of backward integration by hyperscale customers acts as a powerful competitive constraint on all transceiver suppliers, including any combined Halcyon-Vectris entity. If post-merger pricing were to increase or innovation were to slow, hyperscale customers could accelerate their internal development programs and reduce their reliance on external transceiver suppliers. This threat is not hypothetical—it is reflected in the significant investments that multiple hyperscale operators have already made in internal photonics R&D teams, including the recruitment of senior silicon photonics engineers, the filing of photonics-related patents, and the establishment of internal photonics design centers.
The existence of these vertical integration programs imposes a ceiling on the pricing that transceiver suppliers can charge and ensures that suppliers remain focused on continuous innovation and cost reduction. This competitive discipline would continue to operate post-merger with at least as much force as it does today.
D. Customer Testimony
Halcyon understands that the Division has been in contact with certain customers of Halcyon and Vectris as part of its investigation. While Halcyon cannot speak for its customers, Halcyon believes that customer feedback obtained during the investigation is expected to demonstrate that customers do not view the proposed transaction as likely to result in higher prices, reduced innovation, or diminished competitive options. Major hyperscale customers are sophisticated participants in the optical interconnect market and understand that the competitive landscape provides them with ample alternatives to constrain the behavior of any transceiver supplier.
Halcyon notes its willingness to facilitate additional customer interviews and to make available additional customer-facing personnel to assist the Division in obtaining a complete picture of customer perspectives on the transaction. Halcyon is confident that a thorough examination of customer views will confirm that the proposed acquisition does not raise competitive concerns.
VII. THE TRANSACTION GENERATES SUBSTANTIAL PRO-COMPETITIVE EFFICIENCIES
The proposed acquisition of Vectris by Halcyon is motivated by a sound business rationale: the combination of the two companies' complementary capabilities will create a more competitive and innovative enterprise that can better serve the needs of data center operators and compete more effectively against other integrated optical transceiver platforms. The Horizontal Merger Guidelines recognize that cognizable efficiencies can counteract potential anticompetitive effects and may constitute an independent ground for permitting a transaction to proceed. See Horizontal Merger Guidelines § 10 (2010). The efficiencies generated by this transaction are substantial, merger-specific, and verifiable.
A. Complementary Capabilities
Halcyon's core strengths lie in analog and mixed-signal semiconductor design, power management integrated circuit technology, and advanced fabrication capabilities at its three fabs in Munich, Dresden, and Singapore. The company has decades of experience in precision semiconductor manufacturing, advanced process control, and the integration of complex analog and digital functions on a single chip. These capabilities are directly relevant to the manufacture of silicon photonics transceiver modules, which require precise analog signal processing, sophisticated power management, and tight integration of electronic and photonic functions.
Vectris's core strengths, by contrast, lie in silicon photonics transceiver design, co-packaged optics chiplet architecture, and deep relationships with hyperscale data center operators. Vectris has assembled a world-class design engineering team with deep expertise in photonic integrated circuit design, high-speed optical modulation, and advanced transceiver architectures. The company's design leadership is reflected in its strong intellectual property portfolio, its track record of successful product introductions, and its relationships with three of the top five U.S. hyperscale cloud operators.
The combination of these two companies is fundamentally complementary. Halcyon brings fabrication capacity, analog integration expertise, and power management intellectual property that Vectris currently lacks as a fabless company. Vectris brings photonics design leadership, co-packaged optics architecture innovation, and hyperscale customer access that Halcyon currently lacks in its relatively nascent silicon photonics product line. The overlap between the parties' existing product lines is limited—the combined 47% share figure cited by the Division under its narrow market definition reflects overlap only in the silicon photonics transceiver segment where both parties happen to compete, and does not reflect the totality of their businesses or the degree of complementarity between their respective capabilities.
B. Merger-Specific Efficiencies
The integration of Vectris's silicon photonics design capabilities with Halcyon's manufacturing and analog integration expertise would enable a range of significant efficiencies that cannot be achieved through arm's-length commercial relationships.
(1) Monolithic Integration of Photonic and Electronic Functions. Currently, silicon photonics transceiver modules require the assembly of separate photonic and electronic chips, with the two components connected through complex and costly packaging techniques. The combined entity would be able to leverage Halcyon's fabrication capabilities and analog design expertise to develop monolithically integrated photonic-electronic chips, in which photonic and electronic functions are fabricated on the same silicon die. Monolithic integration would significantly reduce manufacturing cost, lower power consumption, reduce module size, and improve performance—benefits that would be passed on to data center operators in the form of lower prices and better products.
(2) Faster Time-to-Market for Next-Generation Co-Packaged Optics Products. The transition to co-packaged optics is one of the most significant trends in data center interconnect technology, and both Halcyon and Vectris have identified CPO as a critical strategic priority. The combined entity would be able to accelerate the development of co-packaged optics products by integrating Vectris's CPO chiplet architecture with Halcyon's advanced packaging and analog integration capabilities. Halcyon's internal analysis estimates that the combination would accelerate development timelines for key CPO product roadmap milestones by approximately 12–18 months, enabling the combined entity to bring next-generation CPO products to market significantly faster than either company could achieve independently.
(3) Improved Manufacturing Yields Through Advanced Process Control. Halcyon's fabrication facilities employ state-of-the-art process control systems that enable exceptionally high manufacturing yields and tight process uniformity. By bringing Vectris's silicon photonics designs in-house for fabrication at Halcyon's fabs, the combined entity would benefit from Halcyon's yield engineering expertise, which is expected to deliver meaningful improvements in manufacturing yield relative to the yields currently achieved by Vectris's third-party foundry partners. Higher yields translate directly into lower per-unit costs and more consistent product quality.
These efficiencies are merger-specific in the meaningful sense required by the Horizontal Merger Guidelines. Vectris, as a fabless company, cannot replicate these integration benefits through arm's-length foundry relationships. While Vectris has access to third-party foundry services, the degree of design-manufacturing co-optimization, the depth of process control integration, and the speed of iterative design-fab feedback loops that are possible within a vertically integrated company are qualitatively different from—and superior to—what can be achieved through a commercial foundry relationship. Halcyon's internal fabrication capabilities enable a level of customization, co-optimization, and process control that third-party foundries, which must serve multiple customers with standardized process nodes, cannot provide.
Similarly, Halcyon cannot develop Vectris's silicon photonics design intellectual property and customer relationsh