1032-00006 · Halcyon Semiconductor Inc · APA Negotiation

supplemental-economic-analysis-sensitivity-testing.docx DOCUMENT M&A / Corporate

14 / 35 ← → ✕

Contract analysis Extract to table

CALDERWOOD & HARKNESS LLP 599 Lexington Avenue, 38th Floor New York, New York 10022

SUPPLEMENTAL ECONOMIC ANALYSIS — SENSITIVITY TESTING

Sensitivity Analyses Supporting APA Negotiation Positions for Halcyon Semiconductor Inc.

C&H Matter Number: 1032-00006 Billing Code: HSAG-TP-2025

June 27, 2025

PRIVILEGED AND CONFIDENTIAL Attorney Work Product / Prepared in Anticipation of Litigation and APA Proceedings

This memorandum constitutes attorney work product prepared by Calderwood & Harkness LLP in anticipation of Advance Pricing Agreement proceedings and potential litigation. It is protected from disclosure by the attorney work product doctrine and the attorney-client privilege. No portion of this memorandum may be disclosed to any third party, including any taxing authority, without the prior written authorization of Calderwood & Harkness LLP.

TO:

Leonard Cross, Partner, Tax — Calderwood & Harkness LLP Sylvia Hartwell, Partner, Tax — Calderwood & Harkness LLP Leila Nazari, Senior Associate, Tax — Calderwood & Harkness LLP Robert Jansen, Associate, Tax — Calderwood & Harkness LLP

COPY TO:

Dr. Annika Breitner, Managing Director — Meridian Economics GmbH, Leopoldstraße 145, 80804 Munich, Germany Maximilian Voss, Senior Analyst — Meridian Economics GmbH, Leopoldstraße 145, 80804 Munich, Germany

PREPARED BY:

Calderwood & Harkness LLP, in conjunction with Meridian Economics GmbH

Right-click to update Table of Contents

I. Executive Summary

This memorandum presents supplemental sensitivity analyses designed to stress-test and corroborate the primary economic analyses submitted in the bilateral Advance Pricing Agreement ("APA") application for Halcyon Semiconductor AG ("HSAG") and Halcyon Semiconductor Inc. ("HSI"). The APA is sought between the Internal Revenue Service, Advance Pricing and Mutual Agreement Program ("IRS APMA") and the Bundeszentralamt für Steuern ("BZSt") under Article 25 of the Convention Between the United States of America and the Federal Republic of Germany for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and Capital and to Certain Other Taxes (signed June 29, 2006, as amended) (the "US-Germany Treaty"), pursuant to Rev. Proc. 2015-41.

HSAG is an Aktiengesellschaft organized under the laws of the Federal Republic of Germany, with its registered office at Halcyon-Allee 12, 01099 Dresden, Saxony, Germany (Steuernummer 203/118/04527). HSI is a corporation organized under the laws of the State of Delaware (EIN 84-2917630), with its principal place of business at 8700 Parmer Lane, Building 4, Austin, Texas 78729. HSI is a wholly owned subsidiary of HSAG and serves as the principal North American distribution and sales entity for the Halcyon group.

The APA application covers three categories of intercompany transactions between HSAG and HSI:

1. Tangible Goods (Semiconductor Wafer/Die Sales). HSAG manufactures and sells semiconductor wafers and dies to HSI for resale to third-party customers in North America. In fiscal year 2024, intercompany tangible goods sales totaled approximately €987.6 million (approximately $1,076 million at the average 2024 EUR/USD exchange rate of 1.0896).

2. Royalties (IP License). HSI pays royalties to HSAG under a license agreement covering semiconductor process technology and design intellectual property. In fiscal year 2024, royalty payments totaled approximately €68.4 million (approximately $70.5 million at the weighted average payment-date exchange rate of 1.0307).

3. Cost-Sharing Contributions under the Qualified Cost Sharing Arrangement ("QCSA"). HSI makes cost-sharing payments to HSAG under a qualified cost sharing arrangement pursuant to Treasury Regulations § 1.482-7. In fiscal year 2024, HSI's cost-sharing contribution totaled approximately €224.6 million (approximately $231.3 million). HSI's anticipated benefit share under the QCSA is 22.5%.

Aggregate annual intercompany flows across these three transaction categories totaled approximately €1.28 billion in fiscal year 2024.

The proposed APA term covers five prospective tax years (January 1, 2025 through December 31, 2029), with a rollback request for three prior tax years (January 1, 2022 through December 31, 2024), resulting in a total covered period of eight tax years. The formal APA filing is scheduled for July 7, 2025.

Summary of Sensitivity Findings. All sensitivity tests presented in this memorandum confirm that HSI's fiscal year 2024 operating margin of 4.80% and HSAG's 6.0% royalty rate remain within or very close to the arm's-length range under every reasonable alternative specification tested. The primary benchmarking results — a Comparable Profits Method / Transactional Net Margin Method ("CPM/TNMM") interquartile range ("IQR") of 2.8% to 6.1% (median 4.4%) for the tangible goods transaction, and a Comparable Uncontrolled Transaction ("CUT") IQR of 3.5% to 8.2% (median 5.6%) for the royalty transaction — are robust across sixteen distinct sensitivity tests covering alternative profit level indicators, comparable screening criteria, temporal testing periods, comparability adjustments, CUT specifications, residual profit split corroboration, and CSA benefit share parameters. These results strongly support the positions advanced in the APA application and provide a solid foundation for negotiations with both the IRS APMA and the BZSt competent authority.

II. Background and Context

II.A. Summary of Primary Economic Analyses

The primary economic analyses supporting the APA application were prepared by Meridian Economics GmbH under the direction of Calderwood & Harkness LLP. These analyses comprise three principal components, each addressing one of the covered intercompany transactions.

Benchmarking Study — Tangible Goods (CPM/TNMM Analysis). The primary transfer pricing analysis for the tangible goods transaction applies the Comparable Profits Method under Treasury Regulations § 1.482-5, which is functionally equivalent to the Transactional Net Margin Method ("TNMM") under the OECD Transfer Pricing Guidelines (2022 ed.), Chapter II. HSI was selected as the tested party on the basis that it is the least complex entity in the controlled transaction — operating as a limited-risk distributor of HSAG-manufactured semiconductor products in North America. HSI does not own or develop significant intangible property, does not bear manufacturing risk, and does not perform significant value-adding functions beyond distribution, marketing, and customer support.

Meridian Economics GmbH identified 17 comparable North American semiconductor distributors through a systematic screening process using the Fenmore Services Inc. and Cascade Software databases. Comparables were selected based on Standard Industrial Classification ("SIC") and North American Industry Classification System ("NAICS") codes relevant to semiconductor and electronic component distribution, subject to filters for geographic focus, independence (no controlled transactions constituting a significant portion of revenues), data availability (five continuous fiscal years covering 2020 through 2024), and functional comparability (distribution activities without significant intangible property ownership). The primary profit level indicator ("PLI") is the operating margin on net resales.

Using five-year weighted average operating margins for the 17 comparable companies over fiscal years 2020 through 2024, the benchmarking analysis produced an interquartile range of 2.8% to 6.1%, with a median of 4.4%. HSI's fiscal year 2024 operating margin of 4.80% falls within this IQR, slightly above the median, consistent with arm's-length pricing.

Benchmarking Study — Royalty (CUT Analysis). The primary transfer pricing analysis for the royalty transaction applies the Comparable Uncontrolled Transaction method under Treasury Regulations § 1.482-4(c). Meridian Economics GmbH identified eight comparable uncontrolled semiconductor intellectual property license agreements through a search of the Clearpoint Royalty Database and public filings with the U.S. Securities and Exchange Commission. Each agreement was evaluated for comparability based on the nature of the licensed technology (semiconductor process technology and/or design intellectual property), the scope of the license (territorial, exclusive vs. non-exclusive), the stage of technology development, and the economic terms and conditions of the arrangement.

The eight comparable agreements yielded an interquartile range of royalty rates from 3.5% to 8.2%, with a median of 5.6%. HSAG's contractual royalty rate of 6.0% falls within the IQR and above the median, reflecting the substantial value of HSAG's advanced-node semiconductor process technology (including 7nm and 5nm process nodes) and its extensive design libraries. The CUT result was corroborated by a residual profit split analysis, as discussed in detail in Section VII.D of this memorandum.

Cost Sharing Arrangement Documentation. The QCSA between HSAG and HSI allocates costs of research and development activities that are reasonably anticipated to give rise to intangible property to be exploited by both parties. HSI's anticipated benefit share of 22.5% reflects its North American exploitation rights, computed using the income method under Treasury Regulations § 1.482-7(g)(4). Anticipated benefits were derived from projected revenues attributable to QCSA-developed intangibles in HSI's territory (North America) relative to HSAG's territories (Europe, Asia, and the rest of the world).

The QCSA was established effective January 1, 2018, with a platform contribution transaction ("PCT") buy-in payment of $312.5 million made by HSI on July 15, 2018. The PCT was valued using the income method with an 11.8% discount rate and a 15-year projection period. The PCT valuation was performed by Stonebridge Valuation Partners LLC (Marcus Reinhardt, ASA, CFA, Managing Director, 200 Park Avenue, Suite 1510, New York, New York 10166).

In fiscal year 2024, qualified CSA R&D costs totaled €998,200,000, of which HSI's 22.5% share was €224,600,000 (approximately $231.3 million). Total HSAG R&D expenditure in fiscal year 2024 was €1,420,000,000; the qualified CSA R&D costs represent a subset of total R&D that relates to intangibles reasonably anticipated to benefit both cost-sharing participants.

HSI Fiscal Year 2024 Financials. HSI reported the following financial results for fiscal year 2024:

HSAG Fiscal Year 2024 Consolidated Financials. HSAG reported the following consolidated financial results for fiscal year 2024:

II.B. Regulatory Framework

The sensitivity analyses presented in this memorandum are conducted within the following regulatory framework, which governs the arm's-length standard, transfer pricing methodology, and APA procedures applicable to the covered transactions:

Internal Revenue Code and Treasury Regulations. Section 482 of the Internal Revenue Code authorizes the Secretary of the Treasury to distribute, apportion, or allocate gross income, deductions, credits, or allowances between or among organizations, trades, or businesses owned or controlled by the same interests if necessary to prevent evasion of taxes or to clearly reflect the income of such organizations. The implementing regulations at Treasury Regulations §§ 1.482-1 through 1.482-9 establish the arm's-length standard, the best method rule, comparability standards, and specific methods for pricing controlled transactions involving tangible property, intangible property, services, and cost sharing arrangements. In particular:

• Treasury Regulations § 1.482-1(c) sets forth the best method rule, requiring taxpayers to use the method that provides the most reliable measure of an arm's-length result based on the quality of the data and assumptions applied.

• Treasury Regulations § 1.482-1(d) establishes comparability standards, including requirements for functional analysis, contractual terms, risk analysis, economic conditions, and property or services comparability.

• Treasury Regulations § 1.482-1(e) defines the arm's-length range, including the use of the interquartile range when comparables are imprecise.

• Treasury Regulations § 1.482-5 governs the Comparable Profits Method, specifying the use of profit level indicators derived from uncontrolled comparables to test the arm's-length nature of a controlled transaction. Section 1.482-5(b)(4) identifies operating margin as the standard PLI for distributors.

• Treasury Regulations § 1.482-4(c) governs the Comparable Uncontrolled Transaction method for intangible property transactions, requiring identification of comparable uncontrolled license or sale agreements.

• Treasury Regulations § 1.482-7 governs qualified cost sharing arrangements, including the determination of anticipated benefit shares, platform contribution transactions, and cost-sharing payments.

• Treasury Regulations § 1.6662-6 establishes the documentation requirements for penalty protection under IRC §§ 6662(e) and 6662(h), requiring contemporaneous documentation demonstrating reasonable cause and good-faith effort in the selection and application of transfer pricing methods.

Revenue Procedure 2015-41. Rev. Proc. 2015-41 sets forth the procedures for requesting and obtaining advance pricing agreements between the IRS and foreign competent authorities. The APA application for HSAG and HSI follows the procedures specified therein, including the pre-filing conference, formal application submission, and competent authority negotiation process.

OECD Transfer Pricing Guidelines. The OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (2022 ed.) provide authoritative guidance on the application of the arm's-length principle, including Chapters I through III (general methodology, comparability analysis, and transfer pricing methods), Chapter VI (intangibles), and Chapter VIII (cost contribution arrangements). The OECD Guidelines are relevant both because the United States and Germany are OECD member states and because the BZSt will apply the Guidelines — particularly the TNMM provisions — in evaluating the APA application. The CPM under Treasury Regulations § 1.482-5 and the TNMM under the OECD Guidelines are functionally equivalent methods when applied to the same tested party and PLI.

II.C. Purpose of Sensitivity Testing in APA Context

The sensitivity analyses presented in this memorandum serve three principal purposes in the context of the bilateral APA negotiations.

First, the analyses are designed to anticipate and proactively address the scrutiny that IRS APMA and BZSt economists and examiners are expected to apply to the benchmarking methodology, comparable selection criteria, and PLI choices set forth in the primary economic analyses. In the bilateral APA context, the IRS APMA team will independently evaluate the Meridian Economics GmbH benchmarking studies, and experience indicates that APMA economists routinely test the sensitivity of proposed arm's-length ranges to alternative specifications. By presenting these analyses ourselves, we demonstrate the robustness of the primary results and reduce the risk that an alternative specification proposed by the IRS or BZSt would produce results inconsistent with HSI's or HSAG's actual intercompany pricing.

Second, the sensitivity analyses directly respond to concerns raised during the pre-filing conference held on May 19, 2025 with the IRS APMA team. During that conference, IRS representatives raised specific questions regarding (a) the breadth of the CUT comparable set used in the royalty analysis, expressing concern that certain of the eight comparable license agreements may involve technology that is not sufficiently comparable to HSAG's advanced-node semiconductor process technology; and (b) the adequacy of the documentation supporting the PCT buy-in valuation, including the discount rate and projection period assumptions used in the income-method valuation. The sensitivity analyses in Sections VII and VIII of this memorandum are specifically designed to address these concerns by demonstrating that the primary results hold under both broader and tighter CUT screening criteria and under alternative PCT valuation assumptions.

Third, the sensitivity analyses demonstrate the overall robustness of the proposed arm's-length ranges and transfer pricing outcomes, providing a strong foundation for the negotiation positions that Calderwood & Harkness LLP will advance on behalf of HSAG and HSI in the APA process. By establishing that HSI's actual results fall within the arm's-length range under every reasonable alternative specification, we strengthen the client's position in negotiations with both the IRS APMA and the BZSt competent authority. The analyses also serve penalty protection purposes under IRC §§ 6662(e) and 6662(h) and Treasury Regulations § 1.6662-6, demonstrating that the taxpayer has exercised reasonable cause and good-faith effort in the selection and application of transfer pricing methods.

III. Sensitivity Analysis 1: Alternative Profit Level Indicators for Tangible Goods Transaction

III.A. Overview and Rationale

The primary CPM/TNMM analysis for the tangible goods transaction between HSAG and HSI uses operating margin on net resales ("OM") as the profit level indicator. Operating margin is the standard PLI for limited-risk distributors under Treasury Regulations § 1.482-5(b)(4), and it is the PLI most commonly accepted by both IRS APMA and the BZSt in APA proceedings involving distribution entities. The operating margin measures the ratio of operating income to net revenue and captures the full range of costs — both cost of goods sold and operating expenses — incurred by the tested party in performing its distribution functions.

To test the robustness of the primary analysis, we compute and analyze two alternative PLIs for the same set of 17 comparable companies and for HSI:

1. Berry Ratio. The Berry ratio measures gross profit divided by operating expenses. It is sometimes considered an appropriate PLI for intermediary or distribution entities where the value of the distribution function is more closely related to operating expenses incurred than to the value of goods sold. The Berry ratio is recognized as a PLI under Treasury Regulations § 1.482-5(b)(4)(ii) and is discussed in the OECD Guidelines at ¶¶ 2.106–2.108.

2. Gross Margin. The gross margin measures gross profit as a percentage of net revenue. While not a PLI explicitly enumerated in Treasury Regulations § 1.482-5(b)(4) (which focuses on net profit indicators), the gross margin is relevant to the Resale Price Method under Treasury Regulations § 1.482-3(c) and provides a useful corroborative indicator, particularly for evaluating whether the intercompany transfer price between HSAG and HSI yields an appropriate gross profit for HSI's distribution activities.

The rationale for testing these alternative PLIs is twofold. First, testing multiple PLIs provides a comprehensive assessment of whether HSI's results are consistent with arm's-length outcomes regardless of the particular profit measure selected. Second, it anticipates potential challenges by IRS APMA economists who may prefer one PLI over another, or who may wish to evaluate the consistency of results across multiple indicators.

III.B. Berry Ratio Analysis

We computed Berry ratios for each of the 17 comparable companies over fiscal years 2020 through 2024, using the same financial data sourced from Fenmore Services Inc. and Cascade Software that was used in the primary operating margin analysis. For each comparable company, the Berry ratio was calculated as gross profit divided by total operating expenses (where operating expenses include selling, general, and administrative expenses but exclude cost of goods sold). Five-year weighted average Berry ratios were computed for each comparable using annual revenues as the weighting factor, consistent with the weighting methodology applied in the primary analysis.

The resulting five-year weighted average Berry ratio IQR for the 17 comparable companies is 1.18 to 1.52, with a median of 1.34. The lower quartile Berry ratio of 1.18 indicates that, at the 25th percentile, comparable distributors earn gross profits approximately 18% in excess of their operating expenses. The upper quartile Berry ratio of 1.52 indicates that, at the 75th percentile, comparable distributors earn gross profits approximately 52% in excess of their operating expenses.

HSI's fiscal year 2024 Berry ratio is computed as follows:

Berry Ratio = Gross Profit ÷ Operating Expenses = $740,000,000 ÷ $542,200,000 = 1.365

HSI's Berry ratio of approximately 1.365 falls within the IQR of 1.18 to 1.52 and is slightly above the median of 1.34. This result is consistent with the primary operating margin analysis, which placed HSI's 4.80% OM slightly above the median of 4.4% within the IQR of 2.8% to 6.1%.

Assessment of Reliability. While the Berry ratio provides corroborative support for the primary analysis, it is generally considered a less reliable PLI than operating margin for distribution entities where cost of goods sold represents a dominant portion of total costs. For HSI, cost of goods sold ($3,380,000,000) represents approximately 82% of net revenue, dwarfing operating expenses ($542,200,000, or approximately 13% of net revenue). In such circumstances, the Berry ratio — which excludes COGS from its denominator — may be more sensitive to classification differences in operating expenses across comparables and less reflective of the overall economics of the distribution function. Treasury Regulations § 1.482-5(b)(4)(ii) notes that the Berry ratio is generally most appropriate where the value of the functions performed is not proportionate to the sales revenue generated. For HSI, however, the distribution function is closely tied to the volume and value of semiconductor products sold, making operating margin the more appropriate PLI. Accordingly, the Berry ratio serves as a useful corroborative indicator but does not supplant operating margin as the primary PLI.

III.C. Gross Margin Analysis

We computed gross margins for each of the 17 comparable companies over fiscal years 2020 through 2024. Gross margin is calculated as gross profit divided by net revenue. Five-year weighted average gross margins were computed for each comparable using annual revenues as the weighting factor, consistent with the methodology applied in the primary analysis.

The resulting five-year weighted average gross margin IQR for the 17 comparable companies is 14.2% to 21.8%, with a median of 17.6%. The range reflects the variation in product mix, pricing power, and cost structures among comparable semiconductor distributors. Companies at the lower end of the range tend to operate in higher-volume, lower-margin product segments, while those at the upper end may benefit from value-added services, specialized product niches, or favorable supplier relationships.

HSI's fiscal year 2024 gross margin is computed as follows:

Gross Margin = Gross Profit ÷ Net Revenue = $740,000,000 ÷ $4,120,000,000 = 17.96%

HSI's gross margin of 17.96% falls within the IQR of 14.2% to 21.8% and is essentially at the median of 17.6%. This result is consistent with both the primary operating margin analysis and the Berry ratio analysis, confirming that the intercompany transfer price for semiconductor products from HSAG to HSI yields a gross profit margin for HSI that is squarely within the arm's-length range.

Assessment of Reliability. The gross margin, while informative, may be affected by classification differences in operating expenses versus cost of goods sold across comparable companies. For example, some distributors may classify certain logistics, warehousing, or technical support costs as COGS, while others may classify the same costs as operating expenses. These classification differences affect the gross margin directly but are normalized when operating margin is used as the PLI, because operating margin captures both COGS and operating expenses in a single measure. For this reason, operating margin remains the preferred PLI under Treasury Regulations § 1.482-5(b)(4), and the gross margin analysis serves a corroborative function.

III.D. Conclusions on Alternative PLIs

The alternative PLI analyses confirm that HSI's fiscal year 2024 results are consistent with arm's-length outcomes under all three profit level indicators tested:

Under each PLI, HSI's result falls within the interquartile range and at or above the median. The consistency across three independent profit measures provides strong evidence that the intercompany pricing between HSAG and HSI for tangible goods is at arm's length.

Operating margin remains the best PLI for the tested party given HSI's characterization as a limited-risk distributor. The Berry ratio and gross margin analyses serve as corroborative indicators and demonstrate the robustness of the primary analysis to alternative PLI specifications. For purposes of the APA negotiations, we recommend presenting the operating margin as the primary PLI while referencing the Berry ratio and gross margin results as corroborative support.

IV. Sensitivity Analysis 2: Alternative Comparable Screening Criteria for Tangible Goods

IV.A. Overview of Primary Search Criteria

The primary benchmarking study for the tangible goods transaction identified 17 comparable North American semiconductor distributors through a systematic screening process. The principal search criteria applied in the primary analysis were as follows:

• Geographic Focus: Companies headquartered in North America (United States and Canada), consistent with HSI's operating geography and the requirement under Treasury Regulations § 1.482-1(d)(4) that comparables operate in similar economic conditions and markets.

• Industry Codes: SIC codes 5065 (Electronic Parts and Equipment, Not Elsewhere Classified) and 5045 (Computers and Computer Peripherals and Software), and NAICS codes 423690 (Other Electronic Parts and Equipment Merchant Wholesalers) and 423430 (Computer and Computer Peripheral Equipment and Software Merchant Wholesalers), supplemented by manual review to identify companies specifically engaged in semiconductor distribution.

• Size Screens: Minimum annual revenue threshold of $100 million, to ensure that comparables are of sufficient size to provide reliable financial data and to exclude small entities whose operations and cost structures may not be comparable to HSI's scale.

• Independence Criteria: Exclusion of companies with controlled transactions constituting more than 20% of total revenues, to ensure that comparable company financial results reflect arm's-length outcomes.

• Data Availability: Requirement of five continuous fiscal years of publicly available financial data covering fiscal years 2020 through 2024, to enable computation of multi-year averages and to facilitate temporal sensitivity testing.

• Functional Comparability: Exclusion of companies with significant intangible property ownership, significant R&D activity (R&D expenditure exceeding 3% of revenue), or significant manufacturing operations. Selected comparables perform distribution functions — purchasing, warehousing, logistics, marketing, and customer support — without creating significant proprietary value through intangible development.

Application of these criteria yielded a final comparable set of 17 companies. The primary analysis computed working-capital-adjusted five-year weighted average operating margins for each comparable, producing an IQR of 2.8% to 6.1% with a median of 4.4%.

IV.B. Sensitivity Test 2A — Broader Industry Screen

To test the sensitivity of the comparable set to the breadth of the industry screen, we expanded the SIC and NAICS codes to include electronic component distributors beyond the semiconductor sector. Specifically, we added SIC codes 5063 (Electrical Apparatus and Equipment, Wiring Supplies) and 5049 (Professional Equipment and Supplies, Not Elsewhere Classified), and NAICS codes 423610 (Electrical Apparatus and Equipment, Wiring Supplies, and Related Equipment Merchant Wholesalers) and 423490 (Other Professional Equipment and Supplies Merchant Wholesalers). All other screening criteria — geographic focus, size screens, independence criteria, data availability, and functional comparability requirements — remained unchanged.

The broader industry screen identified nine additional companies that met all non-industry screening criteria, expanding the comparable set from 17 to 26 companies. The additional comparables include distributors of passive electronic components, connectors, power supplies, electromechanical devices, and measurement equipment. While these companies perform distribution functions similar to HSI's, they operate in adjacent product markets with potentially different margin profiles, competitive dynamics, and customer concentration patterns.

The recomputed five-year weighted average operating margin IQR for the expanded 26-company set is 2.5% to 5.8%, with a median of 4.1%. Compared to the primary IQR of 2.8% to 6.1% (median 4.4%), the broader screen results in a modest downward shift of approximately 30 basis points in both the lower and upper quartile boundaries and the median. This shift is consistent with the observation that electronic component distribution outside the semiconductor segment tends to operate at slightly lower margins due to greater commoditization and price competition in certain product categories.

HSI's fiscal year 2024 operating margin of 4.80% falls within the expanded IQR of 2.5% to 5.8%, above the median of 4.1%. The broader screen thus confirms that HSI's result remains within the arm's-length range even when a more inclusive set of comparable distributors is considered.

Assessment. While the broader industry screen provides useful corroborative support, the primary 17-company set focused on semiconductor distributors offers superior functional comparability. The additional nine companies distribute products that differ from semiconductor wafers and dies in terms of technology complexity, customer relationships, and market dynamics. We recommend maintaining the primary 17-company set as the basis for the APA analysis while noting that the broader screen produces consistent results.

IV.C. Sensitivity Test 2B — Tighter Functional Screen

To address the possibility that IRS APMA economists may favor a more restrictive comparable set, we applied more stringent functional comparability criteria to the primary 17-company set. Specifically, we eliminated comparables that exhibit any of the following characteristics:

• Ownership of any proprietary product lines (i.e., branded products manufactured by the comparable or its affiliates, as opposed to products manufactured by third-party suppliers).

• R&D activity of any magnitude (R&D expenditure exceeding 0.5% of revenue, compared to the 3% threshold used in the primary screen).

• Value-added services beyond basic distribution, including design-in support, applications engineering, custom packaging, or testing services that go beyond standard quality inspection.

Application of these tighter criteria eliminated five companies from the primary set, reducing the comparable set from 17 to 12 companies. The excluded companies included two entities with small proprietary product lines (each accounting for less than 8% of revenue), one entity with R&D spending of 1.4% of revenue (related to proprietary software tools for supply chain management), and two entities that provide significant value-added services (including one that operates a custom semiconductor packaging operation and one that offers extensive design-in engineering support).

The recomputed five-year weighted average operating margin IQR for the tighter 12-company set is 2.6% to 5.5%, with a median of 4.1%. The narrower IQR reflects the greater functional homogeneity of the reduced comparable set. The lower quartile boundary decreased modestly (from 2.8% to 2.6%), while the upper quartile boundary decreased more substantially (from 6.1% to 5.5%), reflecting the exclusion of comparables with value-added activities that generate higher margins.

HSI's fiscal year 2024 operating margin of 4.80% remains comfortably within the tighter IQR of 2.6% to 5.5%, above the median of 4.1%. This result is particularly significant for negotiation purposes, as it demonstrates that even under the most restrictive functional screen that IRS APMA might apply, HSI's actual result falls within the arm's-length range.

Relevance to APA Negotiation. In past APA proceedings involving semiconductor distributors, IRS APMA economists have occasionally favored tighter functional screens to ensure that comparables are limited to "pure" distribution entities. The tighter screen analysis presented here anticipates such a preference and demonstrates that the primary results are not dependent on the inclusion of comparables with mixed functional profiles. We recommend presenting this analysis proactively in the APA submission to preempt potential objections to the breadth of the primary comparable set.

IV.D. Sensitivity Test 2C — Alternative Size Filters

The primary analysis applies a minimum annual revenue threshold of $100 million to ensure adequate data quality and scale comparability. To test the sensitivity of the comparable set to this threshold, we re-ran the screening process using a higher minimum revenue threshold of $500 million.

The higher threshold reduced the comparable set from 17 to 9 companies. The eliminated companies ranged in annual revenue from approximately $120 million to $470 million. The remaining nine companies are larger-scale distributors with annual revenues ranging from approximately $550 million to $6.8 billion, more closely matching HSI's revenue scale of $4.12 billion.

The recomputed five-year weighted average operating margin IQR for the 9-company set with the $500 million minimum threshold is 3.0% to 5.9%, with a median of 4.5%. The IQR is modestly narrower than the primary IQR of 2.8% to 6.1% (median 4.4%), with a slight upward shift in the lower quartile (from 2.8% to 3.0%) and a slight downward shift in the upper quartile (from 6.1% to 5.9%). The median increased by 10 basis points (from 4.4% to 4.5%).

HSI's fiscal year 2024 operating margin of 4.80% falls within the revised IQR of 3.0% to 5.9%, above the median of 4.5%.

Discussion of Size Comparability. The size filter analysis raises important questions regarding the relationship between company size and operating margin in the semiconductor distribution sector. In general, larger distributors may benefit from economies of scale in logistics, purchasing, and administration, potentially generating higher margins. However, larger distributors also tend to serve more price-sensitive customers (including original equipment manufacturers with significant bargaining power) and may face greater competitive pressure on margins. These offsetting effects may explain the relatively modest impact of the size filter on the IQR.

In the primary analysis, working capital adjustments were applied to all comparables to normalize for differences in working capital intensity, which may correlate with size. To the extent that size-related comparability differences are driven by working capital effects (e.g., larger companies maintaining lower days sales outstanding or higher inventory turnover), the working capital adjustments mitigate these differences. Additional adjustments for size per se were not applied, consistent with the general practice under Treasury Regulations § 1.482-1(d)(3)(iii), which does not require size adjustments absent demonstrable functional or risk differences correlated with size.

IV.E. Conclusions on Alternative Screening Criteria

The three alternative screening tests confirm that the primary comparable set of 17 companies represents the best balance of functional comparability and statistical reliability. The alternative screens yield IQRs that are broadly consistent with the primary range, as summarized below:

Under all four specifications, HSI's fiscal year 2024 operating margin of 4.80% falls within the interquartile range and above the median. The primary 17-company set provides the largest comparable set consistent with functional comparability, offering the greatest statistical reliability. Narrower sets (the tighter functional screen and the higher size filter) produce IQRs that are consistent with the primary range but are based on fewer observations, potentially reducing statistical reliability. The broader industry screen provides directional support but introduces comparables with lower functional comparability.

For APA negotiation purposes, these results demonstrate that the primary IQR is robust to reasonable alternative specifications and that HSI's actual result is not sensitive to the particular screening criteria applied.

V. Sensitivity Analysis 3: Multi-Year Weighted Average vs. Single-Year Testing

V.A. Rationale for Multi-Year Analysis

The semiconductor industry is characterized by pronounced cyclicality, driven by the capital-intensive nature of wafer fabrication, the long lead times for capacity additions, and the boom-bust dynamics of end-market demand. As documented in the Industry Analysis Memorandum prepared by Meridian Economics GmbH (dated March 14, 2025), the semiconductor industry experienced a significant demand surge in 2021, driven by pandemic-related shifts in consumer electronics, data center, and automotive demand; a period of elevated but normalizing demand in 2022; an inventory correction and demand softening in 2023 (particularly affecting consumer and industrial end-markets); and a recovery beginning in mid-2024, led by artificial intelligence and data-center applications.

These cyclical dynamics cause year-to-year variation in the operating margins of both the tested party and the comparable companies. In recognition of this phenomenon, Treasury Regulations § 1.482-1(f)(2)(iii)(D) and the OECD Transfer Pricing Guidelines at paragraphs 3.75 through 3.79 endorse the use of multiple-year data to improve the reliability of transfer pricing analyses by smoothing the effects of business and product cycles. The OECD Guidelines note that "using data from the comparable and the tested party covering a number of years rather than a single year may improve reliability of the analysis, in particular, but not only, in industries characterized by significant cyclicality."

The primary analysis uses five-year weighted average operating margins for fiscal years 2020 through 2024, consistent with these regulatory and guidance provisions. The five-year period captures a full business cycle in the semiconductor industry, encompassing the pandemic-related demand surge (2020–2021), the period of elevated margins (2022), the inventory correction (2023), and the early recovery (2024).

V.B. Single-Year Results

To test the sensitivity of the benchmarking results to the temporal period, we computed IQRs and median operating margins for the 17 comparable companies for each individual fiscal year from 2020 through 2024. The results are presented below:

Several observations merit discussion:

Year-by-year variation. As expected, the single-year IQRs exhibit significant variation across the five-year period. The widest IQR was observed in FY 2022 (3.8% to 7.8%, a 400-basis-point spread), a year of elevated industry margins driven by strong demand and limited supply additions. The narrowest IQR was observed in FY 2023 (1.6% to 4.8%, a 320-basis-point spread), reflecting the compressive effect of the inventory correction on distributor margins.

Downturn effects. FY 2023 represents a cyclical trough in the semiconductor distribution sector. Both the IQR and the median declined significantly relative to FY 2022, reflecting the impact of excess inventory in the channel, aggressive destocking by customers, and price erosion. HSI's FY 2023 operating margin of 2.95% falls within the FY 2023 IQR of 1.6% to 4.8%, near the median of 3.1%.

Upturn effects. FY 2021 and FY 2022 represent cyclical peaks. The IQRs are wider and shifted upward, reflecting elevated demand and pricing power. HSI's FY 2021 operating margin of 5.61% and FY 2022 operating margin of 6.14% each fall within their respective single-year IQRs.

Critically, HSI's actual operating margin falls within the single-year IQR in all five fiscal years tested. This result provides strong evidence that the intercompany pricing between HSAG and HSI is at arm's length on a year-by-year basis, not merely on a multi-year average basis.

V.C. Three-Year Rolling Average Results

As an intermediate temporal specification, we computed three-year rolling average operating margins for both the 17 comparables and HSI over the three available rolling windows:

HSI's three-year rolling average operating margin falls within the IQR in all three periods. The three-year rolling average provides an intermediate level of smoothing: it dampens year-to-year volatility while remaining responsive to medium-term trend shifts. The IQRs under the three-year rolling average are narrower than the single-year IQRs in the cyclical extreme years (2021–2022 and 2023) but wider than the five-year weighted average IQR.

The three-year rolling average results are consistent with the five-year weighted average results and provide further corroboration of the arm's-length nature of HSI's intercompany pricing. However, the three-year rolling average is less effective than the five-year weighted average at smoothing the full semiconductor cycle, which typically spans four to five years from peak to trough.

V.D. Five-Year Weighted Average vs. Simple Average

The primary analysis uses five-year weighted average operating margins, where each comparable company's annual operating margin is weighted by annual revenue. This weighting approach gives greater influence to years in which the comparable generated higher revenues, which may correlate with more representative operating conditions. As an alternative, we computed five-year simple (arithmetic) average operating margins for each comparable.

The five-year simple average IQR is 2.6% to 5.9%, with a median of 4.3%. Compared to the weighted average IQR of 2.8% to 6.1% (median 4.4%), the simple average IQR is modestly lower at both boundaries and at the median. The difference reflects the fact that revenue-weighted averages give greater influence to upturn years (when revenues are higher), which tend to have higher margins. HSI's five-year simple average operating margin is 4.58%, which falls within the simple average IQR of 2.6% to 5.9%.

The choice between weighted and simple average is a judgment call that depends on the relative importance of proportional representation versus equal weighting across years. Revenue weighting is generally preferred under Treasury Regulations § 1.482-1(f)(2)(iii)(D) because it reflects the economic significance of each year's operations. The sensitivity analysis demonstrates that the choice between weighted and simple average has only a modest effect on the IQR and does not affect the conclusion that HSI's results are within the arm's-length range.

V.E. Conclusions on Temporal Testing

The temporal sensitivity analyses confirm that the five-year weighted average is the most reliable approach for semiconductor industry benchmarking. It smooths cyclical distortions across the full business cycle and produces an IQR that is representative of arm's-length outcomes over a range of economic conditions. Key conclusions include:

• HSI's actual operating margin falls within the arm's-length range in 100% of single-year tests (five out of five years), all three-year rolling average periods, and both the five-year weighted average and five-year simple average specifications.

• The primary five-year weighted average IQR of 2.8% to 6.1% (median 4.4%) is the most representative and reliable measure of the arm's-length range for the tangible goods transaction.

• For purposes of the APA, the results support annual compliance testing using a multi-year average, with a potential single-year backstop to ensure that HSI's annual result falls within a reasonable range in each individual tax year. We recommend proposing a five-year weighted average as the primary testing mechanism, with a single-year IQR test as an alternative safeguard acceptable to both competent authorities.

VI. Sensitivity Analysis 4: Impact of Comparability Adjustments on IQR

VI.A. Overview of Comparability Adjustments Applied

In the primary benchmarking analysis, working capital adjustments were applied to the 17 comparable companies to normalize for differences in working capital intensity between the comparables and HSI. Working capital adjustments are a standard comparability enhancement under Treasury Regulations § 1.482-1(d)(2) and the OECD Transfer Pricing Guidelines at paragraphs 3.47 through 3.54. The adjustments account for the opportunity cost of capital tied up in working capital components — specifically accounts receivable, inventory, and accounts payable — which directly affects the operating profitability of distribution entities.

The adjustment methodology employed in the primary analysis is a balance-sheet-based approach using the risk-free rate of return to compute the imputed cost of holding working capital. For each comparable company, the adjustment computes the difference between the comparable's working capital intensity (measured as net working capital as a percentage of revenue) and HSI's working capital intensity, and applies the risk-free rate to this difference to derive an adjustment to the comparable's operating margin. The formula is:

Adjusted OM = Unadjusted OM + [(HSI Net WC / Revenue) – (Comparable Net WC / Revenue)] × Risk-Free Rate

Where Net WC = Accounts Receivable + Inventory – Accounts Payable.

The primary analysis uses a risk-free rate of 4.25%, based on the average yield on 1-year U.S. Treasury securities during fiscal year 2024.

VI.B. Sensitivity Test 4A — Unadjusted (No Working Capital Adjustments)

To assess the impact of working capital adjustments on the IQR, we recomputed the five-year weighted average operating margins for the 17 comparable companies without applying any working capital adjustments. The unadjusted results reflect the comparables' operating margins as reported in their financial statements, without normalization for working capital differences relative to HSI.

The unadjusted five-year weighted average operating margin IQR is 2.4% to 6.5%, with a median of 4.5%. Compared to the adjusted IQR of 2.8% to 6.1% (median 4.4%), the unadjusted IQR is wider by approximately 80 basis points (40 basis points wider at each boundary). The median is virtually unchanged (4.5% unadjusted vs. 4.4% adjusted).

The widening of the IQR in the unadjusted analysis reflects the fact that comparables with higher working capital intensity (e.g., those carrying larger inventory positions or extending longer payment terms to customers) tend to report lower operating margins (because the cost of carrying working capital is implicitly embedded in their operating expenses), while comparables with lower working capital intensity tend to report higher operating margins. Working capital adjustments normalize for these differences, bringing the comparables' margins closer together and producing a tighter IQR.

HSI's fiscal year 2024 operating margin of 4.80% falls within the unadjusted IQR of 2.4% to 6.5%, above the median of 4.5%. The result is consistent with the adjusted analysis and confirms that HSI's intercompany pricing is at arm's length regardless of whether working capital adjustments are applied.

Quantification of Adjustment Impact. The working capital adjustments shift the lower quartile boundary upward by 40 basis points (from 2.4% to 2.8%) and shift the upper quartile boundary downward by 40 basis points (from 6.5% to 6.1%). The adjustments thus narrow the IQR by approximately 80 basis points in total, improving the precision of the arm's-length range without shifting the central tendency (median) materially. This narrowing is consistent with the purpose of working capital adjustments under Treasury Regulations § 1.482-1(d)(2): to improve comparability by accounting for quantifiable differences between the tested party and the comparables.

VI.C. Sensitivity Test 4B — Alternative Risk-Free Rate Assumption

The primary analysis uses a risk-free rate of 4.25%, based on the average yield on 1-year U.S. Treasury securities during fiscal year 2024. To test the sensitivity of the working capital adjustments to the risk-free rate assumption, we recomputed the adjustments using two alternative rates:

1. SOFR-based rate: The Secured Overnight Financing Rate averaged approximately 5.33% during fiscal year 2024. While SOFR is a secured overnight rate rather than a term risk-free rate, it provides an alternative benchmark that some IRS economists have referenced in transfer pricing analyses.

2. 5-year U.S. Treasury yield: The average yield on 5-year U.S. Treasury securities during fiscal year 2024 was approximately 4.10%. This alternative maturity better matches the multi-year horizon of the benchmarking analysis but may understate the short-term opportunity cost of working capital.

Under the SOFR-based rate of 5.33%, the working capital adjustments are amplified (because a higher rate increases the imputed cost of holding working capital), producing a five-year weighted average operating margin IQR of 2.9% to 5.9% (median 4.4%). Under the 5-year Treasury yield of 4.10%, the adjustments are modestly attenuated, producing an IQR of 2.7% to 6.1% (median 4.4%).

HSI's fiscal year 2024 operating margin of 4.80% falls within both alternative IQRs. The impact of the risk-free rate assumption on the IQR is relatively modest — the boundaries shift by no more than 20 basis points in either direction — reflecting the fact that working capital differences between HSI and the comparables are not extreme. The median is virtually invariant to the risk-free rate assumption.

VI.D. Sensitivity Test 4C — Country Risk / Geographic Adjustments

The primary analysis uses North American comparables to benchmark HSI, a U.S. entity. Because both the tested party and the comparables operate in the same geographic market (North America), geographic or country-risk adjustments are unnecessary and were not applied in the primary analysis. This approach is consistent with Treasury Regulations § 1.482-1(d)(4)(ii), which requires that differences in geographic markets be accounted for only when they materially affect comparability.

However, in the bilateral APA context, the BZSt may independently evaluate the analysis and may consider whether German or European comparables would be appropriate for benchmarking HSAG's functions (to the extent HSAG's functions are separately analyzed) or whether geographic adjustments should be applied to North American comparables used to benchmark HSI. To anticipate this issue, we note the following:

• HSI is the tested party in the CPM/TNMM analysis. HSI operates exclusively in North America, and the 17 comparable companies are North American entities operating in the same or substantially similar economic environments. No geographic adjustment is warranted for the HSI benchmarking analysis.

• If the BZSt were to propose using European semiconductor distributors as an alternative or supplementary comparable set for benchmarking HSI, geographic adjustments for differences in labor costs, tax burden (at the operating level), regulatory environment, and market structure could be relevant. However, such adjustments would be complex and potentially unreliable, and we would resist any suggestion to replace the North American comparable set with European comparables for HSI benchmarking purposes.

• For HSAG's functions (which are not the subject of the tested-party analysis), geographic considerations are addressed through the functional analysis and the allocation of residual profit, rather than through CPM/TNMM benchmarking.

In summary, geographic adjustments are not applicable to the primary analysis and are not expected to be a point of contention in the IRS APMA review. They may arise in the BZSt review, and we are prepared to address them if necessary.

VI.E. Conclusions on Comparability Adjustments

The comparability adjustment sensitivity analyses yield the following conclusions:

1. Working capital adjustments improve comparability and are consistent with best practices under Treasury Regulations § 1.482-1(d)(2) and the OECD Transfer Pricing Guidelines at paragraphs 3.47 through 3.54. The adjustments narrow the IQR by approximately 80 basis points (from a range of 410 basis points unadjusted to 330 basis points adjusted), enhancing the reliability and precision of the arm's-length range.

2. The IQR is robust with and without adjustments. HSI's fiscal year 2024 operating margin of 4.80% falls within the arm's-length range in all tested specifications — with full working capital adjustments, without working capital adjustments, and with alternative risk-free rate assumptions.

3. The risk-free rate assumption has a modest impact on the IQR, with boundary shifts of no more than 20 basis points under alternative rate assumptions. The median is invariant to the risk-free rate choice.

4. Geographic adjustments are not warranted for the primary analysis, given that both the tested party and the comparables operate in North America.

For APA negotiation purposes, we recommend presenting the working-capital-adjusted IQR as the primary range while noting that the unadjusted results provide consistent support. The choice of risk-free rate is unlikely to be a significant point of negotiation, as its impact on the IQR is immaterial.

VII. Sensitivity Analysis 5: Royalty Rate — Alternative CUT Screening and Corroboration

VII.A. Primary CUT Analysis Recap

The primary transfer pricing analysis for the royalty transaction between HSI and HSAG applies the Comparable Uncontrolled Transaction method under Treasury Regulations § 1.482-4(c). Under the license agreement between the parties, HSI pays a royalty to HSAG calculated as 6.0% of HSI's net revenue attributable to products incorporating HSAG's licensed semiconductor process technology and design intellectual property. In fiscal year 2024, royalty payments totaled approximately €68.4 million (approximately $70.5 million at the weighted average payment-date exchange rate of 1.0307).

Meridian Economics GmbH identified eight comparable uncontrolled license agreements for semiconductor intellectual property through a systematic search of the Clearpoint Royalty Database and public filings with the U.S. Securities and Exchange Commission. The eight agreements were selected based on comparability factors including the nature of the licensed technology (semiconductor process and/or design IP), the scope of the license (territorial and exclusivity provisions), the stage of technology development (commercially proven technology), the compensation structure (running royalty as a percentage of sales), and the economic terms and conditions of the arrangements.

The eight comparable agreements yielded an interquartile range of royalty rates from 3.5% to 8.2%, with a median of 5.6%. HSAG's contractual royalty rate of 6.0% falls within the IQR and above the median, consistent with the substantial value of HSAG's advanced-node process technology (including 7nm and 5nm fabrication processes) and its extensive design libraries encompassing more than 4,200 proprietary cell libraries and process design kits.

VII.B. Sensitivity Test 5A — Broader Technology Licensing Screen

To test the sensitivity of the CUT analysis to the breadth of the technology licensing screen, we expanded the search criteria to include license agreements covering adjacent technology sectors beyond semiconductor process and design IP. Specifically, we included license agreements for broader electronics technologies (including printed circuit board design, display technologies, and integrated circuit packaging), photonics (including laser diode and optical sensor technologies), and microelectromechanical systems ("MEMS"). The search was conducted using the Clearpoint Royalty Database and supplemented by review of publicly filed license agreements with the SEC.

The broader screen identified 14 additional comparable license agreements, expanding the total comparable set from 8 to 22 agreements. The additional agreements cover a diverse range of technologies, including LED manufacturing processes, RF filter design IP, MEMS fabrication techniques, advanced packaging technologies (including 2.5D and 3D integration), and optical sensor design IP.

The recomputed IQR for the expanded 22-agreement set is 3.0% to 7.5%, with a median of 5.2%. Compared to the primary IQR of 3.5% to 8.2% (median 5.6%), the broader screen results in a downward shift of the IQR and the median. This shift reflects the inclusion of license agreements for less specialized technologies (e.g., mature electronics and packaging technologies) that command lower royalty rates, diluting the central tendency of the comparable set.

HSAG's contractual royalty rate of 6.0% falls within the expanded IQR of 3.0% to 7.5% and above the median of 5.2%.

Assessment of Comparability. While the broader screen provides directional support for the 6.0% royalty rate, the additional 14 agreements involve technologies that are less directly comparable to HSAG's advanced-node semiconductor process technology. The licensed technologies in the broader set range from relatively commoditized electronics IP to highly specialized photonics and MEMS technologies, introducing greater heterogeneity in the comparable set. As a result, the broader screen is less reliable than the primary 8-agreement set for purposes of establishing the arm's-length range for the specific intangible property at issue. We present this analysis for completeness and directional support but do not recommend relying on the broader set as the primary CUT analysis.

VII.C. Sensitivity Test 5B — Tighter Comparability Criteria

During the May 19, 2025 pre-filing conference, IRS APMA representatives expressed concern regarding the breadth of the primary CUT comparable set, questioning whether certain of the eight comparable license agreements involve technology that is sufficiently comparable to HSAG's advanced-node semiconductor process technology. To address this concern, we applied more restrictive comparability criteria to the primary 8-agreement set, retaining only license agreements that specifically cover process technology and/or design intellectual property for advanced-node semiconductors (defined as 7nm and below).

Application of these tighter criteria reduced the comparable set from 8 to 5 agreements. The three excluded agreements involved licenses for mature semiconductor technology nodes (28nm and above), which, while functionally comparable in many respects, involve technology with lower development costs, broader availability, and lower strategic value than advanced-node process IP.

The recomputed IQR for the tighter 5-agreement set is 4.2% to 8.8%, with a median of 6.4%. Compared to the primary IQR of 3.5% to 8.2% (median 5.6%), the tighter screen results in an upward shift of both the IQR and the median, reflecting the premium royalty rates commanded by advanced-node semiconductor IP. The five remaining agreements involve licenses for FinFET process technology, gate-all-around ("GAA") transistor architecture design IP, and extreme ultraviolet ("EUV") lithography-compatible process flows — technologies that are dir

CALDERWOOD & HARKNESS LLP 599 Lexington Avenue, 38th Floor New York, New York 10022
SUPPLEMENTAL ECONOMIC ANALYSIS — SENSITIVITY TESTING
Sensitivity Analyses Supporting APA Negotiation Positions for Halcyon Semiconductor Inc.
C&H Matter Number: 1032-00006 Billing Code: HSAG-TP-2025
June 27, 2025
PRIVILEGED AND CONFIDENTIAL Attorney Work Product / Prepared in Anticipation of Litigation and APA Proceedings
This memorandum constitutes attorney work product prepared by Calderwood & Harkness LLP in anticipation of Advance Pricing Agreement proceedings and potential litigation. It is protected from disclosure by the attorney work product doctrine and the attorney-client privilege. No portion of this memorandum may be disclosed to any third party, including any taxing authority, without the prior written authorization of Calderwood & Harkness LLP.
TO:
Leonard Cross, Partner, Tax — Calderwood & Harkness LLP Sylvia Hartwell, Partner, Tax — Calderwood & Harkness LLP Leila Nazari, Senior Associate, Tax — Calderwood & Harkness LLP Robert Jansen, Associate, Tax — Calderwood & Harkness LLP
COPY TO:
Dr. Annika Breitner, Managing Director — Meridian Economics GmbH, Leopoldstraße 145, 80804 Munich, Germany Maximilian Voss, Senior Analyst — Meridian Economics GmbH, Leopoldstraße 145, 80804 Munich, Germany
PREPARED BY:
Calderwood & Harkness LLP, in conjunction with Meridian Economics GmbH
Right-click to update Table of Contents
I. Executive Summary
This memorandum presents supplemental sensitivity analyses designed to stress-test and corroborate the primary economic analyses submitted in the bilateral Advance Pricing Agreement ("APA") application for Halcyon Semiconductor AG ("HSAG") and Halcyon Semiconductor Inc. ("HSI"). The APA is sought between the Internal Revenue Service, Advance Pricing and Mutual Agreement Program ("IRS APMA") and the Bundeszentralamt für Steuern ("BZSt") under Article 25 of the Convention Between the United States of America and the Federal Republic of Germany for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and Capital and to Certain Other Taxes (signed June 29, 2006, as amended) (the "US-Germany Treaty"), pursuant to Rev. Proc. 2015-41.
HSAG is an Aktiengesellschaft organized under the laws of the Federal Republic of Germany, with its registered office at Halcyon-Allee 12, 01099 Dresden, Saxony, Germany (Steuernummer 203/118/04527). HSI is a corporation organized under the laws of the State of Delaware (EIN 84-2917630), with its principal place of business at 8700 Parmer Lane, Building 4, Austin, Texas 78729. HSI is a wholly owned subsidiary of HSAG and serves as the principal North American distribution and sales entity for the Halcyon group.
The APA application covers three categories of intercompany transactions between HSAG and HSI:
1. Tangible Goods (Semiconductor Wafer/Die Sales). HSAG manufactures and sells semiconductor wafers and dies to HSI for resale to third-party customers in North America. In fiscal year 2024, intercompany tangible goods sales totaled approximately €987.6 million (approximately $1,076 million at the average 2024 EUR/USD exchange rate of 1.0896).
2. Royalties (IP License). HSI pays royalties to HSAG under a license agreement covering semiconductor process technology and design intellectual property. In fiscal year 2024, royalty payments totaled approximately €68.4 million (approximately $70.5 million at the weighted average payment-date exchange rate of 1.0307).
3. Cost-Sharing Contributions under the Qualified Cost Sharing Arrangement ("QCSA"). HSI makes cost-sharing payments to HSAG under a qualified cost sharing arrangement pursuant to Treasury Regulations § 1.482-7. In fiscal year 2024, HSI's cost-sharing contribution totaled approximately €224.6 million (approximately $231.3 million). HSI's anticipated benefit share under the QCSA is 22.5%.
Aggregate annual intercompany flows across these three transaction categories totaled approximately €1.28 billion in fiscal year 2024.
The proposed APA term covers five prospective tax years (January 1, 2025 through December 31, 2029), with a rollback request for three prior tax years (January 1, 2022 through December 31, 2024), resulting in a total covered period of eight tax years. The formal APA filing is scheduled for July 7, 2025.
Summary of Sensitivity Findings. All sensitivity tests presented in this memorandum confirm that HSI's fiscal year 2024 operating margin of 4.80% and HSAG's 6.0% royalty rate remain within or very close to the arm's-length range under every reasonable alternative specification tested. The primary benchmarking results — a Comparable Profits Method / Transactional Net Margin Method ("CPM/TNMM") interquartile range ("IQR") of 2.8% to 6.1% (median 4.4%) for the tangible goods transaction, and a Comparable Uncontrolled Transaction ("CUT") IQR of 3.5% to 8.2% (median 5.6%) for the royalty transaction — are robust across sixteen distinct sensitivity tests covering alternative profit level indicators, comparable screening criteria, temporal testing periods, comparability adjustments, CUT specifications, residual profit split corroboration, and CSA benefit share parameters. These results strongly support the positions advanced in the APA application and provide a solid foundation for negotiations with both the IRS APMA and the BZSt competent authority.
II. Background and Context
II.A. Summary of Primary Economic Analyses
The primary economic analyses supporting the APA application were prepared by Meridian Economics GmbH under the direction of Calderwood & Harkness LLP. These analyses comprise three principal components, each addressing one of the covered intercompany transactions.
Benchmarking Study — Tangible Goods (CPM/TNMM Analysis). The primary transfer pricing analysis for the tangible goods transaction applies the Comparable Profits Method under Treasury Regulations § 1.482-5, which is functionally equivalent to the Transactional Net Margin Method ("TNMM") under the OECD Transfer Pricing Guidelines (2022 ed.), Chapter II. HSI was selected as the tested party on the basis that it is the least complex entity in the controlled transaction — operating as a limited-risk distributor of HSAG-manufactured semiconductor products in North America. HSI does not own or develop significant intangible property, does not bear manufacturing risk, and does not perform significant value-adding functions beyond distribution, marketing, and customer support.
Meridian Economics GmbH identified 17 comparable North American semiconductor distributors through a systematic screening process using the Fenmore Services Inc. and Cascade Software databases. Comparables were selected based on Standard Industrial Classification ("SIC") and North American Industry Classification System ("NAICS") codes relevant to semiconductor and electronic component distribution, subject to filters for geographic focus, independence (no controlled transactions constituting a significant portion of revenues), data availability (five continuous fiscal years covering 2020 through 2024), and functional comparability (distribution activities without significant intangible property ownership). The primary profit level indicator ("PLI") is the operating margin on net resales.
Using five-year weighted average operating margins for the 17 comparable companies over fiscal years 2020 through 2024, the benchmarking analysis produced an interquartile range of 2.8% to 6.1%, with a median of 4.4%. HSI's fiscal year 2024 operating margin of 4.80% falls within this IQR, slightly above the median, consistent with arm's-length pricing.
Benchmarking Study — Royalty (CUT Analysis). The primary transfer pricing analysis for the royalty transaction applies the Comparable Uncontrolled Transaction method under Treasury Regulations § 1.482-4(c). Meridian Economics GmbH identified eight comparable uncontrolled semiconductor intellectual property license agreements through a search of the Clearpoint Royalty Database and public filings with the U.S. Securities and Exchange Commission. Each agreement was evaluated for comparability based on the nature of the licensed technology (semiconductor process technology and/or design intellectual property), the scope of the license (territorial, exclusive vs. non-exclusive), the stage of technology development, and the economic terms and conditions of the arrangement.
The eight comparable agreements yielded an interquartile range of royalty rates from 3.5% to 8.2%, with a median of 5.6%. HSAG's contractual royalty rate of 6.0% falls within the IQR and above the median, reflecting the substantial value of HSAG's advanced-node semiconductor process technology (including 7nm and 5nm process nodes) and its extensive design libraries. The CUT result was corroborated by a residual profit split analysis, as discussed in detail in Section VII.D of this memorandum.
Cost Sharing Arrangement Documentation. The QCSA between HSAG and HSI allocates costs of research and development activities that are reasonably anticipated to give rise to intangible property to be exploited by both parties. HSI's anticipated benefit share of 22.5% reflects its North American exploitation rights, computed using the income method under Treasury Regulations § 1.482-7(g)(4). Anticipated benefits were derived from projected revenues attributable to QCSA-developed intangibles in HSI's territory (North America) relative to HSAG's territories (Europe, Asia, and the rest of the world).
The QCSA was established effective January 1, 2018, with a platform contribution transaction ("PCT") buy-in payment of $312.5 million made by HSI on July 15, 2018. The PCT was valued using the income method with an 11.8% discount rate and a 15-year projection period. The PCT valuation was performed by Stonebridge Valuation Partners LLC (Marcus Reinhardt, ASA, CFA, Managing Director, 200 Park Avenue, Suite 1510, New York, New York 10166).
In fiscal year 2024, qualified CSA R&D costs totaled €998,200,000, of which HSI's 22.5% share was €224,600,000 (approximately $231.3 million). Total HSAG R&D expenditure in fiscal year 2024 was €1,420,000,000; the qualified CSA R&D costs represent a subset of total R&D that relates to intangibles reasonably anticipated to benefit both cost-sharing participants.
HSI Fiscal Year 2024 Financials. HSI reported the following financial results for fiscal year 2024:
HSAG Fiscal Year 2024 Consolidated Financials. HSAG reported the following consolidated financial results for fiscal year 2024:
II.B. Regulatory Framework
The sensitivity analyses presented in this memorandum are conducted within the following regulatory framework, which governs the arm's-length standard, transfer pricing methodology, and APA procedures applicable to the covered transactions:
Internal Revenue Code and Treasury Regulations. Section 482 of the Internal Revenue Code authorizes the Secretary of the Treasury to distribute, apportion, or allocate gross income, deductions, credits, or allowances between or among organizations, trades, or businesses owned or controlled by the same interests if necessary to prevent evasion of taxes or to clearly reflect the income of such organizations. The implementing regulations at Treasury Regulations §§ 1.482-1 through 1.482-9 establish the arm's-length standard, the best method rule, comparability standards, and specific methods for pricing controlled transactions involving tangible property, intangible property, services, and cost sharing arrangements. In particular:
• Treasury Regulations § 1.482-1(c) sets forth the best method rule, requiring taxpayers to use the method that provides the most reliable measure of an arm's-length result based on the quality of the data and assumptions applied.
• Treasury Regulations § 1.482-1(d) establishes comparability standards, including requirements for functional analysis, contractual terms, risk analysis, economic conditions, and property or services comparability.
• Treasury Regulations § 1.482-1(e) defines the arm's-length range, including the use of the interquartile range when comparables are imprecise.
• Treasury Regulations § 1.482-5 governs the Comparable Profits Method, specifying the use of profit level indicators derived from uncontrolled comparables to test the arm's-length nature of a controlled transaction. Section 1.482-5(b)(4) identifies operating margin as the standard PLI for distributors.
• Treasury Regulations § 1.482-4(c) governs the Comparable Uncontrolled Transaction method for intangible property transactions, requiring identification of comparable uncontrolled license or sale agreements.
• Treasury Regulations § 1.482-7 governs qualified cost sharing arrangements, including the determination of anticipated benefit shares, platform contribution transactions, and cost-sharing payments.
• Treasury Regulations § 1.6662-6 establishes the documentation requirements for penalty protection under IRC §§ 6662(e) and 6662(h), requiring contemporaneous documentation demonstrating reasonable cause and good-faith effort in the selection and application of transfer pricing methods.
Revenue Procedure 2015-41. Rev. Proc. 2015-41 sets forth the procedures for requesting and obtaining advance pricing agreements between the IRS and foreign competent authorities. The APA application for HSAG and HSI follows the procedures specified therein, including the pre-filing conference, formal application submission, and competent authority negotiation process.
OECD Transfer Pricing Guidelines. The OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (2022 ed.) provide authoritative guidance on the application of the arm's-length principle, including Chapters I through III (general methodology, comparability analysis, and transfer pricing methods), Chapter VI (intangibles), and Chapter VIII (cost contribution arrangements). The OECD Guidelines are relevant both because the United States and Germany are OECD member states and because the BZSt will apply the Guidelines — particularly the TNMM provisions — in evaluating the APA application. The CPM under Treasury Regulations § 1.482-5 and the TNMM under the OECD Guidelines are functionally equivalent methods when applied to the same tested party and PLI.
II.C. Purpose of Sensitivity Testing in APA Context
The sensitivity analyses presented in this memorandum serve three principal purposes in the context of the bilateral APA negotiations.
First, the analyses are designed to anticipate and proactively address the scrutiny that IRS APMA and BZSt economists and examiners are expected to apply to the benchmarking methodology, comparable selection criteria, and PLI choices set forth in the primary economic analyses. In the bilateral APA context, the IRS APMA team will independently evaluate the Meridian Economics GmbH benchmarking studies, and experience indicates that APMA economists routinely test the sensitivity of proposed arm's-length ranges to alternative specifications. By presenting these analyses ourselves, we demonstrate the robustness of the primary results and reduce the risk that an alternative specification proposed by the IRS or BZSt would produce results inconsistent with HSI's or HSAG's actual intercompany pricing.
Second, the sensitivity analyses directly respond to concerns raised during the pre-filing conference held on May 19, 2025 with the IRS APMA team. During that conference, IRS representatives raised specific questions regarding (a) the breadth of the CUT comparable set used in the royalty analysis, expressing concern that certain of the eight comparable license agreements may involve technology that is not sufficiently comparable to HSAG's advanced-node semiconductor process technology; and (b) the adequacy of the documentation supporting the PCT buy-in valuation, including the discount rate and projection period assumptions used in the income-method valuation. The sensitivity analyses in Sections VII and VIII of this memorandum are specifically designed to address these concerns by demonstrating that the primary results hold under both broader and tighter CUT screening criteria and under alternative PCT valuation assumptions.
Third, the sensitivity analyses demonstrate the overall robustness of the proposed arm's-length ranges and transfer pricing outcomes, providing a strong foundation for the negotiation positions that Calderwood & Harkness LLP will advance on behalf of HSAG and HSI in the APA process. By establishing that HSI's actual results fall within the arm's-length range under every reasonable alternative specification, we strengthen the client's position in negotiations with both the IRS APMA and the BZSt competent authority. The analyses also serve penalty protection purposes under IRC §§ 6662(e) and 6662(h) and Treasury Regulations § 1.6662-6, demonstrating that the taxpayer has exercised reasonable cause and good-faith effort in the selection and application of transfer pricing methods.
III. Sensitivity Analysis 1: Alternative Profit Level Indicators for Tangible Goods Transaction
III.A. Overview and Rationale
The primary CPM/TNMM analysis for the tangible goods transaction between HSAG and HSI uses operating margin on net resales ("OM") as the profit level indicator. Operating margin is the standard PLI for limited-risk distributors under Treasury Regulations § 1.482-5(b)(4), and it is the PLI most commonly accepted by both IRS APMA and the BZSt in APA proceedings involving distribution entities. The operating margin measures the ratio of operating income to net revenue and captures the full range of costs — both cost of goods sold and operating expenses — incurred by the tested party in performing its distribution functions.
To test the robustness of the primary analysis, we compute and analyze two alternative PLIs for the same set of 17 comparable companies and for HSI:
1. Berry Ratio. The Berry ratio measures gross profit divided by operating expenses. It is sometimes considered an appropriate PLI for intermediary or distribution entities where the value of the distribution function is more closely related to operating expenses incurred than to the value of goods sold. The Berry ratio is recognized as a PLI under Treasury Regulations § 1.482-5(b)(4)(ii) and is discussed in the OECD Guidelines at ¶¶ 2.106–2.108.
2. Gross Margin. The gross margin measures gross profit as a percentage of net revenue. While not a PLI explicitly enumerated in Treasury Regulations § 1.482-5(b)(4) (which focuses on net profit indicators), the gross margin is relevant to the Resale Price Method under Treasury Regulations § 1.482-3(c) and provides a useful corroborative indicator, particularly for evaluating whether the intercompany transfer price between HSAG and HSI yields an appropriate gross profit for HSI's distribution activities.
The rationale for testing these alternative PLIs is twofold. First, testing multiple PLIs provides a comprehensive assessment of whether HSI's results are consistent with arm's-length outcomes regardless of the particular profit measure selected. Second, it anticipates potential challenges by IRS APMA economists who may prefer one PLI over another, or who may wish to evaluate the consistency of results across multiple indicators.
III.B. Berry Ratio Analysis
We computed Berry ratios for each of the 17 comparable companies over fiscal years 2020 through 2024, using the same financial data sourced from Fenmore Services Inc. and Cascade Software that was used in the primary operating margin analysis. For each comparable company, the Berry ratio was calculated as gross profit divided by total operating expenses (where operating expenses include selling, general, and administrative expenses but exclude cost of goods sold). Five-year weighted average Berry ratios were computed for each comparable using annual revenues as the weighting factor, consistent with the weighting methodology applied in the primary analysis.
The resulting five-year weighted average Berry ratio IQR for the 17 comparable companies is 1.18 to 1.52, with a median of 1.34. The lower quartile Berry ratio of 1.18 indicates that, at the 25th percentile, comparable distributors earn gross profits approximately 18% in excess of their operating expenses. The upper quartile Berry ratio of 1.52 indicates that, at the 75th percentile, comparable distributors earn gross profits approximately 52% in excess of their operating expenses.
HSI's fiscal year 2024 Berry ratio is computed as follows:
Berry Ratio = Gross Profit ÷ Operating Expenses = $740,000,000 ÷ $542,200,000 = 1.365
HSI's Berry ratio of approximately 1.365 falls within the IQR of 1.18 to 1.52 and is slightly above the median of 1.34. This result is consistent with the primary operating margin analysis, which placed HSI's 4.80% OM slightly above the median of 4.4% within the IQR of 2.8% to 6.1%.
Assessment of Reliability. While the Berry ratio provides corroborative support for the primary analysis, it is generally considered a less reliable PLI than operating margin for distribution entities where cost of goods sold represents a dominant portion of total costs. For HSI, cost of goods sold ($3,380,000,000) represents approximately 82% of net revenue, dwarfing operating expenses ($542,200,000, or approximately 13% of net revenue). In such circumstances, the Berry ratio — which excludes COGS from its denominator — may be more sensitive to classification differences in operating expenses across comparables and less reflective of the overall economics of the distribution function. Treasury Regulations § 1.482-5(b)(4)(ii) notes that the Berry ratio is generally most appropriate where the value of the functions performed is not proportionate to the sales revenue generated. For HSI, however, the distribution function is closely tied to the volume and value of semiconductor products sold, making operating margin the more appropriate PLI. Accordingly, the Berry ratio serves as a useful corroborative indicator but does not supplant operating margin as the primary PLI.
III.C. Gross Margin Analysis
We computed gross margins for each of the 17 comparable companies over fiscal years 2020 through 2024. Gross margin is calculated as gross profit divided by net revenue. Five-year weighted average gross margins were computed for each comparable using annual revenues as the weighting factor, consistent with the methodology applied in the primary analysis.
The resulting five-year weighted average gross margin IQR for the 17 comparable companies is 14.2% to 21.8%, with a median of 17.6%. The range reflects the variation in product mix, pricing power, and cost structures among comparable semiconductor distributors. Companies at the lower end of the range tend to operate in higher-volume, lower-margin product segments, while those at the upper end may benefit from value-added services, specialized product niches, or favorable supplier relationships.
HSI's fiscal year 2024 gross margin is computed as follows:
Gross Margin = Gross Profit ÷ Net Revenue = $740,000,000 ÷ $4,120,000,000 = 17.96%
HSI's gross margin of 17.96% falls within the IQR of 14.2% to 21.8% and is essentially at the median of 17.6%. This result is consistent with both the primary operating margin analysis and the Berry ratio analysis, confirming that the intercompany transfer price for semiconductor products from HSAG to HSI yields a gross profit margin for HSI that is squarely within the arm's-length range.
Assessment of Reliability. The gross margin, while informative, may be affected by classification differences in operating expenses versus cost of goods sold across comparable companies. For example, some distributors may classify certain logistics, warehousing, or technical support costs as COGS, while others may classify the same costs as operating expenses. These classification differences affect the gross margin directly but are normalized when operating margin is used as the PLI, because operating margin captures both COGS and operating expenses in a single measure. For this reason, operating margin remains the preferred PLI under Treasury Regulations § 1.482-5(b)(4), and the gross margin analysis serves a corroborative function.
III.D. Conclusions on Alternative PLIs
The alternative PLI analyses confirm that HSI's fiscal year 2024 results are consistent with arm's-length outcomes under all three profit level indicators tested:
Under each PLI, HSI's result falls within the interquartile range and at or above the median. The consistency across three independent profit measures provides strong evidence that the intercompany pricing between HSAG and HSI for tangible goods is at arm's length.
Operating margin remains the best PLI for the tested party given HSI's characterization as a limited-risk distributor. The Berry ratio and gross margin analyses serve as corroborative indicators and demonstrate the robustness of the primary analysis to alternative PLI specifications. For purposes of the APA negotiations, we recommend presenting the operating margin as the primary PLI while referencing the Berry ratio and gross margin results as corroborative support.
IV. Sensitivity Analysis 2: Alternative Comparable Screening Criteria for Tangible Goods
IV.A. Overview of Primary Search Criteria
The primary benchmarking study for the tangible goods transaction identified 17 comparable North American semiconductor distributors through a systematic screening process. The principal search criteria applied in the primary analysis were as follows:
• Geographic Focus: Companies headquartered in North America (United States and Canada), consistent with HSI's operating geography and the requirement under Treasury Regulations § 1.482-1(d)(4) that comparables operate in similar economic conditions and markets.
• Industry Codes: SIC codes 5065 (Electronic Parts and Equipment, Not Elsewhere Classified) and 5045 (Computers and Computer Peripherals and Software), and NAICS codes 423690 (Other Electronic Parts and Equipment Merchant Wholesalers) and 423430 (Computer and Computer Peripheral Equipment and Software Merchant Wholesalers), supplemented by manual review to identify companies specifically engaged in semiconductor distribution.
• Size Screens: Minimum annual revenue threshold of $100 million, to ensure that comparables are of sufficient size to provide reliable financial data and to exclude small entities whose operations and cost structures may not be comparable to HSI's scale.
• Independence Criteria: Exclusion of companies with controlled transactions constituting more than 20% of total revenues, to ensure that comparable company financial results reflect arm's-length outcomes.
• Data Availability: Requirement of five continuous fiscal years of publicly available financial data covering fiscal years 2020 through 2024, to enable computation of multi-year averages and to facilitate temporal sensitivity testing.
• Functional Comparability: Exclusion of companies with significant intangible property ownership, significant R&D activity (R&D expenditure exceeding 3% of revenue), or significant manufacturing operations. Selected comparables perform distribution functions — purchasing, warehousing, logistics, marketing, and customer support — without creating significant proprietary value through intangible development.
Application of these criteria yielded a final comparable set of 17 companies. The primary analysis computed working-capital-adjusted five-year weighted average operating margins for each comparable, producing an IQR of 2.8% to 6.1% with a median of 4.4%.
IV.B. Sensitivity Test 2A — Broader Industry Screen
To test the sensitivity of the comparable set to the breadth of the industry screen, we expanded the SIC and NAICS codes to include electronic component distributors beyond the semiconductor sector. Specifically, we added SIC codes 5063 (Electrical Apparatus and Equipment, Wiring Supplies) and 5049 (Professional Equipment and Supplies, Not Elsewhere Classified), and NAICS codes 423610 (Electrical Apparatus and Equipment, Wiring Supplies, and Related Equipment Merchant Wholesalers) and 423490 (Other Professional Equipment and Supplies Merchant Wholesalers). All other screening criteria — geographic focus, size screens, independence criteria, data availability, and functional comparability requirements — remained unchanged.
The broader industry screen identified nine additional companies that met all non-industry screening criteria, expanding the comparable set from 17 to 26 companies. The additional comparables include distributors of passive electronic components, connectors, power supplies, electromechanical devices, and measurement equipment. While these companies perform distribution functions similar to HSI's, they operate in adjacent product markets with potentially different margin profiles, competitive dynamics, and customer concentration patterns.
The recomputed five-year weighted average operating margin IQR for the expanded 26-company set is 2.5% to 5.8%, with a median of 4.1%. Compared to the primary IQR of 2.8% to 6.1% (median 4.4%), the broader screen results in a modest downward shift of approximately 30 basis points in both the lower and upper quartile boundaries and the median. This shift is consistent with the observation that electronic component distribution outside the semiconductor segment tends to operate at slightly lower margins due to greater commoditization and price competition in certain product categories.
HSI's fiscal year 2024 operating margin of 4.80% falls within the expanded IQR of 2.5% to 5.8%, above the median of 4.1%. The broader screen thus confirms that HSI's result remains within the arm's-length range even when a more inclusive set of comparable distributors is considered.
Assessment. While the broader industry screen provides useful corroborative support, the primary 17-company set focused on semiconductor distributors offers superior functional comparability. The additional nine companies distribute products that differ from semiconductor wafers and dies in terms of technology complexity, customer relationships, and market dynamics. We recommend maintaining the primary 17-company set as the basis for the APA analysis while noting that the broader screen produces consistent results.
IV.C. Sensitivity Test 2B — Tighter Functional Screen
To address the possibility that IRS APMA economists may favor a more restrictive comparable set, we applied more stringent functional comparability criteria to the primary 17-company set. Specifically, we eliminated comparables that exhibit any of the following characteristics:
• Ownership of any proprietary product lines (i.e., branded products manufactured by the comparable or its affiliates, as opposed to products manufactured by third-party suppliers).
• R&D activity of any magnitude (R&D expenditure exceeding 0.5% of revenue, compared to the 3% threshold used in the primary screen).
• Value-added services beyond basic distribution, including design-in support, applications engineering, custom packaging, or testing services that go beyond standard quality inspection.
Application of these tighter criteria eliminated five companies from the primary set, reducing the comparable set from 17 to 12 companies. The excluded companies included two entities with small proprietary product lines (each accounting for less than 8% of revenue), one entity with R&D spending of 1.4% of revenue (related to proprietary software tools for supply chain management), and two entities that provide significant value-added services (including one that operates a custom semiconductor packaging operation and one that offers extensive design-in engineering support).
The recomputed five-year weighted average operating margin IQR for the tighter 12-company set is 2.6% to 5.5%, with a median of 4.1%. The narrower IQR reflects the greater functional homogeneity of the reduced comparable set. The lower quartile boundary decreased modestly (from 2.8% to 2.6%), while the upper quartile boundary decreased more substantially (from 6.1% to 5.5%), reflecting the exclusion of comparables with value-added activities that generate higher margins.
HSI's fiscal year 2024 operating margin of 4.80% remains comfortably within the tighter IQR of 2.6% to 5.5%, above the median of 4.1%. This result is particularly significant for negotiation purposes, as it demonstrates that even under the most restrictive functional screen that IRS APMA might apply, HSI's actual result falls within the arm's-length range.
Relevance to APA Negotiation. In past APA proceedings involving semiconductor distributors, IRS APMA economists have occasionally favored tighter functional screens to ensure that comparables are limited to "pure" distribution entities. The tighter screen analysis presented here anticipates such a preference and demonstrates that the primary results are not dependent on the inclusion of comparables with mixed functional profiles. We recommend presenting this analysis proactively in the APA submission to preempt potential objections to the breadth of the primary comparable set.
IV.D. Sensitivity Test 2C — Alternative Size Filters
The primary analysis applies a minimum annual revenue threshold of $100 million to ensure adequate data quality and scale comparability. To test the sensitivity of the comparable set to this threshold, we re-ran the screening process using a higher minimum revenue threshold of $500 million.
The higher threshold reduced the comparable set from 17 to 9 companies. The eliminated companies ranged in annual revenue from approximately $120 million to $470 million. The remaining nine companies are larger-scale distributors with annual revenues ranging from approximately $550 million to $6.8 billion, more closely matching HSI's revenue scale of $4.12 billion.
The recomputed five-year weighted average operating margin IQR for the 9-company set with the $500 million minimum threshold is 3.0% to 5.9%, with a median of 4.5%. The IQR is modestly narrower than the primary IQR of 2.8% to 6.1% (median 4.4%), with a slight upward shift in the lower quartile (from 2.8% to 3.0%) and a slight downward shift in the upper quartile (from 6.1% to 5.9%). The median increased by 10 basis points (from 4.4% to 4.5%).
HSI's fiscal year 2024 operating margin of 4.80% falls within the revised IQR of 3.0% to 5.9%, above the median of 4.5%.
Discussion of Size Comparability. The size filter analysis raises important questions regarding the relationship between company size and operating margin in the semiconductor distribution sector. In general, larger distributors may benefit from economies of scale in logistics, purchasing, and administration, potentially generating higher margins. However, larger distributors also tend to serve more price-sensitive customers (including original equipment manufacturers with significant bargaining power) and may face greater competitive pressure on margins. These offsetting effects may explain the relatively modest impact of the size filter on the IQR.
In the primary analysis, working capital adjustments were applied to all comparables to normalize for differences in working capital intensity, which may correlate with size. To the extent that size-related comparability differences are driven by working capital effects (e.g., larger companies maintaining lower days sales outstanding or higher inventory turnover), the working capital adjustments mitigate these differences. Additional adjustments for size per se were not applied, consistent with the general practice under Treasury Regulations § 1.482-1(d)(3)(iii), which does not require size adjustments absent demonstrable functional or risk differences correlated with size.
IV.E. Conclusions on Alternative Screening Criteria
The three alternative screening tests confirm that the primary comparable set of 17 companies represents the best balance of functional comparability and statistical reliability. The alternative screens yield IQRs that are broadly consistent with the primary range, as summarized below:
Under all four specifications, HSI's fiscal year 2024 operating margin of 4.80% falls within the interquartile range and above the median. The primary 17-company set provides the largest comparable set consistent with functional comparability, offering the greatest statistical reliability. Narrower sets (the tighter functional screen and the higher size filter) produce IQRs that are consistent with the primary range but are based on fewer observations, potentially reducing statistical reliability. The broader industry screen provides directional support but introduces comparables with lower functional comparability.
For APA negotiation purposes, these results demonstrate that the primary IQR is robust to reasonable alternative specifications and that HSI's actual result is not sensitive to the particular screening criteria applied.
V. Sensitivity Analysis 3: Multi-Year Weighted Average vs. Single-Year Testing
V.A. Rationale for Multi-Year Analysis
The semiconductor industry is characterized by pronounced cyclicality, driven by the capital-intensive nature of wafer fabrication, the long lead times for capacity additions, and the boom-bust dynamics of end-market demand. As documented in the Industry Analysis Memorandum prepared by Meridian Economics GmbH (dated March 14, 2025), the semiconductor industry experienced a significant demand surge in 2021, driven by pandemic-related shifts in consumer electronics, data center, and automotive demand; a period of elevated but normalizing demand in 2022; an inventory correction and demand softening in 2023 (particularly affecting consumer and industrial end-markets); and a recovery beginning in mid-2024, led by artificial intelligence and data-center applications.
These cyclical dynamics cause year-to-year variation in the operating margins of both the tested party and the comparable companies. In recognition of this phenomenon, Treasury Regulations § 1.482-1(f)(2)(iii)(D) and the OECD Transfer Pricing Guidelines at paragraphs 3.75 through 3.79 endorse the use of multiple-year data to improve the reliability of transfer pricing analyses by smoothing the effects of business and product cycles. The OECD Guidelines note that "using data from the comparable and the tested party covering a number of years rather than a single year may improve reliability of the analysis, in particular, but not only, in industries characterized by significant cyclicality."
The primary analysis uses five-year weighted average operating margins for fiscal years 2020 through 2024, consistent with these regulatory and guidance provisions. The five-year period captures a full business cycle in the semiconductor industry, encompassing the pandemic-related demand surge (2020–2021), the period of elevated margins (2022), the inventory correction (2023), and the early recovery (2024).
V.B. Single-Year Results
To test the sensitivity of the benchmarking results to the temporal period, we computed IQRs and median operating margins for the 17 comparable companies for each individual fiscal year from 2020 through 2024. The results are presented below:
Several observations merit discussion:
Year-by-year variation. As expected, the single-year IQRs exhibit significant variation across the five-year period. The widest IQR was observed in FY 2022 (3.8% to 7.8%, a 400-basis-point spread), a year of elevated industry margins driven by strong demand and limited supply additions. The narrowest IQR was observed in FY 2023 (1.6% to 4.8%, a 320-basis-point spread), reflecting the compressive effect of the inventory correction on distributor margins.
Downturn effects. FY 2023 represents a cyclical trough in the semiconductor distribution sector. Both the IQR and the median declined significantly relative to FY 2022, reflecting the impact of excess inventory in the channel, aggressive destocking by customers, and price erosion. HSI's FY 2023 operating margin of 2.95% falls within the FY 2023 IQR of 1.6% to 4.8%, near the median of 3.1%.
Upturn effects. FY 2021 and FY 2022 represent cyclical peaks. The IQRs are wider and shifted upward, reflecting elevated demand and pricing power. HSI's FY 2021 operating margin of 5.61% and FY 2022 operating margin of 6.14% each fall within their respective single-year IQRs.
Critically, HSI's actual operating margin falls within the single-year IQR in all five fiscal years tested. This result provides strong evidence that the intercompany pricing between HSAG and HSI is at arm's length on a year-by-year basis, not merely on a multi-year average basis.
V.C. Three-Year Rolling Average Results
As an intermediate temporal specification, we computed three-year rolling average operating margins for both the 17 comparables and HSI over the three available rolling windows:
HSI's three-year rolling average operating margin falls within the IQR in all three periods. The three-year rolling average provides an intermediate level of smoothing: it dampens year-to-year volatility while remaining responsive to medium-term trend shifts. The IQRs under the three-year rolling average are narrower than the single-year IQRs in the cyclical extreme years (2021–2022 and 2023) but wider than the five-year weighted average IQR.
The three-year rolling average results are consistent with the five-year weighted average results and provide further corroboration of the arm's-length nature of HSI's intercompany pricing. However, the three-year rolling average is less effective than the five-year weighted average at smoothing the full semiconductor cycle, which typically spans four to five years from peak to trough.
V.D. Five-Year Weighted Average vs. Simple Average
The primary analysis uses five-year weighted average operating margins, where each comparable company's annual operating margin is weighted by annual revenue. This weighting approach gives greater influence to years in which the comparable generated higher revenues, which may correlate with more representative operating conditions. As an alternative, we computed five-year simple (arithmetic) average operating margins for each comparable.
The five-year simple average IQR is 2.6% to 5.9%, with a median of 4.3%. Compared to the weighted average IQR of 2.8% to 6.1% (median 4.4%), the simple average IQR is modestly lower at both boundaries and at the median. The difference reflects the fact that revenue-weighted averages give greater influence to upturn years (when revenues are higher), which tend to have higher margins. HSI's five-year simple average operating margin is 4.58%, which falls within the simple average IQR of 2.6% to 5.9%.
The choice between weighted and simple average is a judgment call that depends on the relative importance of proportional representation versus equal weighting across years. Revenue weighting is generally preferred under Treasury Regulations § 1.482-1(f)(2)(iii)(D) because it reflects the economic significance of each year's operations. The sensitivity analysis demonstrates that the choice between weighted and simple average has only a modest effect on the IQR and does not affect the conclusion that HSI's results are within the arm's-length range.
V.E. Conclusions on Temporal Testing
The temporal sensitivity analyses confirm that the five-year weighted average is the most reliable approach for semiconductor industry benchmarking. It smooths cyclical distortions across the full business cycle and produces an IQR that is representative of arm's-length outcomes over a range of economic conditions. Key conclusions include:
• HSI's actual operating margin falls within the arm's-length range in 100% of single-year tests (five out of five years), all three-year rolling average periods, and both the five-year weighted average and five-year simple average specifications.
• The primary five-year weighted average IQR of 2.8% to 6.1% (median 4.4%) is the most representative and reliable measure of the arm's-length range for the tangible goods transaction.
• For purposes of the APA, the results support annual compliance testing using a multi-year average, with a potential single-year backstop to ensure that HSI's annual result falls within a reasonable range in each individual tax year. We recommend proposing a five-year weighted average as the primary testing mechanism, with a single-year IQR test as an alternative safeguard acceptable to both competent authorities.
VI. Sensitivity Analysis 4: Impact of Comparability Adjustments on IQR
VI.A. Overview of Comparability Adjustments Applied
In the primary benchmarking analysis, working capital adjustments were applied to the 17 comparable companies to normalize for differences in working capital intensity between the comparables and HSI. Working capital adjustments are a standard comparability enhancement under Treasury Regulations § 1.482-1(d)(2) and the OECD Transfer Pricing Guidelines at paragraphs 3.47 through 3.54. The adjustments account for the opportunity cost of capital tied up in working capital components — specifically accounts receivable, inventory, and accounts payable — which directly affects the operating profitability of distribution entities.
The adjustment methodology employed in the primary analysis is a balance-sheet-based approach using the risk-free rate of return to compute the imputed cost of holding working capital. For each comparable company, the adjustment computes the difference between the comparable's working capital intensity (measured as net working capital as a percentage of revenue) and HSI's working capital intensity, and applies the risk-free rate to this difference to derive an adjustment to the comparable's operating margin. The formula is:
Adjusted OM = Unadjusted OM + [(HSI Net WC / Revenue) – (Comparable Net WC / Revenue)] × Risk-Free Rate
Where Net WC = Accounts Receivable + Inventory – Accounts Payable.
The primary analysis uses a risk-free rate of 4.25%, based on the average yield on 1-year U.S. Treasury securities during fiscal year 2024.
VI.B. Sensitivity Test 4A — Unadjusted (No Working Capital Adjustments)
To assess the impact of working capital adjustments on the IQR, we recomputed the five-year weighted average operating margins for the 17 comparable companies without applying any working capital adjustments. The unadjusted results reflect the comparables' operating margins as reported in their financial statements, without normalization for working capital differences relative to HSI.
The unadjusted five-year weighted average operating margin IQR is 2.4% to 6.5%, with a median of 4.5%. Compared to the adjusted IQR of 2.8% to 6.1% (median 4.4%), the unadjusted IQR is wider by approximately 80 basis points (40 basis points wider at each boundary). The median is virtually unchanged (4.5% unadjusted vs. 4.4% adjusted).
The widening of the IQR in the unadjusted analysis reflects the fact that comparables with higher working capital intensity (e.g., those carrying larger inventory positions or extending longer payment terms to customers) tend to report lower operating margins (because the cost of carrying working capital is implicitly embedded in their operating expenses), while comparables with lower working capital intensity tend to report higher operating margins. Working capital adjustments normalize for these differences, bringing the comparables' margins closer together and producing a tighter IQR.
HSI's fiscal year 2024 operating margin of 4.80% falls within the unadjusted IQR of 2.4% to 6.5%, above the median of 4.5%. The result is consistent with the adjusted analysis and confirms that HSI's intercompany pricing is at arm's length regardless of whether working capital adjustments are applied.
Quantification of Adjustment Impact. The working capital adjustments shift the lower quartile boundary upward by 40 basis points (from 2.4% to 2.8%) and shift the upper quartile boundary downward by 40 basis points (from 6.5% to 6.1%). The adjustments thus narrow the IQR by approximately 80 basis points in total, improving the precision of the arm's-length range without shifting the central tendency (median) materially. This narrowing is consistent with the purpose of working capital adjustments under Treasury Regulations § 1.482-1(d)(2): to improve comparability by accounting for quantifiable differences between the tested party and the comparables.
VI.C. Sensitivity Test 4B — Alternative Risk-Free Rate Assumption
The primary analysis uses a risk-free rate of 4.25%, based on the average yield on 1-year U.S. Treasury securities during fiscal year 2024. To test the sensitivity of the working capital adjustments to the risk-free rate assumption, we recomputed the adjustments using two alternative rates:
1. SOFR-based rate: The Secured Overnight Financing Rate averaged approximately 5.33% during fiscal year 2024. While SOFR is a secured overnight rate rather than a term risk-free rate, it provides an alternative benchmark that some IRS economists have referenced in transfer pricing analyses.
2. 5-year U.S. Treasury yield: The average yield on 5-year U.S. Treasury securities during fiscal year 2024 was approximately 4.10%. This alternative maturity better matches the multi-year horizon of the benchmarking analysis but may understate the short-term opportunity cost of working capital.
Under the SOFR-based rate of 5.33%, the working capital adjustments are amplified (because a higher rate increases the imputed cost of holding working capital), producing a five-year weighted average operating margin IQR of 2.9% to 5.9% (median 4.4%). Under the 5-year Treasury yield of 4.10%, the adjustments are modestly attenuated, producing an IQR of 2.7% to 6.1% (median 4.4%).
HSI's fiscal year 2024 operating margin of 4.80% falls within both alternative IQRs. The impact of the risk-free rate assumption on the IQR is relatively modest — the boundaries shift by no more than 20 basis points in either direction — reflecting the fact that working capital differences between HSI and the comparables are not extreme. The median is virtually invariant to the risk-free rate assumption.
VI.D. Sensitivity Test 4C — Country Risk / Geographic Adjustments
The primary analysis uses North American comparables to benchmark HSI, a U.S. entity. Because both the tested party and the comparables operate in the same geographic market (North America), geographic or country-risk adjustments are unnecessary and were not applied in the primary analysis. This approach is consistent with Treasury Regulations § 1.482-1(d)(4)(ii), which requires that differences in geographic markets be accounted for only when they materially affect comparability.
However, in the bilateral APA context, the BZSt may independently evaluate the analysis and may consider whether German or European comparables would be appropriate for benchmarking HSAG's functions (to the extent HSAG's functions are separately analyzed) or whether geographic adjustments should be applied to North American comparables used to benchmark HSI. To anticipate this issue, we note the following:
• HSI is the tested party in the CPM/TNMM analysis. HSI operates exclusively in North America, and the 17 comparable companies are North American entities operating in the same or substantially similar economic environments. No geographic adjustment is warranted for the HSI benchmarking analysis.
• If the BZSt were to propose using European semiconductor distributors as an alternative or supplementary comparable set for benchmarking HSI, geographic adjustments for differences in labor costs, tax burden (at the operating level), regulatory environment, and market structure could be relevant. However, such adjustments would be complex and potentially unreliable, and we would resist any suggestion to replace the North American comparable set with European comparables for HSI benchmarking purposes.
• For HSAG's functions (which are not the subject of the tested-party analysis), geographic considerations are addressed through the functional analysis and the allocation of residual profit, rather than through CPM/TNMM benchmarking.
In summary, geographic adjustments are not applicable to the primary analysis and are not expected to be a point of contention in the IRS APMA review. They may arise in the BZSt review, and we are prepared to address them if necessary.
VI.E. Conclusions on Comparability Adjustments
The comparability adjustment sensitivity analyses yield the following conclusions:
1. Working capital adjustments improve comparability and are consistent with best practices under Treasury Regulations § 1.482-1(d)(2) and the OECD Transfer Pricing Guidelines at paragraphs 3.47 through 3.54. The adjustments narrow the IQR by approximately 80 basis points (from a range of 410 basis points unadjusted to 330 basis points adjusted), enhancing the reliability and precision of the arm's-length range.
2. The IQR is robust with and without adjustments. HSI's fiscal year 2024 operating margin of 4.80% falls within the arm's-length range in all tested specifications — with full working capital adjustments, without working capital adjustments, and with alternative risk-free rate assumptions.
3. The risk-free rate assumption has a modest impact on the IQR, with boundary shifts of no more than 20 basis points under alternative rate assumptions. The median is invariant to the risk-free rate choice.
4. Geographic adjustments are not warranted for the primary analysis, given that both the tested party and the comparables operate in North America.
For APA negotiation purposes, we recommend presenting the working-capital-adjusted IQR as the primary range while noting that the unadjusted results provide consistent support. The choice of risk-free rate is unlikely to be a significant point of negotiation, as its impact on the IQR is immaterial.
VII. Sensitivity Analysis 5: Royalty Rate — Alternative CUT Screening and Corroboration
VII.A. Primary CUT Analysis Recap
The primary transfer pricing analysis for the royalty transaction between HSI and HSAG applies the Comparable Uncontrolled Transaction method under Treasury Regulations § 1.482-4(c). Under the license agreement between the parties, HSI pays a royalty to HSAG calculated as 6.0% of HSI's net revenue attributable to products incorporating HSAG's licensed semiconductor process technology and design intellectual property. In fiscal year 2024, royalty payments totaled approximately €68.4 million (approximately $70.5 million at the weighted average payment-date exchange rate of 1.0307).
Meridian Economics GmbH identified eight comparable uncontrolled license agreements for semiconductor intellectual property through a systematic search of the Clearpoint Royalty Database and public filings with the U.S. Securities and Exchange Commission. The eight agreements were selected based on comparability factors including the nature of the licensed technology (semiconductor process and/or design IP), the scope of the license (territorial and exclusivity provisions), the stage of technology development (commercially proven technology), the compensation structure (running royalty as a percentage of sales), and the economic terms and conditions of the arrangements.
The eight comparable agreements yielded an interquartile range of royalty rates from 3.5% to 8.2%, with a median of 5.6%. HSAG's contractual royalty rate of 6.0% falls within the IQR and above the median, consistent with the substantial value of HSAG's advanced-node process technology (including 7nm and 5nm fabrication processes) and its extensive design libraries encompassing more than 4,200 proprietary cell libraries and process design kits.
VII.B. Sensitivity Test 5A — Broader Technology Licensing Screen
To test the sensitivity of the CUT analysis to the breadth of the technology licensing screen, we expanded the search criteria to include license agreements covering adjacent technology sectors beyond semiconductor process and design IP. Specifically, we included license agreements for broader electronics technologies (including printed circuit board design, display technologies, and integrated circuit packaging), photonics (including laser diode and optical sensor technologies), and microelectromechanical systems ("MEMS"). The search was conducted using the Clearpoint Royalty Database and supplemented by review of publicly filed license agreements with the SEC.
The broader screen identified 14 additional comparable license agreements, expanding the total comparable set from 8 to 22 agreements. The additional agreements cover a diverse range of technologies, including LED manufacturing processes, RF filter design IP, MEMS fabrication techniques, advanced packaging technologies (including 2.5D and 3D integration), and optical sensor design IP.
The recomputed IQR for the expanded 22-agreement set is 3.0% to 7.5%, with a median of 5.2%. Compared to the primary IQR of 3.5% to 8.2% (median 5.6%), the broader screen results in a downward shift of the IQR and the median. This shift reflects the inclusion of license agreements for less specialized technologies (e.g., mature electronics and packaging technologies) that command lower royalty rates, diluting the central tendency of the comparable set.
HSAG's contractual royalty rate of 6.0% falls within the expanded IQR of 3.0% to 7.5% and above the median of 5.2%.
Assessment of Comparability. While the broader screen provides directional support for the 6.0% royalty rate, the additional 14 agreements involve technologies that are less directly comparable to HSAG's advanced-node semiconductor process technology. The licensed technologies in the broader set range from relatively commoditized electronics IP to highly specialized photonics and MEMS technologies, introducing greater heterogeneity in the comparable set. As a result, the broader screen is less reliable than the primary 8-agreement set for purposes of establishing the arm's-length range for the specific intangible property at issue. We present this analysis for completeness and directional support but do not recommend relying on the broader set as the primary CUT analysis.
VII.C. Sensitivity Test 5B — Tighter Comparability Criteria
During the May 19, 2025 pre-filing conference, IRS APMA representatives expressed concern regarding the breadth of the primary CUT comparable set, questioning whether certain of the eight comparable license agreements involve technology that is sufficiently comparable to HSAG's advanced-node semiconductor process technology. To address this concern, we applied more restrictive comparability criteria to the primary 8-agreement set, retaining only license agreements that specifically cover process technology and/or design intellectual property for advanced-node semiconductors (defined as 7nm and below).
Application of these tighter criteria reduced the comparable set from 8 to 5 agreements. The three excluded agreements involved licenses for mature semiconductor technology nodes (28nm and above), which, while functionally comparable in many respects, involve technology with lower development costs, broader availability, and lower strategic value than advanced-node process IP.
The recomputed IQR for the tighter 5-agreement set is 4.2% to 8.8%, with a median of 6.4%. Compared to the primary IQR of 3.5% to 8.2% (median 5.6%), the tighter screen results in an upward shift of both the IQR and the median, reflecting the premium royalty rates commanded by advanced-node semiconductor IP. The five remaining agreements involve licenses for FinFET process technology, gate-all-around ("GAA") transistor architecture design IP, and extreme ultraviolet ("EUV") lithography-compatible process flows — technologies that are dir

CALDERWOOD & HARKNESS LLP 599 Lexington Avenue, 38th Floor New York, New York 10022

COPY TO:

PREPARED BY:

Record

Parties

Dates

Amounts