Slip Op. 25-149
UNITED STATES COURT OF INTERNATIONAL TRADE
Court No. 24-00079
CS WIND MALAYSIA SDN. BHD. and CS WIND CORPORATION, Plaintiffs, v. UNITED STATES, Defendant, and WIND TOWER TRADE COALITION, Defendant-Intervenor.
Before: M. Miller Baker, Judge
OPINION
[Sustaining Commerce in part and remanding in part.]
Dated: December 5, 2025
Jarrod M. Goldfeder, Kenneth N. Hammer, and Sezi Erdin, Trade Pacific PLLC, Washington, DC, on the briefs for Plaintiffs.
Isabelle Aubrun, Trial Attorney, Commercial Litiga- tion Branch, Civil Division, U.S. Department of Jus- tice, Washington, DC, on the brief for Defendant. Of counsel for Defendant was Brien Stonebreaker, Ct. No. 24-00079 Page 2
Attorney, Office of the Chief Counsel for Trade En- forcement & Compliance, U.S. Department of Com- merce, Washington, DC.
Alan H. Price, Robert E. DeFrancesco III, Laura El- Sabaawi, Derick G. Holt, and Kimberly A. Reynolds, Wiley Rein LLP, Washington, DC, on the brief for De- fendant-Intervenor.
Baker, Judge: A Malaysian producer and its South Korean parent challenge certain aspects of the Depart- ment of Commerce’s final determination in an admin- istrative review of a countervailing duty order on wind towers from the former nation. The court sustains the agency in part and remands in part.
I
A country may bankroll key manufacturers, which allows them in turn to sell products abroad at lower prices and undercut—if not demolish—their overseas competitors. To combat this practice, U.S. trade law provides a remedy. When Commerce determines that a foreign government subsidizes exported goods and the International Trade Commission finds that such imports injure domestic industry, the former agency will impose a countervailing duty “equal to the amount of the net countervailable subsidy.” 19 U.S.C. § 1671(a). To do so, it must find that “(1) [the foreign] government provided a financial contribution (2) to a specific industry and (3) a recipient within the indus- try received a benefit as a result of that contribution.” Ct. No. 24-00079 Page 3
Hyundai Steel Co. v. United States, 753 F. Supp. 3d 1355, 1356 (CIT 2025) (cleaned up) (quoting Fine Fur- niture (Shanghai) Ltd. v. United States, 748 F.3d 1365, 1369 (Fed. Cir. 2014)). This case presents questions about when a countervailable benefit exists and how it is calculated.
II
In 2021, Commerce imposed countervailing duties on utility-scale wind towers from Malaysia. 86 Fed. Reg. 41,950. In the first administrative review of that decree—covering most of 2021—the Department se- lected producer CS Wind Malaysia Sdn. Bhd. as a mandatory respondent. Appx1001.
After receiving certain information, the agency pre- liminarily found that the company “received counter- vailable subsidies during the” period of review. Appx1000. Commerce reaffirmed that conclusion in its final determination and imposed corresponding du- ties. Appx1021–1022.
Invoking jurisdiction conferred by 28 U.S.C. § 1581(c), the Malaysian company and its South Ko- rean parent (collectively CS Wind) filed this suit under 19 U.S.C. § 1516a(a)(2)(B)(i) to challenge its counter- vailing duty rate. The Wind Tower Trade Coalition, a group of domestic producers, intervened to defend Commerce’s decision. The parties have fully briefed CS Wind’s motion for judgment on the agency record, which is ripe for decision. Ct. No. 24-00079 Page 4
In § 1516a(a)(2) actions, “[t]he court shall hold un- lawful any determination, finding, or conclusion found . . . to be unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B)(i). The question is not whether the court would have reached the same deci- sion on the same record. Rather, it is whether the ad- ministrative record as a whole permits Commerce’s conclusion:
Substantial evidence has been defined as more than a mere scintilla, as such relevant evidence as a reasonable mind might accept as adequate to support a conclusion. To determine if substan- tial evidence exists, we review the record as a whole, including evidence that supports as well as evidence that fairly detracts from the sub- stantiality of the evidence.
Nippon Steel Corp. v. United States, 337 F.3d 1373, 1379 (Fed. Cir. 2003) (cleaned up); see also SSIH Equip. S.A. v. U.S. Int’l Trade Comm’n, 718 F.2d 365, 382 (Fed. Cir. 1983) (if Commerce makes a choice be- tween “two fairly conflicting views,” the court may not substitute its judgment even if its view would have been different “had the matter been before it de novo”) (quoting Universal Camera Corp. v. NLRB, 340 U.S. 474, 488 (1951)).
The court also reviews to ensure the agency en- gaged in “reasoned decisionmaking,” meaning its re- sult must be “within the scope” of its authority and Ct. No. 24-00079 Page 5
“the process” it uses to reach that outcome “must be logical and rational.” Michigan v. EPA, 576 U.S. 743, 750 (2015). The agency must “examine the relevant data and articulate a satisfactory explanation . . . in- cluding a rational connection between the facts found and the choice made.” Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983) (cleaned up). But courts will “uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned.” Id.
III
A
One way that foreign governments may provide a countervailable benefit is to forgo collecting customs duties on imports used by favored industries. Com- merce has addressed this practice. See generally 19 C.F.R. § 351.519 (referring to such forgone customs duties as “import charges”). As relevant here, a benefit exists to the extent that the exemption covers imports not consumed in making a beneficiary’s final exported product. Id. § 351.519(a)(1)(ii). “If the Secretary deter- mines” that an “exemption of import charges upon ex- port confers a benefit,” the amount of the benefit is “normally” the value of “import charges that otherwise would have been paid on the inputs not consumed in the production of the exported product . . . .” Id. § 351.519(a)(3)(ii). Ct. No. 24-00079 Page 6
But the regulation prescribes an “exception” to “paragraph (a)(3).” Id. § 351.519(a)(4). The Depart- ment considers the entire amount of the refund—that is, all import charges the company did not have to pay—to confer a benefit unless either of two circum- stances exists. Id. The first is where the foreign gov- ernment uses a system to track what inputs are con- sumed, and in what quantities, in making the exported goods, and the system is reasonable, effective, and based on generally accepted commercial practices in the foreign country. Id. § 351.519(a)(4)(i). The second applies when the foreign government uses no such sys- tem—or has one that is unreasonable or ineffective— but nonetheless does monitor inputs to confirm what is used and in what amounts. Id. § 351.519(a)(4)(ii).
Commerce preliminarily found that CS Wind re- ceived a countervailable benefit in the form of an ex- emption from import charges for inputs not used in making wind towers exported to this country. Appx1011–1014. As relevant here, the Department made two key findings for purposes of 19 C.F.R. § 351.519(a)(4). First, the Malaysian government did not have a “system or procedure to confirm which in- puts are consumed in the production of the exported products and in what amounts.” Appx1013. Second, neither did it conduct “an examination of actual inputs involved” for the same end. Appx1014. Ct. No. 24-00079 Page 7
The company did not dispute those findings. It in- stead argued that because its records showed that it made limited, if any, home-country sales, it could not have received any benefit for countervailing duty pur- poses. Appx1052. In effect, it contended that its own records rendered the lack of monitoring by the Malay- sian government beside the point.
In its final decision, the Department rejected that argument, explaining that under § 351.519(a)(4) an exporter’s records are irrelevant to whether the for- eign government satisfies the regulation. Appx1052. Here, it was undisputed that Malaysian authorities neither “ha[d] an independent mechanism in place to confirm the quantity of raw material inputs” nor “con- ducted an examination of actual inputs involved to confirm which [ones] are consumed in the production of the exported product.” Id. Therefore, Commerce found the exemption’s full value countervailable. Appx1053.
CS Wind now argues that the Department unlaw- fully applied § 351.519(a)(4)(i) without first making the “threshold determination of whether a benefit is conferred under . . . § 351.519(a)(1)(ii).” ECF 49, at 25. The crux of its argument is its characterization of § 351.519(a)(4) as an exception merely to “the method- ology prescribed by paragraph (a)(3) for calculating the amount of [the] benefit.” Id. at 26 (emphasis added). Absent any finding of a benefit, so the argument goes, there’s nothing to calculate. Ct. No. 24-00079 Page 8
The company misapprehends § 351.519(a)(4). In relevant part, it provides that “[n]otwithstanding par- agraph (a)(3) of this section, the Secretary will con- sider the entire amount of an exemption . . . to confer a benefit, unless” the agency determines that either criterion discussed above is satisfied. 19 C.F.R. § 351.519(a)(4). Critically, the superordinating word— “notwithstanding”—applies to all parts of § 351.519(a)(3), not merely—as the company would have it—to “the methodology prescribed by paragraph (a)(3) for calculating the amount of [the] benefit.” ECF 49, at 26 (emphasis in original). As a result, the language in § 351.519(a)(3)(ii) on which CS Wind im- plicitly relies—“[i]f the Secretary determines that the exemption of import charges upon export confers a benefit” (emphasis added)—does not apply when nei- ther of the criteria of § 351.519(a)(4) is satisfied.
Thus, contrary to the company’s argument, it doesn’t matter that according to its records, it received no “benefit” because its receipt of duty exemptions was limited to inputs for export sales. As the government correctly puts it, the regulation “presum[es]” that all such exemptions are a benefit unless the export coun- try government “demonstrate[s] that it has met the criteria found in 19 C.F.R. § 351.519(a)(4).” ECF 47, at 25. As it’s undisputed that the Malaysian govern- ment failed to do that here, the court sustains Com- merce’s determination that the full value of duty ex- emptions that CS Wind received conferred a counter- vailable benefit. Ct. No. 24-00079 Page 9
B
Another way that a foreign government can provide a countervailable benefit to an exporter is to supply “goods or services . . . for less than adequate remuner- ation.” 19 U.S.C. § 1677(5)(E)(iv). A regulation ex- plains that Commerce normally measures the suffi- ciency of payment for these purposes “by comparing the government price to a market-determined price for the good or service resulting from actual transactions in the country in question.” 19 C.F.R. § 351.511(a)(2)(i). The Department and litigants call 1
this procedure a “tier-one” benchmark.
But sometimes there is no usable “market-deter- mined price” for comparison. In such cases, the agency turns to “tier two” and compares the foreign govern- ment’s price “to a world market price where it is rea- sonable to conclude that [the latter] would be available to purchasers in the country in question.” Id. § 351.511(a)(2)(ii). If there is more than one such world market price, Commerce averages them “to the extent practicable, making due allowance for factors affecting comparability.” Id.
If, in turn, no such world market price is available, “tier three” applies, and the Department examines
1 Commerce amended this regulation in 2024. The pre-2024
version applies here. Ct. No. 24-00079 Page 10
“whether the government price is consistent with mar- ket principles.” Id. § 351.511(a)(2)(iii). If the agency finds no such consistency, “it will look to construct an external benchmark.” Canadian Solar Inc. v. United States, 537 F. Supp. 3d 1380, 1389 n.6 (CIT 2021) (cit- ing +DEDü 6LQDL 9H 7LEEL *D]ODU ,VWLKVDO (QGVWULVL $û v. United States, 459 F. Supp. 3d 1341, 1348 n.13 (CIT 2020)).
The Department preliminarily found that CS Wind received a countervailable benefit because the Malay- sian government leased land to the company for less- than-adequate remuneration. Appx1005; Appx1007– 1008. Applying a tier-one benchmark to gauge what a market-based rental rate would have been, the agency looked to a publication (the Ellis report) that listed such rates for industrial land in Malaysia during the relevant period. Appx1005. The agency selected a lease of a manufacturing property in the state of Pe- nang as a proxy. Id.; Appx1008.
CS Wind had proposed that Commerce instead use rental prices reported by the Malaysia Investment De- velopment Authority. The Department rejected those figures because the source did not explain either their provenance or whether they were government or pri- vate transactions. Appx1005.
The company disagreed and continued to urge the agency to use the Authority’s data. Appx5592–5596. It Ct. No. 24-00079 Page 11
noted that this information reflected a range of rates, whereas the Department used only a single transac- tion from the Ellis Report. Appx5592, Appx5595. Moreover, the Authority’s data came from the same Malaysian jurisdiction, Pahang, where CS Wind’s fa- cility is located. The Ellis report information, on the other hand, came from a different state, Penang, where real estate costs are considerably higher. Appx5595. Finally, the company argued in the alter- native that “if the Department determines that the rental pricing data from Penang . . . is comparable to that of industrial facility pricing in Pahang,” it “should average the . . . rental prices on the record in order to derive a benchmark . . . .” Appx5596.
In its final decision, the Department again used the Penang transaction price from the Ellis report. It ex- plained that this choice “was appropriate because the rental land was . . . roughly comparable in size to CS Wind’s . . . , was . . . of a similar nature (i.e., industrial), and also reflected a transaction between private par- ties.” Appx1038.
Even so, the agency “agree[d] with CS Wind that reliance on a second benchmark source would also be appropriate.” Id. Although Commerce had reserva- tions about the Authority’s data as previously ex- pressed, it was “based on multiple transaction prices for the particular locales in question.” Id. The Depart- ment had also previously used that source in other pro- ceedings. Id. The agency therefore found it appropriate Ct. No. 24-00079 Page 12
to average that benchmark with the Ellis report infor- mation, as doing so was “consistent with [its] general practice with multiple usable sources.” Appx1038– 1039. This revised approach reduced the company’s countervailing duty rate based on the land rental.
CS Wind now challenges the Department’s calcula- tion of the adequacy of remuneration for the property lease from the Malaysian government. The company points to the regulation’s tier-one requirement that the agency consider “product similarity . . . and other factors affecting comparability.” 19 C.F.R. § 351.511(a)(2)(i) (emphasis added). It then argues that Commerce ignored record evidence that land rental rates for a factory in Penang (according to Ellis) were about three times higher than in Pahang (as re- ported by the Authority) during the same period. See ECF 49, at 55 (citing Appx5868–5873, Appx10080). The Department “failed to explain why it was reason- able to include the Penang land transaction” despite the pricing disparity. Id. CS Wind also contends that Commerce ignored the size difference between its (larger) lot in Pahang and the parcel cited in the Ellis data. Id. at 56. The company maintains that these fail- ures rendered the agency’s use of the Ellis figures ar- bitrary.
The government does not defend the price compa- rability of the Ellis data. Instead, it asserts that “Com- merce determined that both sources offered usable, Ct. No. 24-00079 Page 13
albeit imperfect, benchmark prices.” ECF 47, at 49. But the Department made no such finding. It simply— and impermissibly—elided CS Wind’s argument that the Ellis rental price in Penang is not comparable to its lease rate in Pahang due to the disparity in real estate costs between the two Malaysian jurisdictions. Cf. Özdemir Boru San. ve Tic. Ltd. Sti. v. United States, 273 F. Supp. 3d 1225, 1252 (CIT 2017) (“Com- merce must consider relevant record evidence in deter- mining the comparability of land parcels it uses in cre- ating a reasonable benchmark that lacks distortive pricing.”) (citing 19 C.F.R. § 351.511(a)(2)(i)). A re- mand is thus necessary for the agency to tackle that question.
Commerce must also address the company’s argument that the size of its property in Pahang is not comparable to the land in the Ellis report transaction. Although the government argues that CS Wind waived this argument by failing to exhaust it, see ECF 47, at 53, recall that the agency did not consider this topic in its preliminary determination. Instead, it only did so in its final decision. Exhaustion therefore does not apply here. See PAO TMK v. United States, Slip Op. 23-150, at 8–9, 2023 WL 6939242, at *3 (CIT 2023) (finding exhaustion doctrine inapplicable when agency changed course in final decision). As neither the government nor the Coalition defends Commerce’s finding that the size of the company’s property is comparable to the land in the Ellis transaction, the Ct. No. 24-00079 Page 14
agency must explain this determination on remand in light of CS Wind’s contrary evidence.
Finally, CS Wind argues that even if the Depart- ment otherwise properly used the Ellis transaction in determining the adequacy of remuneration for the company’s lease from the Malaysian government, the agency’s calculation of “a land benchmark based on a simple average of the single land transaction from Pe- nang with the multiple land transactions [in Pahang] reflected” in the Authority’s data was unreasonable and “distortive.” ECF 49, at 60–61 (emphasis in origi- nal). The government objects, pointing out that “CS Wind suggested averaging the data in the first place.” ECF 47, at 53.
The court agrees with the government. The com- pany’s lead argument before Commerce was that the agency “should use the average industrial land rental prices in Pahang state [from the Authority] to calcu- late the benchmark.” Appx5592 (emphasis added). In the alternative, it argued that if the agency were to find the Ellis and Authority data comparable, it “should average the . . . rental prices on the record in order to derive a benchmark.” Appx5596. Given CS Wind’s primary argument, there were two prices on the record to choose from—the average of the Author- ity data (from multiple transactions) proffered by the company and the Ellis information (from a single transaction). The Department then did what the com- pany asked it to. Ct. No. 24-00079 Page 15
Under the invited error doctrine (as applied in this administrative law context), “a party may not com- plain on appeal of errors that he himself invited or pro- voked or caused the [agency] . . . to commit.” 9C Wright & Miller, Federal Practice and Procedure § 2558 (3d ed. 2025 update) (quoting Harvis v. Roadway Express, Inc., 923 F.2d 59, 60 (6th Cir. 1991)). Because CS Wind solicited the asserted error that it now bemoans, the court sustains the Department’s decision to calculate a land benchmark based on a simple average of the El- lis price—assuming that this transaction was other- wise comparable with the company’s, a question to be reconsidered on remand—with the Authority’s cost in- formation.
C
In addition to providing land, the Malaysian gov- ernment also supplied electricity to CS Wind. In its preliminary decision, the Department found it lacked sufficient information to use either a tier-one or tier- two benchmark under 19 C.F.R. § 351.511(a)(2) to value the company’s electricity purchases, so it resor- ted to tier three. Appx1010; Appx1006–1007. 2 Citing confidentiality concerns, the Malaysian government refused to provide Commerce with information to al- low an examination of whether its pricing was based
2 Recall that tier three requires the Department to look to
“market principles.” 19 C.F.R. § 351.511(a)(2)(iii). Ct. No. 24-00079 Page 16
on market principles. Appx1007. In its preliminary de- termination, the agency resorted to facts otherwise available under 19 U.S.C. §§ 1677e(a)(1) and (a)(2)(A), (C), and (D) 3 and used Singaporean rates. Appx1006– 1007, Appx1010–1011. It found that country’s data ap- propriate because they represented costs for a compa- rable class or type of electricity customer and were market-based. Appx1006.
CS Wind objected to the selection of Singaporean rates as a benchmark and argued that Commerce should have used Turkish data instead. The company said the latter “reflect[ed] contemporaneous electricity pricing for industrial customers for a country at a com- parable level of economic development to Malaysia.” Appx1053. Singapore, on the other hand, “is at such a grossly disparate level of economic development rela- tive to Malaysia that its electricity pricing cannot pos- sibly be comparable to prices available to industrial consumers in Malaysia.” Id. CS Wind asserted that the Department must compare a country’s level of eco- nomic development to that of the country at issue
3 The cited provisions refer to necessary information being
missing from the record ((a)(1)) and to withholding re- quested information ((a)(2)(A)), significantly impeding the proceeding ((a)(2)(C)), and providing information that could not be verified ((a)(2)(D)). For background on facts otherwise available, see Hung Vuong Corp. v. United States, 483 F. Supp. 3d 1321, 1336–38 (CIT 2020). Ct. No. 24-00079 Page 17
when evidence of such is available on the record. Appx1054.
The agency disagreed and continued to use Singa- porean data in its final determination. In addition to the reasons cited in the preliminary decision, it noted that all the cases CS Wind cited involved “the selection of a tier-three benchmark for land” and stated that the company “provided insufficient evidence demonstrat- ing why the level of economic development would be a similarly relevant consideration when examining elec- tricity.” Appx1057–1058 (emphasis added). It observed that the question in a tier-three analysis is how the foreign government sets its prices and whether they adhere to market principles, and it explained that CS Wind made no effort to show that the Malaysian and Singaporean electricity frameworks are not compara- ble or that the latter’s is not market-based. Appx1058. Finally, the agency cited a previous proceeding in which it found geographic proximity relevant in a tier- three analysis, especially when two countries are part of the same utility market. Id. It noted that Malaysia’s electric grid is interconnected with Singapore and Thailand, which it found suggestive of “regional avail- ability of electricity” between those countries. Id.
CS Wind challenges how Commerce applied the tier-three benchmark for electricity. It relies on a 2021 agency decision, Corrosion Inhibitors from China, that it characterizes as requiring the Department to con- Ct. No. 24-00079 Page 18
sider whether an external benchmark source is from a country at “an economically comparable level of devel- opment” if such information is on the record. ECF 49, at 63 (citing Corrosion Inhibitors I&D Memo at 35–36, accompanying 86 Fed. Reg. 7,537). There, the agency stated that it relied on “two important factors in deter- mining whether a country’s land prices were suitable benchmarks”—geographic proximity to the country at issue and comparable level of economic development. Corrosion Inhibitors I&D Memo at 35 (emphasis added).
Commerce addressed this contention and explained that CS Wind did not show why the level of economic development is relevant to electricity pricing. Appx1057–1058. The agency emphasized that con- sistency with market principles is the touchstone in a tier-three analysis and observed that CS Wind did not dispute that Singapore’s electricity rates satisfy that requirement. Appx1058. Moreover, in other proceed- ings the Department had “determined that countries that are a part of the same utility market are appro- priate as tier-three benchmarks.” Id.
Although CS Wind contends that Commerce failed to account for “evidence that detracts from the reason- ableness of its determination,” ECF 49, at 62–63— namely, the asserted “gross economic disparity be- tween Malaysia and Singapore,” id. at 65—the rele- vance of this gap to the reasonableness of choosing a benchmark for electricity prices is not self-evident. Ct. No. 24-00079 Page 19
The Department adequately responded to this concern by explaining that the company did not show why it mattered.
In this court, CS Wind appears to contend that the level of economic development is essential to analyzing price comparability. See ECF 49, at 72–73. But the reg- ulation only requires the Department to consider price comparability for tiers one and two. See 19 C.F.R. § 351.511(a)(2)(i) (tier one, referring to “comparing” prices); id. § 351.511(a)(2)(ii) (tier two, referring to “comparing” government and world-market pricing); cf. id. § 351.511(a)(2)(iii) (tier three, referring to “as- sessing whether the government price is consistent with market principles”); VHHDOVR+DEDü6LQDLYH7LEEL *D]ODU ,VWLKVDO (QGVWULVL $û Y 8QLWHG 6WDWHV, 536 F. Supp. 3d 1333, 1345 (CIT 2021) (noting that tier three does not require consideration of price compara- bility). Here, Commerce used tier three for electricity, so the company’s “economic development” argument fails on its own terms.
As the Coalition points out (see ECF 53, at 36) the Department recently again reiterated that while the level of economic development is not relevant to choos- ing a benchmark for electricity rates, geographical proximity is. See Crystalline Silicon Photovoltaic Cells from Malaysia I&D Memo at 41–43 (Apr. 18, 2025), accompanying 90 Fed. Reg. 17,384. The agency found proximity “appropriate” in “selecting a tier-three Ct. No. 24-00079 Page 20
benchmark for a utility that is transmitted and con- sumed on a grid.” Id. at 43.
CS Wind argues, however, that the interconnection between the two countries’ power grids “does not sup- port that Singaporean electricity is available to . . . consumers in Malaysia or that [they] consider geo- graphic proximity in deciding whether to purchase” from Singapore or Malaysia. ECF 49, at 65–66. But the Department noted that its use of “a facts available tier- three benchmark” meant it “need not consider whether Singaporean prices are available to users in Malay- sia.” Appx1057 n.213 (emphasis in original); see also Appx1010 (noting agency practice of assuming that foreign electricity prices “are normally not available to purchasers in the country in question”) (citing 63 Fed. Reg. 65,348, 65,377). And if the company is serious about “availability,” it cannot seriously argue that Commerce should have used data from far-more-dis- tant Turkey.
In short, Commerce’s selection of Singaporean rates as a tier-three benchmark to value what CS Wind paid for electrical power supplied by the Malay- sian government is supported by substantial evidence. The court therefore sustains it. Ct. No. 24-00079 Page 21
D
As relevant here, a Commerce regulation directs the agency to calculate a countervailable subsidy rate by dividing the amount of the benefit allocated to the period of review (the numerator) by the sales value during that same period of the product(s) to which the agency attributes the subsidy (the denominator). 19 C.F.R. § 351.525(a). The Department normally deter- mines the sales value “on an f.o.b. (port) basis (if the product is exported) or on an f.o.b. (factory) basis (if . . . sold for domestic consumption).” Id. 4 The agency “will normally attribute an export subsidy only to products exported by a firm,” 19 C.F.R. § 351.525(b)(2), and a “domestic subsidy to all products sold by a firm, includ- ing products that are exported,” id. § 351.525(b)(3).
Sometimes, however, the sales value that would normally be the denominator does not match the goods’ value at U.S. import—for example, when the ex- porter has an affiliate that sells in the U.S. at a markup. This “can result in an overcollection of duties, because the mark-up in price already offsets some of
4 FOB, or “free on board,” means that “the invoice price in-
cludes delivery at seller’s expense to that location.” FOB, Black’s Law Dictionary 578 (4th ed. 1979). “Title to goods usually passes from seller to buyer at the FOB location.” Id. (citing U.C.C. § 2-319(1)). Ct. No. 24-00079 Page 22
the benefit afforded by the countervailable subsidy.” Canadian Solar, 537 F. Supp. 3d at 1395.
To prevent such an overcollection, an exporter may apply for an “entered-value adjustment,” which in- creases the calculation’s denominator. At least in pre- vious years, Commerce did this “by calculating the mark-up ratio on U.S. sales and then multiplying that ratio by the original subsidy rate.” Id.
For purposes of this previously used methodology, “Commerce established six criteria that respondents must demonstrate to qualify for an” entered-value ad- justment. Id. The Department stated them as follows:
(1) the price on which the alleged subsidy is based differs from the U.S. invoiced price,
(2) the exporters and the party that invoices the customer are affiliated,
(3) the U.S. invoice establishes the customs value to which the CVD duties are applied,
(4) there is a one-to-one correlation between the invoice that reflects the price on which subsidies are received and the invoice with the mark-up that accompanies the shipment,
(5) the merchandise is shipped directly to the United States, and Ct. No. 24-00079 Page 23
(6) the invoices can be tracked as back-to-back invoices that are identical except for price.
Id. (quoting Jiangsu Zhongji Lamination Materials Co. v. United States, 405 F. Supp. 3d 1317, 1327 (CIT 2019)).
But in more recent years, Commerce appears to have adopted a different methodology for calculating an entered-value adjustment. The Department “seem- ingly now adjusts a subsidy rate by changing the de- nominator of the original subsidy rate from the re- spondent’s total worldwide export sales to the affili- ate’s marked-up total worldwide export sales.” Id. at 1396 (emphasis added) (citing Crystalline Silicon Pho- tovoltaic Cells from China I&D Memo at 72–75 (Aug. 19, 2019), accompanying 84 Fed. Reg. 45,125, and Jiangsu Zhongji, 405 F. Supp. 3d at 1327–28). “This calculation method does not specifically account for a U.S. mark-up and does not require U.S.-specific sales data.” Id.
The agency preliminarily calculated a countervail- ing duty rate using CS Wind Malaysia’s sales tied to domestic and export subsidies as the denominator. Appx1004–1005. The company then applied for an en- tered-value adjustment, asking the Department to use its Korean parent’s U.S. sales pricing instead. Appx1022. Ct. No. 24-00079 Page 24
Commerce declined this request. Without articulat- ing which methodology it uses to calculate the denom- inator, it asserted that the company must prove that it satisfied the six-part test for the U.S. sales–based methodology. Appx1027. It explained that after CS Wind Malaysia asked for the adjustment, the Depart- ment issued a supplemental questionnaire seeking, for each of four transactions, the invoice from that com- pany to its Korean parent, the invoice from the latter company to the U.S. customer, the bills of lading and packing lists, and copies of Customs Form 7501. 5 Appx1027.
For two of the relevant sales, the Malaysian com- pany responded that it could not provide the Customs forms. Id. It provided no alternative documentation showing declared value upon entry. Appx1028. As these forms were “critical to demonstrate that the U.S. invoice establishes the customs value to which CVD duties are applied,” Appx1027, the Department could not determine whether the third criterion had been met for these two sales. Appx1028.
For the other two, the company did provide all the requested documentation. Id. Even so, “there was a discrepancy between the . . . value listed on the
5 Form 7501 is an “entry summary” Customs uses to collect
various information about imported goods to make a record of the importation and to collect the proper duties and taxes. See https://www.cbp.gov/sites/default/files/2025-11/ cbp_form_7501.pdf. Ct. No. 24-00079 Page 25
invoices and the total entered value on the . . . 7501 forms.” Id. “Given these deficiencies,” Commerce “lack[ed] sufficient information to determine what the customs value is for certain sales, much less that the U.S. invoice establishes the customs value to which the [countervailing] duties are applied”—i.e., the third criterion for the adjustment. Id. (internal quotation marks omitted).
Citing Canadian Solar, CS Wind argues that the Department’s rationale for denying an entered-value adjustment—that the company did not satisfy the third criterion of the six-part test—is unreasonable. “Commerce’s current methodology for calculating” such an adjustment—“by replacing the denominator of the subsidy rate with the total worldwide marked-up sales value—does not require it to” look to U.S. sales data. ECF 49, at 44 (emphasis added). As Canadian Solar explained, the agency’s current worldwide-based “methodology does not match up with the information it asserts is needed.” 537 F. Supp. 3d at 1399; see also id. at 1396 (“Despite not needing U.S. sales data for its” calculation, the Department “has not appeared to adjust its evidentiary requirements to reflect this new methodology.”). In short, there is a disconnect between the agency’s six-part test and what seems to be its cur- rent methodology.
The government responds that “Commerce con- tinue to require” that applicants satisfy the agency’s Ct. No. 24-00079 Page 26
six-part test for an entered-value adjustment. ECF 47, at 39. That begs the question, however, whether con- tinued use of this test is reasonable in view of the ap- parently changed methodology.
The court uses the term “apparently” because the Department here neglected to articulate what method- ology it uses to calculate the denominator. A remand is thus necessary for the agency to undertake this basic exercise. And insofar as it explains that it now looks to the affiliate’s total marked-up worldwide sales or some other metric rather than the affiliate’s marked-up U.S. sales, Commerce must explain the rel- evance of the third criterion 6 on which it relied to deny CS Wind’s requested adjustment. Cf. Canadian Solar, 537 F. Supp. 3d at 1399 (remanding for the agency to “clarify its calculation methodology” and “explain [why] the evidence it requires . . . is reasonable given” that methodology). 7 The six criteria used by the De- partment in this case are merely a means to the end of
6 That “the U.S. invoice establishes the customs value to
which the CVD duties are applied.” Appx1026. 7 The rest of the story is that on remand the agency ran up
the white flag and granted the adjustment. See Canadian Solar Inc. v. United States, Slip Op. 22-49, at 13, 2022 WL 1583428, at *5 (CIT May 19, 2022). In so doing, it stated its intent “to re-evaluate its . . . methodology and the circum- stances under which” an adjustment might be granted. Id. Today’s remand offers an opportunity for that promised reevaluation. Ct. No. 24-00079 Page 27
applying a particular methodology, not ends in them- selves that exist in a vacuum.
* * *
The court grants CS Wind’s motion for judgment on the agency record (ECF 49) in part, sustains Com- merce’s decision in part, and remands for further pro- ceedings consistent with this opinion.
Dated: December 5, 2025 /s/ M. Miller Baker New York, NY Judge