Canadian Solar Inc. et al v. United States

Slip Op. 21-114
UNITED STATES COURT OF INTERNATIONAL TRADE
CANADIAN SOLAR INC., ET AL.,
Plaintiffs,
CHANGZHOU TRINA SOLAR ENERGY
CO., LTD. ET AL.,
Consolidated Plaintiffs,
JINKO SOLAR CO., LTD. ET AL.,
Intervenor Plaintiffs,
v.
UNITED STATES,
Defendant.
Before: Jane A. Restani, Judge
Consol. Court No. 19-00178
PUBLIC VERSION
OPINION AND ORDER
Dated: September 3, 2021
[Commerce’s Final Results in the Fifth Administrative Review of Commerce’s Countervailing
Duty Order on crystalline silicon photovoltaic cells from the People’s Republic of China are
partially sustained and partially remanded for reconsideration consistent with this opinion.]
Bryan P. Cenko and Sarah Wyss, Mowry & Grimson, PLLC, of Washington D.C. argued
for Plaintiffs. With them on the brief was Jeffrey S. Grimson.
Robert G. Gosselink, Jonathan M. Freed, and Kenneth N. Hammer, Trade Pacific, PLLC,
of Washington, D.C., for Consolidated Plaintiffs.
Jordan C. Kahn, Grunfeld Desiderio Lebowitz Silverman & Klestadt, LLP, of Washington,
D.C., argued for Intervenor Plaintiffs Jinko Solar Co., Ltd., Jinko Solar Import & Export Co., Ltd.,
Zhejang Jinko Solar Co., Ltd. On the brief were Andrew T. Schutz and Kavita Mohan.
Richard L.A. Weiner, Rajib Pal, Shawn M. Higgins, and Justin R. Becker, Sidley Austin,
LLP, of Washington, D.C., for Intervenor Plaintiffs Yingli Green Energy Holding Co., Ltd.,
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 1 of 33

Consol. Court No. 19-00178 	Page 2
Public Version
Baoding Tianwei Yingli New Energy Resources Co., Ltd., Tianjin Yingli New Energy Resources
Co., Ltd., Hengshui Yingli New Energy Resources Co., Ltd., Lixian Yingli New Energy Resources
Co., Ltd., Baoding Jiasheng Photovoltaic Technology Co., Ltd., Beijing Tianneng Yingli New
Energy Resources Co., Ltd., Hainan Yingli New Energy Resources Co., Ltd., Shenzhen Yingli
New Energy Resources Co., Ltd., Yingli Green Energy International Trading Co., Ltd., Yingli
Green Energy Americas, Inc., Yingli Energy (China) Co., Ltd.
Craig A. Lewis, Hogan Lovells U.S. LLP, of Washington D.C., for Intervenor Plaintiff
Shanghai BYD Co., Ltd.
Justin R. Miller, International Trade Field Office, U.S. Department of Justice, of New
York, NY argued for the Defendant. With him on the brief was Tara K. Hogan, Assistant Director.
Of counsel on the brief was Paul K. Keith, Office of Chief Counsel for Trade Enforcement and
Compliance, U.S. Department of Commerce, of Washington, D.C.
Restani, Judge: This action is a challenge to the final determination made by the United
States Department of Commerce (“Commerce”) in the Fifth Administrative Review of the
countervailing duty order on crystalline silicon photovoltaic cells, whether or not assembled into
modules from the People’s Republic of China (“PRC”) covering the period from January 1,
2016, to December 31, 2016.
Plaintiffs Canadian Solar Inc., Canadian Solar International, Ltd., Canadian Solar
Manufacturing (Luoyang) Inc., Canadian Solar Manufacturing (Changshu) Inc., Canadian Solar
(USA) Inc., CSI Cells Co., Ltd., CSI Solar Power (China) Inc., CSI Solar Power Group Co. Ltd.,
CSI Solartronics (Changshu) Co., Ltd., CSI Solar Technologies Inc., CSI New Energy Holding
Co., Ltd., CSI-GCL Solar Manufacturing (YanCheng) Co., Ltd., Changshu Tegu New Materials
Technology Co., Ltd., Changshu Tlian Co., Ltd., Suzhou Sanysolar Materials Technology Co.,
Ltd. (collectively, “Canadian Solar”), Consolidated Plaintiffs Changzhou Trina Solar Energy
Co., Ltd., Trina Solar (Changzhou) Science & Technology Co., Ltd., Yancheng Trina Solar
Energy Technology Co., Ltd., Changzhou Trina Solar Yabang Energy Co., Ltd. (collectively,
“Changzhou”), Intervenor Plaintiffs Jinko Solar Co., Ltd., Jinko Solar Import & Export Co., Ltd.,
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 2 of 33

Consol. Court No. 19-00178 	Page 3
Public Version
Zhejang Jinko Solar Co., Ltd. (collectively, “Jinko”), Intervenor Plaintiffs Yingli Green Energy
Holding Co., Ltd., Baoding Tianwei Yingli New Energy Resources Co., Ltd., Tianjin Yingli
New Energy Resources Co., Ltd., Hengshui Yingli New Energy Resources Co., Ltd., Lixian
Yingli New Energy Resources Co., Ltd., Baoding Jiasheng Photovoltaic Technology Co., Ltd.,
Beijing Tianneng Yingli New Energy Resources Co., Ltd., Hainan Yingli New Energy
Resources Co., Ltd., Shenzhen Yingli New Energy Resources Co., Ltd., Yingli Green Energy
International Trading Co., Ltd., Yingli Green Energy Americas, Inc., Yingli Energy (China) Co.,
Ltd. (collectively, “Yingli”) and Intervenor Plaintiff Shanghai BYD Co., Ltd. request that the
court hold aspects of Commerce’s final determination unsupported by substantial evidence or
otherwise not in accordance with law.
The United States (“Government”) asks that the court grant remand for some aspects of
Commerce’s final determination and sustain other parts of Commerce’s Amended Final Results
of its Fifth Administrative Review. 1
BACKGROUND
Commerce published a countervailing duty order on crystalline silicon photovoltaic cells,
whether or not assembled into modules (“solar cells”) from the PRC on December 7, 2012. See
Crystalline Silicon Photovoltaic Cells, Whether or Not Assembled Into Modules, From the
1 Consolidated Plaintiffs, Changzhou, and Intervenor Plaintiffs Jinko, Yingli and Shanghai BYD
have adopted the arguments raised in Plaintiffs Canadian Solar’s brief, Mem. of P. & A. in Supp.
of Rule 56.2 Mot. for J. Upon the Agency R. of Pls. Canadian Solar, ECF No. 70 (confidential),
ECF No. 71 (public) (August 31, 2020) (“Canadian Solar Br.”). See Mem. in Supp. of Mot. for J.
Upon the Agency R. of Consolidated Pls. Changzhou Trina Solar Energy Co., Ltd. et al. at 8,
ECF No. 73-2 (Aug. 31, 2020); Mem. of Law in Supp. of Pl.-Intervenors’ Mot. for J. on the
Agency R. at 2, ECF No. 72 (Aug. 31, 2020) (“Jinko Br.”); Mot. of Pl.-Intervenors Yingli Green
Energy Holding Co., Ltd. et al. for J. on the Agency R. at 2, ECF No. 66 (Aug. 31, 2020);.Mem.
in Supp. of Rule 56.2 Mot. for J. Upon Agency R. by Pl.-Intervenor Shanghai BYD Co., Ltd.,
ECF No. 67 (Aug. 31, 2020).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 3 of 33

Consol. Court No. 19-00178 	Page 4
Public Version
People’s Republic of China: Countervailing Duty Order, 77 Fed. Reg. 73,017 (Dep’t Commerce
Dec. 7, 2012). In February, 2018, Commerce began its Fifth Administrative Review of this
countervailing duty order, covering the period from January 1, 2016, to December 31, 2016.
Initiation of Antidumping and Countervailing Duty Administrative Reviews, 83 Fed. Reg. 8058
(Dep’t Commerce Feb. 23, 2018). On April 17, 2018, the U.S. International Trade Administration
selected Canadian Solar, and Jinko as mandatory respondents (“Mandatory Respondents”) in this
review. See Respondent Selection Memorandum at 1–2, P.R. 38 (Apr. 17, 2018). 2
Commerce published its preliminary results on February 20, 2019, see Crystalline Silicon
Photovoltaic Cells, Whether or Not Assembled Into Modules, From the People’s Republic of
China: Preliminary Results of Countervailing Duty Administrative Review and Intent to Rescind
the Review, in Part; 2016, 84 Fed. Reg. 5051 (Dep’t Commerce Feb. 20, 2019), along with the
accompanying Preliminary Issues and Decision Memorandum, Decision Memorandum for the
Preliminary Results of the Countervailing Duty Administrative Review, Crystalline Silicon
Photovoltaic Cells, Whether or Not Assembled Into Modules, from the People’s Republic of
China, C-570-980, POR: 01/01/2016-12/31/2016 (Dep’t Commerce Feb. 12, 2019) (“PDM”).
Commerce published its final determination on August 28, 2019. See Crystalline Silicon
Photovoltaic Cells, Whether or Not Assembled Into Modules, From the People’s Republic of
China: Final Results of Countervailing Duty Administrative Review and Recession of Review in
2 Commerce thereafter issued an initial questionnaire to Canadian Solar, Jinko, and the
Government of China (the “GOC”); Canadian Solar and the GOC submitted responses on June
19, 2018, Canadian Solar Section III Questionnaire Response, P.R. 68–72, C.R. 20–60 (June 19,
2018); GOC Initial CVD Questionnaire Response, P.R. 73–76, C.R. 61–70 (June 19, 2018), and
Jinko filed its response on June 20, 2018, Jinko Section III Questionnaire Response, P.R. 77–90,
C.R. 71–166 (June 20, 2018). Commerce also issued a supplemental questionnaire to the GOC
and the GOC submitted a response on December 19, 2018. GOC Supplemental CVD
Questionnaire Response, P.R. 156 (Dec. 19, 2018).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 4 of 33

Consol. Court No. 19-00178 	Page 5
Public Version
Part; 2016, 84 Fed. Reg. 45,125 (Dep’t Commerce Aug. 28, 2019) (“Final Results”), as amended
by Crystalline Silicon Photovoltaic Cells, Whether or Not Assembled Into Modules, From the
People’s Republic of China: Amended Final Results of Countervailing Duty Administrative
Review; 2016, 84 Fed. Reg. 68,102 (Dep’t Commerce Dec. 13, 2019) (“Amended Final Results”);
see also Decision Memorandum for Final Results of Countervailing Duty Administrative Review:
Crystalline Silicon Photovoltaic Cells, Whether or Not Assembled Into Modules, from the
People’s Republic of China, C-570-980, POR: 01/01/2016-12/31/2016 (Dep’t Commerce Aug. 19,
2019) (“I&D Memo”). Commerce calculated a subsidy rate of 9.7% ad valorem for Canadian Solar
and of 12.7% ad valorem for Jinko. Amended Final Results at 68,103.
JURISDICTION & STANDARD OF REVIEW
The court has jurisdiction pursuant to 28 U.S.C. § 1581(c) (2021) and 19 U.S.C.
§ 1516a(a)(2)(B)(i) (2021). The court will uphold Commerce’s determinations in a countervailing
duty proceeding unless they are “unsupported by substantial evidence on the record, or otherwise
not in accordance with law[.]” Id. § 1516a(b)(1)(B)(i).
DISCUSSION
I. 	Commerce’s Unopposed Remand Requests on Three Issues
Commerce requests a remand on three of the issues before the court that are substantially
similar to the issues presented in the Third Administrative Review of the order on the merchandise
at issue: 1) reconsidering the benchmark for aluminum extrusions, see Def.’s Resp. in Opp’n. to
Pls.’ Mots. for J. Upon the Agency R. at 19–21, ECF No. 79 (confidential), ECF No. 80 (public)
(December 4, 2020) (“Gov. Br.”), 2) choosing the benchmark for solar grade polysilicon, see id.
at 22–24, and 3) its application of adverse facts available in its specificity finding for the provision
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 5 of 33

Consol. Court No. 19-00178 	Page 6
Public Version
of electricity for less than adequate renumeration (“LTAR”), see id. at 24–26. Plaintiffs support
the Government’s requests for remand.3
The court concludes that in view of its opinions covering the Third Administrative Review
of certain photovoltaic cells from the PRC, see Changzhou Trina Solar Energy Co., Ltd. v. United
States, 42 CIT __, __, 352 F. Supp. 3d 1316, 1332–33, 1336–37, 1341–42 (2018) (“Changzhou
Remand I”); Changzhou Trina Solar Energy Co., Ltd. v. United States, Slip Op. 19-137, 2019 WL
5856438, at *9–10 (CIT Nov. 8, 2019) (“Changzhou Remand II”); Changzhou Trina Solar Energy
Co., Ltd. v. United States, 45 CIT __, __, 466 F. Supp. 3d 1287, 1293–95, 1299–1303 (2020), and
the Fourth Administrative Review, see Canadian Solar Inc. v. United States, Slip Op. 20-23, 2020
WL 898557, at *3 (CIT Feb. 25, 2020) (“Canadian Solar I”); Canadian Solar Inc. v. United States,
Slip Op. 20-149, 2020 WL 6129754, at *4–5 (CIT Oct. 19, 2020) (“Canadian Solar II”), remand
here is appropriate. The administrative records of the Third, Fourth, and Fifth Administrative
Reviews and the Government’s legal rationales are similar and thus the determination at hand
suffers from essentially the same deficiencies that the court noted in these prior opinions. The court
remands with instructions for Commerce to consult these prior opinions and reevaluate its
decisions on these three issues accordingly.
II. 	LTAR Specificity Finding for Aluminum Extrusions
Commerce may assess a countervailing duty if, after investigating a subsidy, Commerce
finds that “an authority 1) provides a financial contribution to a person, 2) a benefit is thereby
3 See Reply Br. in Supp. of Rule 56.2 Mot. for J. on the Agency R. of Pls. Canadian Solar at 4–8,
ECF No. 85 (public), (Jan. 26, 2021), ECF No. 90 (confidential) (Feb. 1, 2021) (“Canadian Solar
Reply Br.”); Reply Br. of Consolidated Pls. Changzhou at 3, ECF No. 86, (Jan. 26, 2021); Pl.-
Intervenors’ Reply Br. at 7, ECF No. 87 (Jan. 26, 2021); Reply of Pl.-Intervenor Shanghai BYD
Co., Ltd. in Supp. of Their Rule 56.2 Mot. for J. on the Agency R. at 4, ECF No. 88 (Jan. 26,
2021).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 6 of 33

Consol. Court No. 19-00178 	Page 7
Public Version
conferred, and 3) the subsidy is specific.” Bethlehem Steel Corp. v. United States, 26 CIT 1003,
1009, 223 F. Supp. 2d 1372, 1378, (2002) (citing 19 U.S.C. § 1677(5)(A)–(B)); see also
Changzhou Remand I, 42 CIT at __, 352 F. Supp. 3d at 1329–30 (reciting same). No party
challenges Commerce's determinations regarding financial contribution or benefit conferred.
Canadian Solar contends that Commerce’s specificity determination concerning the provision of
aluminum extrusions for LTAR was unsupported by substantial evidence and unlawful, because
aluminum extrusions were provided widely at LTAR. Canadian Solar Br. at 14–18. Commerce
argues that its specificity finding for the provision of aluminum extrusions for LTAR is supported
by substantial evidence and in accordance with law, because Commerce determined that the
number and nature of users of aluminum extrusions is limited and does not involve the bulk of the
Chinese economy. Gov. Br. at 16–19.
The GOC’s provision of aluminum extrusions for LTAR is a domestic subsidy subject to
Section 771(5A)(D) of the Tariff Act of 1930. See 19 U.S.C. § 1667(5A)(D) (2020); see also I&D
Memo at 45–46. A domestic subsidy is specific, and therefore countervailable, when “[t]he actual
recipients of the subsidy, whether considered on an enterprise or industry basis, are limited in
number.” 19 U.S.C. § 1677(5A)(D)(iii)(I). “In determining whether a subsidy is provided to a
group of enterprises or industries, Commerce is not required to ‘determine whether there are shared
characteristics among the enterprises or industries’ that receive or are eligible for a subsidy: variety
amongst the industries receiving a given subsidy is not the test for specificity.” Changzhou
Remand I, 42 CIT at __, 352 F. Supp. 3d at 1330 (quoting 19 C.F.R. § 351.502(b) (2021)).
Commerce must “compare the industries receiving the subsidy to the industry makeup of the
country at issue as a whole[,]” id., in order “to avoid the imposition of countervailing duties in
situations where, because of the widespread availability and use of a subsidy, the benefit of the
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 7 of 33

Consol. Court No. 19-00178 	Page 8
Public Version
subsidy is spread throughout an economy[,]” Uruguay Round Agreements Act, Statement of
Administrative Action, H.R. Rep. No. 103-316, vol. 1, at 930 (1994), reprinted in 1994
U.S.C.C.A.N. 4040, 4242.
In its response to Commerce’s initial questionnaire, the GOC denied the existence of a
countervailable subsidy for aluminum extrusions and maintained that “no specificity exists in the
provision of aluminum extrusions[]” because “[t]here are a vast number of uses for aluminum
extrusions.” GOC Initial CVD Questionnaire Response at 80; see also I&D Memo at 45. The GOC
also denied the existence of any price controls, production level laws or policies, or any export
controls in the aluminum extrusions industry. GOC Initial CVD Questionnaire Response at 94–
96. “While the GOC indicates aluminum extrusions are used in a variety of industries and sectors
across China,” Commerce found that “the industries within those sectors that actually consume
aluminum extrusions are limited in number.” I&D Memo at 45. Comparing “the industries
receiving the subsidy to the industry makeup of the country at issue as a whole[,]” Changzhou
Remand I, 42 CIT at __, 352 F. Supp. 3d at 1330, Commerce concluded that “the number and
nature of the users (as identified by the GOC) are limited compared to the overall structure of the
Chinese economy.” I&D Memo at 45.
In the Third Administrative Review in 2013, the GOC reported six industries that use
aluminum extrusions: “building and construction, transportation, electrical, machinery and
equipment, consumer durables, and other industries.” I&D Memo at 45. The GOC did not provide
any further information on users of aluminum extrusions in this review besides noting that they
are used in a variety of sectors and industries. Id. Relying on a 2007 comprehensive survey from
the China Statistical Yearbook, Commerce examined a list of the types of manufacturers in China
and determined that a significant number and types of manufacturers in China “do not appear to
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 8 of 33

Consol. Court No. 19-00178 	Page 9
Public Version
use aluminum extrusions as an input[.]” Id.; see Administrative Review of Crystalline Silicon
Photovoltaic Cells, Whether or Not Assembled into Modules, from the People’s Republic of
China: Documents for the Record at Attachment A, P.R. 215 (Feb. 12, 2019). Based on this record
evidence, Commerce found that while there were a number of industries that use aluminum
extrusions, there are also comparatively numerous industries that do not. 4 See I&D Memo at 45.
Thus, Commerce determined that “the number and nature of the users (as identified by the GOC)
are limited compared to the overall structure of the Chinese economy.” Id.
Because the GOC failed to submit information about the applications of aluminum
extrusions into the record for this review, Commerce’s analysis on this issue differed slightly from
past administrative reviews because Commerce relied on a generic list of the types of
manufacturers in the PRC rather than on information about specific applications of aluminum
extrusions. Compare I&D Memo at 45, with Changzhou Remand II, 2019 WL 5856438, at *5–6,
and Canadian Solar II, 2020 WL 6129754, at *3–4. The overarching goal of the analysis, however,
is the same, and, although not dispositive, this was a reasonable means of determining whether
beneficiaries of the subsidy were found throughout the Chinese economy. 5 Here, Commerce
4 Specifically, Commerce found that “manufacturers in China produce at least the following
products (the majority of which do not appear to use aluminum extrusions as an input): foods,
beverages, tobacco, textiles, apparel, leather products, furniture, paper and paper products,
recording media, articles for culture, education and sports activities, raw chemical materials and
chemical products, medicines, chemical fibers, rubber, plastics, mineral products, and machinery
for cultural activity and office work, and artwork.” I&D Memo at 45.
5 Canadian Solar argues that Commerce erred by relying on information from a past administrative
review in its analysis of specificity, by focusing on the sectors that the GOC identified as users of
aluminum extrusions in 2013. Canadian Solar Br. at 15–16 (citing Albermarle Corp. &
Subsidiaries v. United States, 821 F.3d 1345, 1356–57 (Fed. Cir. 2016)). In the Final Results,
Commerce did consider the evidence presented by the GOC in the current review by analyzing
“actual users of aluminum extrusions” from the “variety of industries and sectors across China[]”
that use aluminum extrusions, as noted by the GOC. I&D Memo at 45; see GOC Initial CVD
Questionnaire Response at 80; see also Gov. Br. at 18–19.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 9 of 33

Consol. Court No. 19-00178 	Page 10
Public Version
properly compared “the number and nature of the users” of aluminum extrusions to the “overall
structure” of the Chinese economy in reaching its determination that the provision of aluminum
for LTAR is de facto specific as defined in 19 U.S.C. § 1677(5A)(D)(iii)(I). I&D Memo at 45–46.
Although there is some evidence on the record that there are many users of aluminum
extrusions across a broad set of sectors in the Chinese economy, there is evidence in the record to
support Commerce’s determination that “users of aluminum extrusions do not make up something
akin to the whole of the Chinese economy.” I&D Memo at 45; see also Changzhou Remand II,
2019 WL 5856438, at *6. Accordingly, Commerce’s specificity determination regarding the
GOC’s provision of aluminum extrusions for LTAR is not “unsupported by substantial evidence
on the record, or otherwise not in accordance with law,” 19 U.S.C. § 1516a(b)(1)(B)(i), and is
sustained.
III. 	Land Value Benchmark
Prior to finding a countervailable subsidy, Commerce must establish that a benefit was
conferred upon a respondent. 19 U.S.C. § 1677(5)(B). “A foreign government’s provision of
goods to a respondent for LTAR constitutes a benefit.” Changzhou Trina Solar Energy Co., Ltd.
v. United States, Slip Op. 18-31, 2018 WL 1649629, at *2 (CIT Mar. 27, 2018) (citing 19 U.S.C.
§ 1677(5)(E)(iv)). In such circumstances, “Commerce determines the amount of the subsidy by
comparing remuneration actually paid” with a market-determined price for the goods or services,
under “a three-tiered hierarchy” employed by Commerce “to determine the appropriate
remuneration benchmark.” Changzhou Remand I, 42 CIT at __, 352 F. Supp. 3d at 1332; see 19
C.F.R. § 351.511(a)(2)(i)–(iii) (2021). 6
6 Commerce derives a tier one benchmark, “by comparing the government price to a market-
determined price for the good or service resulting from actual transactions in the country in
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 10 of 33

Consol. Court No. 19-00178 	Page 11
Public Version
At issue here is the benchmark set by Commerce in assessing the value of land use-rights.
Canadian Solar argues that Commerce’s use of a tier three benchmark, the 2010 Coldwell
Banker Richard Ellis (“CBRE”) Asian Marketview Report for Thailand Industrial Land Report
indexed using a Consumer Price Index (“CPI”), was unlawful and unsupported by substantial
evidence because: (1) under the tiered hierarchy Commerce should have used the preferred tier
two benchmark data that Canadian Solar provided, world average prices from the 2016 and 2017
CBRE Global Prime Logistics Rents reports, see Letter from Mowry & Grimson PLLC to Sec’y
of Commerce Pertaining to Canadian Solar Benchmark Submission Ex. 5B at 18, P.R. 107-110
(Nov. 5, 2018) (“CS Land Benchmark Submission”), 7 and (2) Commerce’s chosen dataset was
more stale than the data provided by Canadian Solar. Canadian Solar Br. at 32–37. Commerce
maintains that its benchmark determination is lawful, because the nature of land as an in situ
good does not lend itself to use of a tier two benchmark, see I&D Memo at 58–59; Gov. Br. 31–
question.” 19 C.F.R. § 351.511(a)(2)(i). In the absence of such a benchmark, Commerce turns to
a tier two benchmark “by comparing the government price to a world market price where it is
reasonable to conclude that such price would be available to purchasers in the country in
question.” Id. § 351.511(a)(2)(ii). “If there is no world market price available to purchasers in the
country in question,” however, Commerce moves on to a tier three analysis and “measures[s] the
adequacy of remuneration by assessing whether the government price is consistent with market
principles.” Id. § 351.511(a)(2)(iii); see also Countervailing Duties, 63 Fed. Reg. 65,348, 65,378
(Dep’t Commerce Nov. 25, 1998); Nucor Corp. v. United States, 42 C.I.T. __, __, 286 F. Supp.
3d 1364, 1373 & n.13 (2018), aff'd, 927 F.3d 1243 (Fed. Cir. 2019). If Commerce determines
that the government price is not consistent with market principles, it will look to construct an
external benchmark. See Habaş Sinai Ve Tibbi Gazlar Istihsal Endüstrisi A.Ş. et al.v. United
States, 44 CIT __, __, 459 F. Supp. 3d 1341, 1348 n.13 (2020).
7 The 2016 and 2017 CBRE reports include rents for prime industrial and logistics facilities in
global hubs in the Americas, Asia Pacific, Europe, Middle East, and Africa as of Q4 2015
through Q1 of 2017. See CS Land Benchmark Submission Ex. 5B at 18. Canadian Solar argues
that these properties are “similar to the type of land used by Canadian Solar, because Canadian
Solar engaged in logistics operations involving loading and shipping of containerized finished
products.” Id. at 3; see also Canadian Solar Br. at 32. Respondents ask that Commerce use a
global average of the most expensive rents for these facilities as a tier two benchmark. See I&D
Memo at 59; CS Land Benchmark Submission Ex. 5A, Ex. 5B at 14.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 11 of 33

Consol. Court No. 19-00178 	Page 12
Public Version
32, and its selection of the 2010 Thailand data as a tier three benchmark based on geographic
proximity and economic comparability is supported by substantial evidence on the record, see
Gov. Br. at 32–36. Because Commerce properly conducted a tiered benchmark analysis before
choosing a reasonable benchmark, Commerce’s chosen land benchmark is supported by
substantial evidence and is lawful.
First, Commerce determined that it could not rely on a tier one benchmark due to the
GOC’s significant role in domestic land markets and the resulting institutional constraints on
individual land use-rights. See I&D Memo at 58; Land for LTAR Memorandum at 2, 13–27,
P.R. 216 (Feb. 12, 2019) (“Land Use Memo”). Next, Commerce determined that a tier two
world-wide average price would not be appropriate because such a benchmark can be selected
only when it is “reasonable to conclude that such price would be available to purchasers in the
country in question.” 19 C.F.R. § 351.511(a)(2)(ii); I&D Memo at 58–59. Commerce considered
the nature and scope of the market for land and reasonably concluded that land, as an in situ
good, is generally not available to an in-country purchaser on the world market and therefore not
suitable for a tier two benchmark. I&D Memo at 59. Accordingly, Commerce’s determination
that a tier two benchmark was inappropriate was reasonable and its subsequent rejection of
Canadian Solar’s world price benchmark data as a tier two benchmark was lawful.
Next, Commerce moved through a tier three analysis, and reasonably determined “that
government pricing of land-use rights in China is not consistent with market principles.” Land
Use Memo at 30. Substantial evidence on the record shows “there is no meaningful separation
between the government and the land system[]” and “[t]he limited, attenuated market forces
within China’s land system . . . are not sufficient to give the market a decisive and systemic role
in resource allocations.” Land Use Memo at 27–30 (outlining the impact of such policies on the
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 12 of 33

Consol. Court No. 19-00178 	Page 13
Public Version
market philosophy of the Chinese land market). 8 Finding that the GOC’s influence on the land
market rendered the government price not “in accordance with market principles,” Commerce
properly determined that a constructed tier three benchmark was appropriate. Land Use Memo at
30–32; see Habaş Sinai, 44 CIT at __, 459 F. Supp. 3d at 1348 n.13.
Commerce said it would “consider data sources submitted by interested parties to the
administrative record, and carefully evaluate the available record evidence based on . . .
comparability factors” on a case-by-case basis for finding an appropriate benchmark. Land Use
Memo at 31. Commerce identified two important factors for its analysis: geographic proximity
and economic comparability. See id. No party challenges Commerce’s consideration of
geographic proximity and economic comparability as factors when choosing a benchmark for
land prices. Canadian Solar argues instead that Commerce erred by overemphasizing the
geographic proximity factor over the contemporaneity of the data, resulting in an unlawful
benchmark choice. See Canadian Solar Reply Br. at 19–21; Canadian Solar Br. at 35–37. It is
within Commerce’s discretion to weigh the relevant factors. See Nippon Steel Corp. v. United
States, 27 CIT 1856, 1859, 301 F. Supp. 2d 1355, 1360 (2003) (noting that “[i]n conducting its
review, the court's function is not to reweigh the evidence but rather to ascertain whether the
[agency’s] determinations are supported by substantial evidence on the record.”).
In its reply brief, Canadian Solar also argues that even under tier three, Commerce should
have chosen to use data from the 2016 and 2017 CBRE report that Canadian Solar submitted
because it is more contemporaneous. Canadian Solar Reply Br. at 22. Arguments not raised in
an opening brief are waived. See SmithKline Beecham Corp. v. Apotex Corp., 439 F.3d 1312,
8 Domestic Chinese industrial land, for example, is not owned in fee simple, but rather is
allocated to nationals for periods of 40, 50, or 70 years; rural land is similarly contracted out in
30-year periods. See Land Use Memo at 21–22.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 13 of 33

Consol. Court No. 19-00178 	Page 14
Public Version
1319 (Fed. Cir. 2006); see also Sigvaris, Inc. v. United States, 41 CIT __, __, 211 F. Supp. 3d
1353, 1364 (2017). Canadian Solar failed to raise its argument regarding a tier three benchmark
in its opening brief and did not meaningfully develop it in its reply brief, referring to it only in
one sentence. Canadian Solar Reply Br. at 22. Therefore, Canadian Solar’s argument regarding
using data from its land benchmark submission as a tier three benchmark is waived.
Nonetheless, it was reasonable for Commerce not to use the data submitted by Canadian
Solar as a benchmark due to a lack of geographic proximity and economic comparability. See
I&D Memo at 58–59. Specifically, Commerce noted that the dataset included land prices “from
locations such as, e.g., Warsaw, Poland; Stockholm, Sweden; and Atlanta, Georgia” which it
determined “are not reasonable alternatives to China as locations for Asian production.” Id. at
59. While the dataset also included more geographically proximate locations, Commerce
correctly noted that Canadian Solar had not included data in its submission that allowed it “to
evaluate these locations’ economic comparability with respect to China,” id., and while there
may be some economically comparable countries in Canadian Solar’s dataset, (specifically data
from Mexico and Brazil, countries that are on Commerce’s surrogate country list for China in the
antidumping context), they are not geographically proximate to China, a factor that Commerce
says it weighed in its analysis and which neither side disputes as unreasonable.
Based on similarity of geographic location, per capita income, regional population
density, and economic development Commerce’s chosen tier three benchmark: land prices for
industrial zones in Thailand from 2010, is supported by substantial evidence. Land Use Memo at
30–31; see PDM at 18–19; I&D Memo at 59. Both countries have similar levels of economic
development, population density (141 and 127 persons per square kilometer), and regional per
capita income levels; China and Thailand have also in the past competed for the same foreign
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 14 of 33

Consol. Court No. 19-00178 	Page 15
Public Version
direct investment opportunities, underscoring the reasonableness of the Thai land benchmark.
Land Use Memo at 30. To address the stale nature of the data, Commerce reasonably chose to
index these prices using the Thailand CPI from the International Monetary Fund. 9 See
Calculations Memorandum for the Amended Final Results at 553, C.R. 344–45 (Dec. 9, 2016).
Commerce’s use of an indexed 2010 Thailand industrial land price survey as a tier three
benchmark for land prices in China was reasonable and supported by substantial evidence and is
therefore sustained.
IV. 	Creditworthiness of Canadian Solar
In its Final Determination, Commerce concluded that Canadian Solar and certain of its
cross-owned affiliates were uncreditworthy in 2016. I&D Memo at 64–68; see also PDM at 14–
17. 10 Canadian Solar argues that Commerce’s determination was unsupported by substantial
evidence and unlawful because certain of its non-cross-owned affiliates received commercial long-
term loans. See Canadian Solar Br. at 43–46; I&D Memo at 67. The Government claims that
Commerce’s determination is supported by substantial evidence for two reasons. First, the
Government argues that the commercial loans to affiliated companies in 2016 that Canadian Solar
9 Canadian Solar argues that this indexing created a price bloat that inflated the price of land,
because the index included consumer goods. See Canadian Solar Br. at 36–37; Canadian Solar
Reply Br. at 21–22. There is no evidence in the record, however, to suggest that land prices
cannot be indexed using a typical CPI index. In fact, Canadian Solar recommends using a CPI to
adjust the 2016 and 2017 CBRE data it submitted as benchmark data. See CS Land Benchmark
Submission at Ex. 5A. Thus, it was within Commerce’s discretion to weigh the contemporaneous
nature of the data against other factors including economic comparability and geographic
proximity and index the data accordingly.
10 Using the same methodology as Commerce did for Canadian Solar, Commerce reviewed Jinko
Solar’s submission of financial information and concluded that “Jinko Solar’s poor current and
quick ratios” indicated Jinko Solar did not have the liquidity to meet debt obligations and
therefore, it was uncreditworthy pursuant to 19 C.F.R. § 351.505(a)(4)(i)(2021) in 2015 and
2016. I&D Memo at 69. Jinko Solar does not challenge this finding.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 15 of 33

Consol. Court No. 19-00178 	Page 16
Public Version
points 	to 	were 	not 	issued 	to 	cross-owned 	companies 	in 	accordance 	with 	19 	C.F.R.
§ 351.525(b)(6)(vi) 	(2021), 	and 	therefore, 	are 	not 	dispositive 	as 	to 	Canadian 	Solar’s
creditworthiness. See Gov. Br. at 36–40. Second, the Government contends Commerce’s
creditworthiness analysis, and its decision not to consider these loans, was properly based on
“financial indicators as calculated from [Canadian Solar’s] financial statements and accounts.” Id.
at 37–38 (citing I&D Memo at 67).
Commerce examines the creditworthiness of a company as a means of establishing a
benchmark for measuring benefits from government loans. See 19 C.F.R. § 351.505(a)(4); Nucor
Corp. v. United States, 45 CIT __, __, 494 F. Supp. 3d 1377, 1380–81 (2021). “Commerce must
determine in a creditworthiness analysis whether a company could have obtained long-term loans
from 	conventional 	commercial 	sources.” 	Id. 	at 	1380; 	see 	19 	C.F.R. 	§ 	351.505(a)(4)(i)
(“[Commerce] will consider a firm to be uncreditworthy if [Commerce] determines that, based on
information available at the time of the government-provided loan, the firm could not have
obtained long-term loans from conventional commercial sources.”). Regulations enumerate a non-
exhaustive list of factors Commerce may consider in making such a determination, which include
(1) “receipt . . . of comparable commercial long-term loans;” (2) “[t]he present and past financial
health of the firm, as reflected in various financial indicators calculated from the firm’s financial
statements and accounts;” (3) “[t]he firm’s recent past and present ability to meet its costs and
fixed financial obligations with its cash flow; and” (4) “[e]vidence of the firm’s future financial
position[.]” 19 C.F.R. §§ 351.505(a)(4)(i)(A)–(D); “As the regulatory language makes clear,
Commerce has ‘flexibility and discretion in determining which factors to consider and weigh in
making its creditworthiness decision.’” Archer Daniels Midland Co. v. United States, 37 CIT 760,
774, 917 F. Supp. 2d 1331, 1345 (2013) (internal citation omitted).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 16 of 33

Consol. Court No. 19-00178 	Page 17
Public Version
Here, Commerce examined the creditworthiness of five Canadian Solar cross-owned
affiliates (CS Luoyang, CSI Cells, CSI Solar Power, CSI New Energy, and CSI Yancheng) from
2014 to 2016, and concluded that Canadian Solar was uncreditworthy in 2016 based on financial
indicators that revealed a deteriorating financial position beginning in 2015. I&D Memo at 66–68;
PDM at 14–15. 11 Canadian Solar submitted financial data, which Commerce evaluated under the
guidelines set out in 19 C.F.R. § 351.505(a)(4). See I&D Memo at 67. Commerce verified this
information in its review, and confirmed available data from Canadian Solar’s SEC Form 20-F.
See Verification of the Questionnaire Responses Submitted by Canadian Solar Inc. at 6–7, P.R.
224 (July 22, 2019).
Commerce then analyzed Canadian Solar’s financial data and determined that it did not
receive “comparable commercial long-term loans” in 2016 within the meaning of the statute nor
did evidence on the record indicate that it issued comparable convertible notes during the period
of review. See 19 C.F.R §§ 351.505(a)(4)(i)(A), 351.505(a)(4)(ii); I&D Memo at 67; PDM at 16.
Furthermore, indicators 12 of Canadian Solar’s past and present financial health and ability to meet
11 Commerce relied on credit rating reports from Fitch, Moody’s, and S&P to conclude that
Canadian Solar was creditworthy in 2015. The reports concluded favorable long-term prospects
for the solar industry and stated Canadian Solar’s operating stability and efficiency, financial
flexibility and strong market position would enable stability and growth. I&D Memo at 66–67.
To find Canadian Solar creditworthy in 2014, Commerce relied on “a previous finding of
creditworthiness during the 2014 administrative review.” Id. at 66; see also PDM at 15.
12 Commerce uses quick and current ratios, which measure a company’s short-term liquidity, and
therefore its ability to pay short-term obligations, to determine creditworthiness under 19 C.F.R.
§§ 351.505(a)(4)(i)(B)–(C). Current ratios measure a company’s ability to pay current or short-
term liabilities with current or short-term assets and are calculated by dividing current assets by
current liabilities. Commerce’s typical benchmark for a creditworthy company’s current ratio is
2; Commerce will normally find companies with current ratios below 2 to be uncreditworthy.
I&D Memo at 67; PDM at 16. Quick ratios are a more conservative measure that include only
assets that could be converted to cash within 90 days or less. Quick ratios are calculated by
adding cash (and cash equivalents), current and account receivables, any short-term investments
and dividing all of that by current and short-term liabilities. Commerce’s typical benchmark for a
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 17 of 33

Consol. Court No. 19-00178 	Page 18
Public Version
its costs and financial obligations were unacceptably low, which led Commerce to make an
uncreditworthy 	determination. 	I&D 	Memo 	at 	67–68; 	PDM 	at 	16; 	see 	also 	19 	C.F.R.
§§ 351.505(a)(4)(i)(B)–(C). Commerce calculated Canadian Solar’s current ratio as 1.02 and quick
ratio as 0.38 in 2016, both of which were significantly below the benchmarks (2.0 and 1.0
respectively) Commerce deems as creditworthy. I&D Memo at 67. Commerce also explained that
Canadian Solar’s retained cash flow, which measures cash after paying liabilities, continued to
decrease further into the negative, which, combined with the ratio levels, reflected that Canadian
Solar did not have “liquid funds to cover its upcoming obligations.” I&D Memo at 67–68.
Canadian Solar does not dispute Commerce’s calculation of these ratios, nor the
appropriateness of the benchmarks, but instead argues that commercial loans made to non-cross-
owned affiliates in 2016 indicate its creditworthiness. See Canadian Solar Br. at 43–46; I&D
Memo at 64–67; Canadian Solar Resp. to Ct.’s Order of May 19, 2021 at 11–12, ECF No. 114
(June 2, 2021) (“CS Resp. to Ct. Order”). Commerce explained that it did not consider these loans
in its analysis because they were not provided to cross-owned entities in accordance with 19 C.F.R.
§ 	351.525(b)(6)(vi), 	which 	would 	therefore 	be 	dispositive 	evidence 	of 	Canadian 	Solar’s
creditworthiness under 19 C.F.R § 351.505(a)(4)(ii) . I&D Memo at 67; see Gov. Br. at 37–39.
Canadian Solar does not dispute that the loans were made to non-cross-owned affiliates. Canadian
Solar Reply Br. at 23. Nonetheless, pointing to its SEC Form 20-F, Canadian Solar argues that its
financial reporting of affiliate financial activity shows it was creditworthy in 2016. See CS Resp.
to Ct. Order at 11–12.
creditworthy company’s quick ratio is 1. I&D Memo at 67; PDM at 16. If a company’s quick
ratio is below 1, Commerce will normally deem that entity uncreditworthy. I&D Memo at 67;
PDM at 16.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 18 of 33

Consol. Court No. 19-00178 	Page 19
Public Version
Commerce’s decision not to consider affiliate loans was reasonable because of the
distinction between cross-ownership and affiliation. 13 Evidence on the record does not indicate
that Canadian Solar nor any cross-owned affiliate was party to a commercial loan and therefore,
that this activity is dispositive of Canadian Solar’s creditworthiness. See Canadian Solar Br. at 45–
46; Canadian Solar Reply Br. at 23; CS Resp. to Ct. Order at 11–12. A loan to a non-cross-owned
affiliate, as is the case here, may have no financial bearing on the creditworthiness of another
simply affiliated entity and it was reasonable in these circumstances for Commerce to conclude
so.
Canadian Solar did not submit financial forecasts, market ratings, or other appraisals
indicative 	of 	the 	companies’ 	future 	financial 	position 	for 	2016 	under 	19 	C.F.R.
§ 351.505(a)(4)(i)(D). See I&D Memo at 67–68; see also PDM at 17. Accordingly, Commerce’s
review of the financial information submitted, analysis under the four factors in 19 C.F.R. §
351.505(a)(4)(i), and determination that Canadian Solar was uncreditworthy in 2016 was
reasonable and is sustained.
V. 	Entered Value Adjustment
Commerce establishes countervailing duty (“CVD”) rates by calculating an “individual
countervailable subsidy rate[] for each investigated exporter and producer.” Jiangsu Zhongji
Lamination Materials Co., Ltd. v. United States, 43 CIT __, __, 405 F. Supp. 3d 1317, 1326
(2019) (“Jiangsu I”); see 19 U.S.C. §§ 1671d(c)(1)(B)(i)(I) (2021), 1671b(d)(1)(A)(i) (2021).
The subsidy rate is calculated by dividing the amount of the benefit afforded by the subsidy
13 Cross-ownership “between two or more corporations” exists “where one corporation can use
or direct the individual assets of the other corporation(s) in essentially the same ways it can use
its own assets.” 19 C.F.R. § 351.525(b)(6)(vi). In contrast, an affiliate can be any entity or person
with influence or control over another. See 19 U.S.C. § 1677(33); 19 C.F.R. § 351.102(b)(3)
(2021).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 19 of 33

Consol. Court No. 19-00178 	Page 20
Public Version
during the period of investigation (numerator) by the sales value of the product during the same
period (denominator). 19 C.F.R. § 351.525(a) (2021). CVDs are intended to be remedial rather
than punitive in nature. Chaparral Steel Co. v. United States, 901 F.2d 1097, 1103 (Fed. Cir.
1990). Thus, the CVD must equal the amount of the net countervailable subsidy. 19 U.S.C. §
1671(a) (2021). Furthermore, Commerce has a duty to determine CVD rates as accurately as
possible. Içdaş Celik Enerji Tersane ve Ulasim Sanayi, A.S. v. United States, 45 CIT __, __, 498
F. Supp. 3d 1345, 1353 (2021).
Ideally, the free on board (F.O.B.) export value upon which the subsidy rate is initially
calculated equals the import value of the merchandise entering the United States upon which duties
are collected. Jiangsu I, 43 CIT at __, 405 F. Supp. 3d at 1326. Thus, in such a case the collection
of duties would equal the net countervailable subsidy as required by 19 U.S.C. § 1671(a). Id.
Sometimes, however, merchandise is marked up between export and import into the United States,
“creat[ing] a mismatch between the previously calculated subsidy rate and the final invoiced price
to which the subsidy rate is applied[.]” Id. at 1327. This situation most commonly occurs when
merchandise is sold by the exporter “in a back-to-back intercompany sales transaction to a foreign
affiliate, which then sells the merchandise to the United States with a mark-up.” Id. at 1326. A
mark-up can result in an overcollection of duties, because the mark-up in price already offsets
some of the benefit afforded by the countervailable subsidy. Not accounting for this mark-up
means the CVD rate will not equal the net countervailable subsidy, as required by statute. See 19
U.S.C. § 1671(a). In such a situation, the exporter may request that Commerce adjust the ad
valorem subsidy rate during an administrative review. Jiangsu I, 43 CIT at __, 405 F. Supp. 3d at
1327. This adjustment is known as the entered value adjustment (“EVA”).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 20 of 33

Consol. Court No. 19-00178 	Page 21
Public Version
Commerce has described its granting of an EVA as a “practice,” and often emphasizes that
such an adjustment is not mandated by statute or regulation. Id. For this reason, Commerce asserts
it does not make specific requests regarding EVA in its questionnaires to respondents. Instead,
respondents present evidence to Commerce that they believe demonstrates their eligibility for an
EVA. Recent changes to Commerce’s methodology in calculating the EVA has led to some
confusion about what evidence Commerce requires to grant an EVA.
Commerce first outlined the requirements for granting an EVA in Ball Bearings and Parts
Thereof from Thailand; Final Results of Countervailing Duty Administrative Review, 57 Fed. Reg.
26,646, 26,647 (Dep’t Commerce June 15, 1992). In that case, Commerce adjusted the subsidy
rate by calculating the mark-up ratio on U.S. sales and then multiplying that ratio by the original
subsidy rate to obtain the adjusted subsidy rate for U.S. sales. Id. In this way, Commerce
specifically accounted for the mark-up ratio on merchandise entering the United States. Based on
this review, Commerce established six criteria that respondents must demonstrate to qualify for an
EVA:
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 6) the
invoices can be tracked as back-to-back invoices that are identical except for price.
Jiangsu I, 43 CIT at __, 405 F. Supp. 3d at 1327.
In recent reviews, however, Commerce appears to have taken a new approach for
calculating an EVA. Commerce seemingly now adjusts a subsidy rate by changing the
denominator of the original subsidy rate from the respondent’s total worldwide export sales to the
affiliate’s marked-up total worldwide export sales. See I&D Memo at 72–75; see also Jiangsu I,
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 21 of 33

Consol. Court No. 19-00178 	Page 22
Public Version
Confidential Information Omitted
43 CIT at __, 405 F. Supp. 3d at 1327–28 (explaining that Commerce adjusted the subsidy rate to
include “all sales by Zhongji HK [the affiliate], i.e. the higher valued sales, instead of all sales by
Zhongji to Zhongji HK, i.e. the lower valued sales”). This calculation method does not specifically
account for a U.S. mark-up and does not require U.S.-specific sales data. Despite not needing U.S.
sales data for its EVA calculation, Commerce has not appeared to adjust its evidentiary
requirements to reflect this new methodology. Commerce still requires respondents to demonstrate
“that there is a higher customs value for all of its U.S. sales.” I&D Memo at 75. Commerce asserts
that it must be satisfied “that the sales value adjustment properly reflects an upward adjustment to
the sales value of all merchandise that entered the United States, and on which Customs and Border
Protection assessed dutiable value.” Id. In practice, without additional guidance from Commerce
regarding their submissions, respondents have typically submitted worldwide sales data showing
an aggregate mark-up and a sample U.S. invoice demonstrating Commerce’s six criteria, and the
existence of a U.S. mark-up. See Canadian Solar I, 2020 WL 898557, at *8–9.
Here, Canadian Solar did just that, submitting data showing that its total sales exported to
the U.S. through affiliates were marked-up in the aggregate, and a sample U.S. invoice
demonstrating a U.S. mark-up (although the U.S. mark-up is different than the aggregate
worldwide markup). See Canadian Solar Section III Questionnaire Response, at 11–12, Exs. 8.1
and 8.2; see also Canadian Solar Br. at 48–49. Canadian Solar stated that an EVA was appropriate
because it demonstrated that it met each of the six criteria:
(1) The U.S. invoice is through [[ 	]]. The invoice includes a mark-up from the
invoice issued from our company to [[ 	]]. For example, in the sample sales
documentation, our company's invoice No. [[ 	]] valued USD [[ 	]],
and [[ 	]] invoice No. [[ 	]] valued USD [[ 	]].
(2) Our company, the exporter out of China, is affiliated with [[ 	]], as is explained
in the Affiliated Companies Questionnaire Response.
(3) The U.S. invoice issued by [[ 	]] establishes the customs value to which
countervailing duties would be applied. Please compare [[ 	]] invoice No. [[
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 22 of 33

Consol. Court No. 19-00178 	Page 23
Public Version
Confidential Information Omitted
]] totaling [[ 	]], with the Customs Form 7501, which lists the same total
entry value of [[ 	]].
(4) There is a one-to-one correlation between our company's invoice and [[ 	]]
invoice, as can be seen in the sample sales documentation. Both invoices
include the same job numbers and the same quantity of [[ 	]] PCS.
(5) The merchandise is shipped directly to the United States. The Bill of Lading
indicates the product is shipped from [[ 	]]. Both the Bill of
Lading and [[ 	]] packing list include the same quantity of [[ 	]] KGS.
(6) The invoices can be tracked as back-to-back invoices that are identical except
for price. Specifically, our company's invoice and [[ 	]] invoice include the
same job numbers and the same quantity of [[ 	]].
Canadian Solar Section III Questionnaire Response at 10. Commerce disputes that the one sample
invoice provided by Canadian Solar in this administrative review demonstrates “that the overall
mark-up to its export sales reflects a mark-up on its U.S. export sales” because of an inconsistency
between the mark-up present in the sample U.S. invoice and the aggregate mark-up present in the
worldwide sales data. 14 Gov. Br. at 28–30; see also Canadian Solar I, 2020 WL 898557, *8–9.
Canadian Solar avers that Commerce inappropriately denied it an EVA for two reasons. First, it
demonstrated that its total sales exported to the U.S. through affiliates were marked-up in the
aggregate. Canadian Solar Br. at 48–49. Second, if Canadian Solar was required to demonstrate
each U.S. sale was marked up for all of its U.S. sales to qualify for an EVA, Commerce was
required to notify Canadian Solar of this deficiency in its submission and give it an opportunity to
14 “[T]he court acknowledge[d] that granting an EVA appears feasible only when a mark-up is
added consistently and as a matter of course, given verification concerns.” Canadian Solar I, 2020
WL 898557, at *9. This concern is understandable because a respondent could potentially receive
an excessive EVA based on a sample U.S. invoice showing a higher U.S. markup that is not
consistently applied to all U.S. sales. Nevertheless, a respondent could also be improperly denied
an EVA when the U.S mark-up is inconsistent, even though U.S. sales are still marked up as a
matter of course, potentially resulting in an inaccurate subsidy rate in violation of 19 U.S.C. §
1671(a).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 23 of 33

Consol. Court No. 19-00178 	Page 24
Public Version
provide the necessary additional information pursuant to 19 U.S.C. § 1677m(d) (2020). 15 Id. at
49–52.
Commerce’s desire to verify the existence of a U.S. markup is reasonable because an EVA
is intended to prevent the overcollection of duties, which can only occur upon entry of the
merchandise to the United States. Canadian Solar I, 2020 WL 898557, at *8. Nevertheless,
Commerce’s current methodology for calculating an EVA, by replacing the denominator of the
subsidy rate with the total worldwide marked up sales value, does not necessarily confirm the U.S.
mark-up about which Commerce is most concerned, and does not require U.S. specific sales data.
The court previously noted this discrepancy between the evidence Commerce requires and its
chosen calculation method in Jiangsu I. See 43 CIT at __, 405 F. Supp. 3d at 1330 (noting
“Commerce does not adequately explain why, given the calculation methodology employed in this
case, the identification of U.S. sales or U.S. customers is relevant to the EVA determination”). 16
15 The only “guidance” Commerce provided regarding an EVA was instructing Canadian Solar to
[s]eparately report the value of services sold by your company, if any. In addition,
separately report the value of sales by each cross-owned company, as well as the
value of sales between your company and the cross-owned company. . . . Please
report sales information for all responding cross-owned companies during these
years, not only sales information for the recipient of the subsidy.
Canadian Solar Section III Questionnaire Response at 9. In response to the questionnaire,
Canadian Solar submitted its sales data and requested Commerce make an EVA. Id. at 10.
16 On remand, Commerce “agree[d] with the Court” and clarified that “the way that [it] made the
adjustment in the Preliminary Determination was incorrect, as any adjustment should have
properly been focused on Zhongji’s mark-up to its U.S. sales.” Final Results of Redetermination
Pursuant to Court Order Summary, C-570-054, POR 01/01/2016-12/31/2016 at 6–7 (Dep't
Commerce Jan. 27, 2020). Commerce recognized, however, that it failed to communicate the
change in calculation methodology before the Final Determination to Zhongji and ultimately
granted the EVA. Id. at 8; Jiangsu Zhongji Lamination Materials Co., Ltd. v. United States, Slip
Op. 20-39, 2020 WL 1456531, at *1 (CIT Mar. 24, 2020). Nevertheless, Commerce seems to
describe this method in cases commenced before this redetermination, including this review,
describing an EVA as seeking “an adjustment to [a respondent’s] sales denominator[.]” I&D
Memo at 75.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 24 of 33

Consol. Court No. 19-00178 	Page 25
Public Version
Confidential Information Omitted
Canadian Solar argues that requiring it to demonstrate the six criteria are met for all sales “does
not directly correlate with [the] underlying purpose of an EVA to prevent the over-collection of
duties.” Canadian Solar Br. at 50–51.
Evidence in the record indicates that all of Canadian Solar’s shipments entered the U.S.
through an affiliate. See Canadian Solar Section III Questionnaire Response at 9; see also Jiangsu
I, 43 CIT at __, 405 F. Supp. 3d at 1330. The sample U.S. invoice demonstrates that the shipment
in question was marked up when entering the U.S., although the markup is lower than the markup
demonstrated in Canadian Solar’s aggregate worldwide sales data. See Canadian Solar Section III
Questionnaire Response at 10 (showing a mark-up from [[ 	]] in the U.S. and
showing an overall mark-up from [[ 	]]). This inconsistency in the
markup might present a verification issue and leave a question of whether all U.S. invoices were
marked up. Commerce, however, should have requested supplemental information pursuant to 19
U.S.C. § 1677m(d) if it found the information that Canadian Solar submitted deficient to resolve
this.
The Government’s argument that it is not required to request supplemental documentation
from Canadian Solar because Section 782(d) of the Tariff Act only requires Commerce to request
additional information where it deems that a response to an initial questionnaire was insufficient
is not convincing. See Gov. Br. at 30; I&D Memo at 75–76 (quoting 19 U.S.C. § 1677m(d)).
Commerce reviewed and considered Canadian Solar’s submissions in its EVA analysis, creating
an impression that Canadian Solar’s su
bmission was accepted as a response, without providing any notice to Canadian Solar of
any deficiencies until the Preliminary Determination. Thus, Commerce must give Canadian Solar
an opportunity to correct any deficient information. See
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 25 of 33

Consol. Court No. 19-00178 	Page 26
Public Version
Shelter Forest Int'l Acquisition, Inc. v. United States, 45 CIT __, __, 497 F. Supp. 3d 1388, 1399
(2021). 17
Commerce has two options on remand. It can grant the EVA to Canadian Solar, given that
the evidence provided should be sufficient for completing the EVA calculation and that Canadian
Solar has demonstrated that there is a U.S. mark-up on the sample invoice provided. If Commerce
refuses to grant the EVA because it deems the evidence insufficient to separately verify a U.S.
mark-up, then Commerce must clarify its calculation methodology, explain the evidence it requires
for verification and why that evidence is reasonable given the calculation methodology, give
Canadian Solar an opportunity to submit the additional evidence, and in the case that new
information is submitted, reassess Canadian Solar’s eligibility in the light of the supplemental
documentation. In sum, Commerce’s stated methodology does not match up with the information
it asserts is needed. Where warranted an EVA prevents an overcollection of duties and Commerce
cannot avoid acting fairly just because there is no statute or regulation specifically addressing
EVA. 18
VI. 	Export Buyer’s Credit Program
a. 	Background
17 At oral argument, the Government raised new arguments about Canadian Solar’s sales data
which were not included in the I&D Memo. See Oral Argument at 21:00-25:04, 27:26-27:38.
Because these arguments are post-hoc rationalizations, we do not consider them here. See U.H.F.C.
Co. v. United States, 916 F.2d 689, 700 (Fed. Cir. 1990) (“Post hoc rationalizations of agency
actions first advocated by counsel in court may not serve as the basis for sustaining the agency's
determination.”).
18 The court notes that Commerce requested additional information regarding EVA from a different
respondent in a recent review, so apparently Commerce does recognize this principle. See Decision
Memorandum for the Preliminary Results of the Countervailing Duty Administrative Review of
Certain Aluminum Foil from the People’s Republic of China; 2017-2018, C-570-054, POR:
8/14/17-12/31/18, at 10 (Dep't Commerce June 17, 2020).
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 26 of 33

Consol. Court No. 19-00178 	Page 27
Public Version
The GOC’s Export Buyer’s Credit Program (“EBCP”) promotes exports by providing
credit at preferential interest rates to qualifying foreign purchasers of PRC goods. See Clearon
Corp. v. United States, 43 CIT __, __, 359 F. Supp. 3d 1344, 1347 (2019). Mandatory
Respondents reported that to their knowledge none of their U.S. customers received assistance
under the EBCP, and that Mandatory Respondents did not assist their customers in participating
in the program. See Canadian Solar Section III Questionnaire Response at 31 (stating “none of
these unaffiliated companies received assistance under the Export Buyer’s Credit program” and
“nor did our company. . .provide assistance to any customers in obtaining buyer credits”); Jinko
Section III Questionnaire Response at 40–41 (stating “Jinko did not play any role in assisting its
customers in obtaining buyer credits,” and “none of these unaffiliated companies received
assistance under this program [the EBCP]”). Mandatory Respondents also submitted
certifications of non-use of EBCP from their U.S. customers. See Canadian Solar Section III
Questionnaire Response, Ex. 15, (notarized and signed customer declarations from Canadian
Solar’s customers); Jinko Solar Section III Questionnaire Response, Ex. IQR-JJ-25, (notarized
and signed customer declarations from Jinko’s customers). Furthermore, Mandatory
Respondents indicated that they were available to answer Commerce’s questions and would
encourage any customers to similarly cooperate with Commerce’s verification attempts. See
Canadian Solar Section III Questionnaire Response at 31; Jinko Section III Questionnaire
Response at 40–41.
Based on information from prior investigations, Commerce determined that administrative
changes in the EBCP in 2013 may have (1) removed the previous requirement that a contract be
worth over $2 million to qualify for credits through the EBCP and (2) allowed unspecified third-
party banks to distribute EBCP funds to importers rather than China Export-Import Bank (“EX-
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 27 of 33

Consol. Court No. 19-00178 	Page 28
Public Version
IM Bank”). See I&D Memo at 26–28. 	In order to understand the operation of the EBCP,
Commerce requested that the GOC provide (i) the EBCP Administrative Measures that were
revised in 2013, and (ii) a list of all “partner/correspondent banks” involved in the disbursement
of funds under the EBCP. GOC Supplemental CVD Questionnaire Response at 1; see also I&D
Memo at 28. The GOC declined to provide this information, asserting that these questions were
“not applicable” because no U.S. customers of the Mandatory Respondents used the EBCP. GOC
Supplemental CVD Questionnaire Response at 1. Commerce also requested “a sample buyer’s
credit application. . .the application’s approval and the agreement between the respondent’s
customer and the bank, which established the terms of the assistance provided under the facility.”
GOC Initial CVD Questionnaire Response at 126. The GOC did not provide these documents to
Commerce, stating that no such agreements existed because no U.S. customers used the EBCP. Id.
at 126–27. Without this information, Commerce found that its understanding of the program was
incomplete and unreliable and that it was thus unable to verify the customer certifications of non-
use provided by Mandatory Respondents. Id. at 34–38. Accordingly, Commerce applied adverse
facts available (“AFA”) to determine that Mandatory Respondents used the EBCP. Id. at 8, 38–
39.
Canadian Solar and Jinko objected to Commerce’s application of AFA, arguing that there
was no gap in the record regarding the use of EBCP because customers provided verifiable
certifications attesting to their non-use of EBCP and therefore an adverse inference was
impermissible. See Canadian Solar. Br. at 10–14; Jinko Br. at 23–30. In response, Commerce
maintained that its application of AFA was lawful because the GOC failed to cooperate with its
requests for information. Gov. Br. at 9–13. The Government also noted its position that it was not
possible to verify the customer certifications of non-use. Id. at 13–15.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 28 of 33

Consol. Court No. 19-00178 	Page 29
Public Version
Following oral argument, the court ordered Mandatory Respondents to file supplemental
briefing detailing how Commerce could verify non-use of the EBCP through information collected
from Mandatory Respondents and their customers, and for Commerce to explain whether the
proposed verification method was feasible. Mandatory Respondents presented multiple avenues
for verification, including proposed verification methods at the exporter, importer, and the EX-IM
bank. See Suppl. Br. Per Ct. Order of Pls. Canadian Solar Inc. et al. and Consolidated-Pls. Jinko
Solar Co., Ltd. et al. Related to Export Buyer’s Credit Program, ECF No. 115 (confidential), ECF
No. 116 (public) (June 9, 2021). In response, Commerce asserted that verification was unfeasible
despite Mandatory Respondents proposed method and argued, in short, that it could not verify the
customer certifications of non-use because it did not have the names of partner banks and a sample
“paper trail” of documents associated with an EBCP loan that were necessary for verification. See
Def.’s Suppl. Br. on the Issue of the Export Buyer’s Credit Program at 4–5, 12, ECF No. 121 (July
16, 2021) (“Gov. Supp. Br.”).
b. 	Discussion
If “necessary information is not available on the record” or if a responding party “withholds
information” requested by Commerce, Commerce shall “use the facts otherwise available in
reaching the applicable determination[.]” 19 U.S.C § 1677e(a) (2020). Commerce may use AFA
only when information is missing on the record because a party “has failed to cooperate by not
acting to the best of its ability to comply with a request for information” from Commerce. 19
U.S.C. § 1677e(b). Although Commerce can apply adverse facts that collaterally impact a
cooperating party, “Commerce should seek to avoid such impact if relevant information exists
elsewhere on the record.” Archer Daniels, 37 CIT at 769, 917 F. Supp. 2d at 1342; see also Guizhou
Tyre Co., Ltd. v. United States, 42 CIT __, __, 348 F. Supp. 3d 1261, 1270 (2018) (“To apply
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 29 of 33

Consol. Court No. 19-00178 	Page 30
Public Version
AFA in circumstances where relevant information exists elsewhere on the record — that is, solely
to deter non-cooperation or ‘simply to punish’ — . . . that is a fate this court should sidestep.”)
(citation omitted). Information submitted by parties is subject to verification by Commerce. 19
U.S.C. § 1677m(i)(1). “Commerce need not consider information submitted by respondents that
cannot be verified,” but Commerce “must first reasonably show that such information is, in fact,
unverifiable.” See Changzhou Remand I, 42 CIT at __, 352 F. Supp. 3d at 1327 (citing 19 U.S.C.
§ 1677m(e)).
As this court has previously discussed, there is no gap in this record as it stands regarding
EBCP usage because the Mandatory Respondents have provided evidence of non-use in the form
of customer certifications. See Canadian Solar Section III Questionnaire Response at Ex. 15, Jinko
Solar Section III Questionnaire at Ex. IQR-JJ-25. The question here is whether Commerce’s claim
that customer certifications of non-use are unverifiable is supported by substantial evidence. We
conclude, as we have in the past, that Commerce has not reasonably shown that the customer
certifications are unverifiable on the record before us. See Changzhou Remand I, 42 CIT at __,
352 F. Supp. 3d at 1326–27; see also Changzhou Trina Solar Energy Co., Ltd. v. United States,
Slip Op. 19-143, 2019 WL 6124908, at *1–3 (CIT Nov. 18, 2019).
Despite more generalized complaints, the “missing” information that might potentially
impact verification is the identity of the partner banks. Nonetheless, Commerce’s arguments that
EBCP non-usage is unverifiable without the names of the partner banks lack sufficient support.
See I&D Memo at 32. Commerce posits that to verify use of a loan program, it would typically
examine the company’s subledgers for references to the party making the financial contribution
and request underlying documentation from specific entries to confirm the origin of each loan.
Gov. Supp. Br. at 10; see also I&D Memo at 32–33. Because it does not know the names of the
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 30 of 33

Consol. Court No. 19-00178 	Page 31
Public Version
partner banks that may disburse loans under the EBCP, Commerce claims it would have to review
underlying documentation for all loans reflected in the subledger to determine whether any loan
was associated with the EBCP. Gov. Supp. Br. at 10. This review would be “an unreasonably
onerous undertaking for any company that received more than a small number of loans.” Id. (citing
I&D Memo at 33). Commerce, because it never proceeded to attempt verification, does not indicate
the number of loans in question and does not demonstrate how burdensome verification would be
here, however. The court acknowledges that verification would appear to be easier if Commerce
had access to the names of the partner banks and could review specific entries in a customer’s
subledger, and that Commerce is entitled to consider the burden of verification on agency
resources. See Torrington Co. v. United States, 68 F.3d 1347, 1351–52 (Fed. Cir. 1995). But here,
Commerce has provided insufficient evidence to conclude that verification by the usual spot check
of the customer records would be unduly burdensome. This record also suggests that Commerce
would have likely had adequate opportunity to attempt verification as Mandatory Respondents
offered to aid Commerce’s verification efforts and indicated that they would encourage
unaffiliated customers to do the same. Commerce made no such attempt and its claims of
unverifiability are unavailing. 19
Commerce also argues that the lack of sample EBCP paperwork prevents verification at
the U.S. customer because even if Commerce were to review the underlying documentation of
every customer loan, it would not understand “whether/how that documentation would indicate
China Ex-Im involvement.” I&D Memo at 33. Commerce has not provided a reasonable
explanation for why verification is unfeasible without examples of forms. Commerce simply
19 Normally the court does not expect Commerce to attempt to obtain financial records from non-
parties. The facts here are unusual because customer cooperation has already been demonstrated
to some extent.
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 31 of 33

Consol. Court No. 19-00178 	Page 32
Public Version
assumes that documents demonstrating EBCP use would be unidentifiable and incomprehensible
without guidance from the sample paperwork. Having apparently not reviewed or requested any
customer loans records for an inquiry into EBCP, despite apparently having opportunity to do so,
Commerce has a limited basis for this assumption.
Nor has Commerce shown why verification at the respondent is unfeasible. The record
demonstrates that the exporter would be involved in various stages of the EBCP. See e.g., GOC
Initial CVD Questionnaire Response at 127–28 (stating “Normally, if export buyer’s credits are
provided by the EX-IM bank, the Chinese exporter is aware of the buyers receipt of the loans and
is involved in the loan evaluation proceeding and, in particular, is involved in post-lending loan
management conducted by the EX-IM Bank.”) Commerce argued that it cannot verify usage at
the exporter because Commerce lacks the sample paperwork and information regarding the
disbursing banks requested from the GOC, and it needed a “better understanding of the program
before it could verify it.” I&D Memo at 34–35; Gov. Supp. Br. at 8–10. Specifically, Commerce
noted it would be unable to confirm non-use by “examining books and records which can be
reconciled to audited financial statements or other documents, such as tax returns.” Id. at 35. But
Commerce does not sufficiently explain why it would be unable to conduct an adequate review
of Canadian Solar and Jinko’s accounting records, especially where, as here, the record indicates
that Commerce would likely have Respondent’s cooperation in such a review. Commerce failed
to demonstrate that a fuller understanding of the functioning of the EBCP is required to verify
non-use of the program.
Commerce’s should “seek to avoid” application of AFA that collaterally impacts a
cooperating party. See Archer Daniels, 37 CIT at 769, 917 F. Supp. 2d at 1342. Here, Mandatory
Respondents have cooperated by attesting they are not involved with the EBCP, providing
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 32 of 33

Consol. Court No. 19-00178 	Page 33
Public Version
customer certifications stating that their customers were not involved in the EBCP, offering to
assist in verification of non-use, and stating that they would encourage their customers to do the
same. There is no record evidence suggesting that Mandatory Respondents or their customers used
this program, and no reason to doubt the legitimacy of the Mandatory Respondents’ statements of
nonuse or customers’ certifications. Commerce’s claims that it is unable to verify these
certifications without information from the GOC are not supported by substantial evidence.
Commerce has not attempted verification and thus far has no basis to claim unverifiability. Its
assertion that verification would be overly burdensome did not suffice to justify its use of AFA.
On remand, Commerce may attempt to verify Mandatory Respondents’ claims of non-use through
the means proposed by Mandatory Respondents or any other reasonable procedure at its disposal.
Alternatively, Commerce can elect not to extend verification and simply accept Mandatory
Respondents’ evidence of non-use.
CONCLUSION
The court sustains Commerce’s determination regarding, (1) the specificity finding for
the aluminum extrusions for LTAR program, (2) Commerce’s chosen benchmark for the land for
LTAR program and, (3) Canadian Solar’s lack of creditworthiness in 2016. For the foregoing
reasons, the court remands to Commerce for a determination consistent with this opinion on the
remaining issues. The remand determination shall be issued within 60 days hereof. Comments
may be filed 30 days thereafter and any response 15 days thereafter.
__/s/ Jane A. Restani_________
Jane A. Restani, Judge
Dated: September 3, 2021
New York, New York
Case 1:19-cv-00178-JAR Document 125 Filed 09/03/21 Page 33 of 33