Husteel Co. v. United States

98 F. Supp. 3d 1315, 2015 CIT 100, 37 I.T.R.D. (BNA) 2102, 2015 Ct. Intl. Trade LEXIS 100, 2015 WL 5132123

Opinion

OPINION

RESTANI, Judge:

This action challenges the Department of Commerce’s (“Commerce”) final determination rendered in the antidumping (“AD”) duty investigation of certain oil country tubular goods (“OCTG”) 1 from the Republic of Korea (“Korea”). See Certain Oil Country Tubular Goods From the Republic of Korea: Final Determination of Sales at Less Than Fair Value and Negative Final Determination of Critical Circumstances, 79 Fed.Reg. 41,983 (Dep’t Commerce July 18, 2014) (“Final Determination”). Before the court are the motions for judgment on the agency record of Korean producers Husteel Co., Ltd. (“Husteel”), NEXTEEL Co., Ltd. (“NEXTEEL”), ILJIN Steel Corporation (“IL-JIN”), AJU Besteel Co., Ltd. (“AJU Besteel”), and SeAH Steel Corp. (“SeAH”) (collectively, “plaintiffs”). See Br. of Pl. Husteel Co., Ltd. in Supp. of Its Mot. for J. on the Agency R., DE 95 (“Husteel Br.”); Mem. in Supp. of Consol. Pl. NEXTEEL’s Rule 56.2 Mot. for J. upon the Agency R., DE 106 (“NEXTEEL Br.”); Mem. in Supp. of Consol. Pl. HYSCO’s Rule 56.2 Mot. for J. upon the Agency R., DE 104 (“HYSCO Br.”); Br. of Pl. Intvnr. ILJIN Steel Corp. in Supp. of Its Mot. for J. on the Agency R., DE 89-1 (“ILJIN Br.”); Mot.' of Consol. Pl. AJU Besteel Co., Ltd. for J. upon the Agency R., DE 82-1 (“AJU Besteel Br.”); Br. of Pl. SeAH Steel Corp. in Supp. of Its Rule 56.2 Mot. for J. on the Agency R., DE 86 (“SeAH Br.”). Also before the court are the motions for judgment on the agency record of U.S. producers United States Steel Corporation (“U.S. Steel”) and Maverick Tube Corporation (“Maverick”) (collectively, “petitioners”). See Mot. of Pl. United States Steel Corp. for J. on the Agency R. Under Rule 56.2, DE 97 (“U.S. Steel Br.”); Pl.’s Maverick Tube Corp. Mem. in Supp. of Its Rule 56.2 Mot. for J. on the Agency R., DE 102 (“Maverick Br.”). For the reasons stated below, Commerce’s Final Determination is sustained in part and remanded in part.

BACKGROUND

Following the filing of a petition by U.S. Steel, Maverick, and other domestic producers of OCTG, Commerce initiated an AD investigation of OCTG from Korea on July 22, 2013. See Certain Oil Country Tubular Goods from India, the Republic of Korea, the Republic of the Philippines, Saudi Arabia, Taiwan, Thailand, the Republic of Turkey, Ukraine, and the Socialist Republic of Vietnam: Initiation of Antidumping Duty Investigations, 78 Fed. Reg. 45,505, 45,506, 45,512 (Dep’t Commerce July 29, 2013) (“Initiation Notice ”). On August 26, 2013, Commerce limited the number of respondents for individual examination, selecting the two exporters or producers of OCTG that accounted for the largest volume of imports from Korea to the United States: NEXTEEL and HYS-CO. Respondent Selection Memorandum at 6-8, PD 80 (Aug. 27, 2013) (“Respondent Selection Memo”). Because the two mandatory respondents did not have viable home or third-country markets for OCTG, pursuant to 19 U.S.C. § 1677b(a)(4) (2012), Commerce used a constructed value (“CV”) to determine the appropriate normal value. 2 Issues and Decision Memorandum for the Final Affirmative Determination in the Less than Fair Value Investigation of Certain Oil Country Tubular Goods from the Republic of Korea at 3, A-580-870, (July 10, 2014), available at http://enforcement.trade.gov/frn/ summary/korea-south/2014-16874-l.pdf (last visited Aug. 27, 2015) (“I & D Memo ”). In order to determine whether OCTG from Korea were sold in the United States at less than fair value, Commerce compared HYSCO’s constructed normal value to a constructed export price (“CEP”), 3 because HYSCO reported that it sold the subject merchandise to a wholly-owned subsidiary in the United States that then sold the merchandise to an unaffiliated customer. Decision Memorandum for the Preliminary Determination in the Less-Than Fair Value Investigation of Certain Oil Country Tubular Goods from the Republic of Korea at 15, 19, PD 276 (Feb. 14, 2014) (“Preliminary I & D Memo ”). NEXTEEL’s constructed normal value was compared to NEXTEEL’s export price 4 for certain sales that it made directly to unaffiliated customers, and CEP for sales made through an affiliated customer. See 1 & D Memo at 90.

In February 2014, Commerce issued a negative preliminary determination. Certain Oil Country Tubular Goods From the Republic of Korea: Negative Preliminary Determination of Sales at Less Than Fair Value, Negative Preliminary Determination of Critical Circumstances and Postponement of Final Determination, 79 Fed. Reg. 10,480 (Dep’t Commerce Feb. 25, 2014) (“Preliminary Determination ”). Commerce calculated weighted-average dumping margins of zero for both mandatory respondents. Id. at 10,481.

In July 2014, Commerce issued an affirmative final determination. Final Determination, 79 Fed.Reg. at 41,983. Commerce calculated a dumping margin of 9.89% for NEXTEEL and 15.75% for HYSCO. Id. at 41,984. Korean producers and exporters not individually examined, including Husteel, ILJIN, SeAH, and AJU Besteel, were assigned a margin of 12.82%, which was the weighted average of the mandatory respondents’ dumping margins. See id. The largest factor in the significant change in the dumping margin between the Preliminary Determination and the Final Determination was the profit figure used in the CV calculation. For NEXTEEL, Commerce preliminarily relied on the profit recorded in certain Korean OCTG producers’ financial statements, and for HYSCO, Commerce preliminarily used the profit HYSCO earned on its home market sales of non-OCTG pipe products. I & D Memo at 14. For the Final Determination, Commerce used the profit reflected in the financial statement of Tenaris S.A., a multinational corporation, to calculate CV profit for both mandatory respondents. Id. at 14, 16, The Tenaris financial statement was placed on the record after the Preliminary. Determination. See id. at 28-29.

The International Trade Commission reached an affirmative injury determination in September 2014. See Certain Oil Country Tubular Goods from India, Korea, the Philippines, Taiwan, Thailand, Turkey, Ukraine, and Vietnam, 79 Fed. Reg. 53,080 (ITC Sept. 5, 2014). Commerce issued the AD order effective September 10, 2014. See Certain Oil Country Tubular Goods From India, the Republic of Korea, Taiwan, the Republic of Turkey, and the Socialist Republic of Vietnam: Antidumping Duty Orders; and Certain Oil Country Tubular Goods From the Socialist Republic of Vietnam: Amended Final Determination of Sales at Less Than Fair Value, 79 Fed.Reg. 53,691 (Dep’t Commerce Sept. 10, 2014).

Korean producers NEXTEEL, HYSCO, Husteel, SeAH, AJU Besteel, and ILJIN, and domestic producers U.S. Steel and Maverick, challenge numerous aspects of Commerce’s Final Determination. Each issue will be discussed in turn.

JURISDICTION AND STANDARD OF REVIEW

The court has jurisdiction pursuant to 28 U.S.C. § 1581(c). The court will uphold Commerce’s final determination in an AD investigation, unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i).

DISCUSSION

I. Respondent Selection

A. Background

In the Initiation Notice, Commerce indicated that it would rely on U.S. Customs and Border Protection (“CBP”) data for U.S. imports of OCTG to select mandatory respondents in the event that Commerce determined that the number of known exporters or producers was “large.” Initiation Notice, 78 Fed.Reg. at 45,511. Commerce explained that it would release the CBP data shortly following the Initiation Notice and invited interested parties to comment regarding the CBP data and respondent selection. Id.

Whereas the AD petition listed ten Korean producers or exporters of OCTG, Petition at Ex. 1-5, PD 1-3 (July 2, 2013), the CBP data released by Commerce listed twenty-two producers or exporters. CBP Data, CD 16-17 (July 26, 2013). Of the twenty-two companies listed, several of the companies had almost identical names, suggesting that these firms were double counted, and issues with others cast doubt on their suitábility as respondents. 5 Id. The government states that even if some of the entries in the CBP data were redundant, there were at least twelve potential respondents, although the government maintains that twenty-two is the appropriate figure in determining the number of potential respondents. Def.’s Resp. in Opp’n to Mots, for J. upon the Administrative R. 67-70, ECF No. 144 (confidential version).

Commerce concluded that “[bjecause of the large number of known exporters or • producers involved in this investigation, and after careful consideration of [its] resources, ... it would not be practicable ... to examine all known exporters and producers of the subject merchandise as identified in the Petition and the CBP import data.” Respondent Selection Memo at 6. Rather than review each known exporter or producer, Commerce limited the mandatory respondents to the exporters or producers that accounted for the largest volume of imports of OCTG that reasonably could be examined, pursuant to 19 U.S.C. § 1677f-l(c)(2)(B). See id. at 7. Commerce selected HYSCO and NEXTEEL, as they were the two largest exporters of OCTG. 6 Id. at 8. Commerce indicated that it would consider requests to be treated as voluntary respondents at a future date. Id. at 9.

Husteel, SeAH, and ILJIN requested to be individually examined as voluntary respondents. Treatment of Voluntary Respondents Memorandum at 1, PD 194 (Dec. 30, 2013) (“Voluntary Respondent Memo”). On December 30, 2013, approximately four months after Commerce limit ed the number of mandatory respondents, Commerce determined that it could not examine any voluntary respondents “as this would be unduly burdensome to the Department, and inhibit the timely completion of this investigation.” Id. Commerce noted the complexities involved in its examination of the two mandatory respondents, the truncated timeline for investigations, the need to verify the responses of any additional respondents, its workload, including a number of AD and countervailing duty investigations on OCTG from other countries, and its limited resources as factors bearing on its decision. Id. at 5-7.

Husteel, SeAH, 7 and ILJIN argue that they should have been examined either as either mandatory respondents or voluntary respondents.

B. Mandatory Respondent Selection

Husteel argues that Commerce impermissibly interpreted the statute that authorizes Commerce to limit the number of mandatory respondents “[i]f it is not practicable to make individual weighted average dumping margin determinations [for each known exporter or producer of the subject merchandise] because of the large number of exporters or producers involved in the investigation.” 19 U.S.C. § 1677f-1(c)(2). It argues that Commerce improperly relied on an assessment of its own resource constraints in defining “large number.” Husteel Br. at 39-42. Husteel further contends that the number of exporters or producers involved in the investigation was not “large.” Id. at 42-43.

ILJIN repeats the same arguments made by Husteel, but emphasizes Commerce should have predicted based on the CBP import data and the requests to be reviewed that only a handful of companies were willing to cooperate in the investigation. ILJIN Br. at 19-21. According to ILJIN, Commerce should have considered the number of respondents it in fact was likely to review (i.e., the companies that had indicated they would cooperate) in determining whether it could individually examine each respondent. Id. at 20-21. IL-JIN additionally argues that Commerce acted contrary to law, because it did not examine a “reasonable number” of respondents. Id. at 22-23. ILJIN also contends that Commerce erred by failing to take into account evidence showing that ILJIN was the only Korean producer of seamless OCTG and that any margin based solely on welded OCTG would not be representative. See id. at 23-30.

i. Reliance on Resources

The general rule in AD cases is that Commerce “shall determine the individual weighted average dumping margin for each known exporter and producer of the subject merchandise.” 19 U.S.C. § 1677f-1(c)(1). The statute, however, provides an exception, which Commerce invoked in this case:

(2) Exception

If it is not practicable to make individual weighted average dumping margin determinations under paragraph (1) because of the large number of exporters or producers involved in th'e investigation or review, the administering authority may determine the weighted average dumping margins for a reasonable number of exporters or producers by limiting its examination to—

(A) a sample of exporters, producers, or types of products that is statistically valid based on the information available to the administering authority at the time of selection, or

(B) exporters and producers accounting for the largest volume of the subject merchandise from the exporting country that can be reasonably examined.

19 U.S.C. § 1677f — 1(c)(2). Husteel and ILJIN first argue that Commerce impermissibly determined whether there was a “large” number of potential respondents based upon its resource constraints. They cite several decisions of the court wherein Commerce was criticized for employing such reasoning. See Asahi Seiko Co. v. United States, 34 CIT 1443, 1449-50, 751 F.Supp.2d 1335, 1340-41 (2010) (concluding that Commerce had implicitly construed “large” to mean any number greater than three when Commerce stated in the issues and decision memorandum that “[b]ased upon our analysis of the workload required of this administrative review, we have determined that we can examine a maximum of three exporters/producers” and determining that this construction was unreasonable); Carpenter Tech. Corp. v. United States, 33 CIT 1721, 1726-29, 662 F.Supp.2d 1337, 1341-44 (2009) (concluding that Commerce had interpreted “large” to mean any number greater than two based upon Commerce’s explanation in the issues and decision memorandum that it could “examine a maximum of two exporters/producers” and holding that this interpretation was unreasonable); Zhejiang Native Produce & Animal By-Prods. Imp. & Exp. Corp. v. United States, 33 CIT 1125, 1129, 637 F.Supp.2d 1260, 1263-64 (2009) (rejecting Commerce’s conclusion that four was a large number and explaining that “[t]he statute focuses solely on the practicability of determining individual dumping margins based on the large number of exporters or producers” and thus “Commerce may not rely upon its workload caused by other ... proceedings in assessing whether the number of exporters or producers is ‘large’ ”). This argument lacks merit.

Commerce apparently took account of its limited resources and the workload caused by other proceedings in deciding to limit the number of mandatory respondents. For example, Commerce stated that “[i]n considering what constitutes a large number of exporters and producers as part of selecting respondents for an antidumping duty investigation, the Department carefully considers its resources, including its current and anticipated workload and deadlines coinciding with the proceeding in question.” Respondent Selection Memo at 6. Commerce also stated that although it ideally would examine all potential respondents, “in instances where [Commerce is] forced to limit [its] examination due to the large number of potential respondents relative to [its] resource constraints,” Commerce examines as many exporters or producers as it is able. Id. Although Commerce referenced its resource constraints a number of times in the Respondent Selection Memo, these references do not fatally undermine Commerce’s conclusion that there was a “large” number of exporters or producers involved in the investigation.

This case distinguishable in a number of material respects from Asahi, Carpenter, and Zhejiang. Unlike Commerce’s determinations in Asahi and Carpenter, Commerce’s determination here did not rest on an interpretation that any number greater than two or three is large. Rather, Commerce determined that a large number of potential respondents was involved in the investigation and then limited its examination to two. And the situation faced by Commerce in Zhejiang was materially different, in that the number of respondents initially involved in that case was four, the two respondents initially selected for review refused to cooperate, and the plaintiff was the only company still seeking review. 33 CIT at 1130, 637 F.Supp.2d at 1264. Thus, Commerce determined in that case that between one and four respondents was a large number. See id. Here, Commerce was determining whether twelve 8 constituted a large number of exporters or producers, which is a much larger number. The court recognizes that Zhejiang did state that “Commerce may not rely upon its workload caused by other anti-dumping proceedings in assessing whether the number of exporters or producers is ‘large,’ and thus deciding that individual determinations are impracticable.” Id. at 1129, 637 F.Supp.2d at 1263-64. The court urges Commerce to focus solely on the number of exporters or producers involved in the investigation or review, rather than its workload caused by other proceedings, in determining whether there is a large number of potential respondents. The statement in Zhejiang, however, should be read within its context. The very next sentence stated that “Commerce cannot rewrite the statute based on its staffing issues.” Id. at 1129, 637 F.Supp.2d at 1264. The problem in Zhejiang was that Commerce used its resource constraints to interpret the statute to mean that even numbers that appear to be objectively small, such as one or four, were defined as “large.” Commerce here has not written “large” completely out of the statute, and the court will not reject Commerce’s conclusion that twelve is a large enough number that examining each producer or exporter would be impracticable, solely because Commerce referenced its heavy workload.

ii. Whether Twelve is a Large Number

Husteel and ILJIN next argue that Commerce erred in concluding that there were a “large” number of respondents involved in the investigation. The statute does not define the term “large” and Commerce is afforded some discretion in interpreting that term. Cf. Carpenter, 33 CIT at 1727-28, 662 F.Supp.2d at 1342 (noting that Congress did not define the term “large number of exporters or producers involved in the [administrative proceedings]” and acknowledging that “the term might be seen as inherently ambiguous in some contexts”). The court has suggested that numbers ranging from three, see id. at 1726-29, 662 F.Supp.2d at 1341-44, to eight, see id. at 1730, 662 F.Supp.2d at 1344, do not constitute “large” numbers. The number of exporters or producers involved in this case, twelve, exceeds the number of potential respondents involved in the cases cited by Husteel and ILJIN. In addition to the fact that the number of potential respondents involved in this case is larger than the cases cited, the court notes that this case involves an investigation. As explained in greater detail re garding Commerce’s refusal to examine any voluntary respondents, the statutory deadlines for completing an investigation are shorter than the deadlines for completing a review, and Commerce is required to conduct a verification of respondents’ submissions. As a general matter, “Commerce has more work to do in less time” when conducting an investigation. Mem. in Opp’n to Pis.’ and Pl.-Intvnrs.’ Mot. for J. on the Agency R. Filed by Def.-Intvnr. United States Steel Corp. 91, ECF No. 149 (“U.S. Steel Resp.”). Although Commerce’s shifting resource allocations do not define “large,” “large” may mean something different in investigations. The court concludes that Commerce’s determination that there was a “large” number of known exporters or producers involved in this investigation was reasonable.

Husteel and ILJIN allude to fact that only five companies requested to be examined, and suggest that Commerce should have considered the fact that its investigation likely would have consisted of only those companies. The statute states that Commerce “shall determine the individual weighted average dumping margin for each known exporter and producer of the subject merchandise.” 19 U.S.C. § 1677f-1(c)(1). The Statement of Administrative Action indicates that Commerce’s practice is to attempt to calculate margins “for all producers and exporters of merchandise who are subject to an antidumping investigation.” Statement of Administrative Action Accompanying the Uruguay Round Agreements Act, H.R. Doc. No. 103-316, vol. 1, at 872 (1994), reprinted in 1994 U.S.C.C.A.N. 4040, 4200 (“SAA”). The statute does not limit Commerce’s duty to investigate only respondents that specifically ask to be reviewed. Furthermore, ILJIN’s apparent assumption that the other companies listed in the CBP data would not have cooperated in any investigation is based on nothing more than speculation. The court therefore rejects this contention. 9

Hi “Reasonable Number” of Respondents

ILJIN next cursorily argues that Commerce’s decision to limit the number of mandatory respondents to only two was unreasonable. See ILJIN Br. at 22-23. ILJIN contends that “if the exception can legally be invoked, it provides that the ‘administering authority may determine the weighted average dumping margins for a reasonable number of exporters or producers by limiting its examination to ... [the selected subset identified in subparts (A) and (B) ].’ ” Id. at 22-23 (alterations in original) (quoting 19 U.S.C. § 1677f-l(c)). According to ILJIN, two out of ten is not a “reasonable number.” Id. at 23. In support of this argument, ILJIN cites Zhejiang, Carpenter, and Asahi as establishing a minimum number of respondents that must be reviewed. This argument lacks merit.

First, ILJIN did not exhaust its administrative remedies on this issue. Nowhere in its case brief did ILJIN argue that two was not a “reasonable number” of respondents. See ILJIN Case Brief, PD 446 (June 8, 2014); Pakfood Pub. Co. v. United States, 34 CIT 1122, 1143-44, 724 F.Supp.2d 1327, 1349-50 (2010) (discussing general rule that a party must present all of its arguments in its case brief in order to exhaust its administrative remedies). Second, whether a certain number of mandatory respondents is “reasonable” in any particular case is likely to depend on the facts of that case, such as the subject merchandise at issue, the respondents chosen, the mandatory respondents’ share of the total volume of imports, and other factors. There is no magic number of respondents that must be chosen for the number to be “reasonable,” and the cases cited by ILJIN do not create any such bright line. None of those cases discussed whether the number of respondents selected was a “reasonable number” once the authority to limit the number of respondents was invoked properly. The court therefore rejects this argument.

iv. Representativeness

ILJIN also argues that Commerce failed to take account of information it submitted showing that the other potential respondents in the investigation, including the two respondents that were selected for individual examination, were' not representative of ILJIN. See ILJIN Br. at 23-31. ILJIN notes that it produces only seamless OCTG, whereas each of the other Korean companies produce only welded OCTG. Seamless OCTG requires different manufacturing processes. See ILJIN’s Comments on Respondent Selection at 2-5, PD 56 (Aug. 5, 2013). ILJIN submitted information to Commerce showing that because of the specialized nature of seamless OCTG, the sales price of seamless OCTG was significantly higher than the sales prices for welded OCTG. 10 Id. ILJIN contends that it is “fundamentally unfair to burden ILJIN’s sales of seamless OCTG with the margins calculated on much lower-priced welded OCTG.” ILJIN Br. at 24. This argument has merit.

“[A]n overriding purpose of Commerce’s administration of antidumping laws is to calculate dumping margins as accurately as possible.” Parkdale Int'l v. United States, 475 F.3d 1375, 1380 (Fed.Cir.2007) (citing Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed.Cir. 1990)). The statute expresses a general preference that each exporter or producer receive its own margin. See 19 U.S.C. § 1677f-l(c); see also Carpenter, 33 CIT at 1731, 662 F.Supp.2d at 1345 (construing that the statute should be construed such that “limiting the number of individually examined respondents is intended to be the exceptional circumstance, not the norm”). By individually examining each exporter or producer, Commerce bases dumping margins on each company’s own commercial behavior, which presumably supports the overall goal of calculating dumping margins as accurately as possibly. As explained, in certain circumstances, Commerce is authorized to limit its examination to a “reasonable number” of respondents by using “(A) a sample of exporters, producers, or types of products that is statistically valid based on the information available to the administering authority at the time of selection, or (B) exporters and producers accounting for the largest volume of the subject merchandise from the exporting country that can be reasonably examined.” 19 U.S.C. § 1677f-l(c)(2). It is not unreasonable to assume that the goals of these provisions are to capture a broadly representative sample of the export market, whether through the use of a statistically valid sample based on factors pertinent to the case or by the fact that capturing a large percentage of the imported merchandise generally will reflect the various commercial realities in the home market. This assumption is at the heart of ILJIN’s argument.

As explained, ILJIN submitted information to Commerce showing that it was the sole producer of seamless OCTG, which for a number of reasons markedly differs from welded OCTG. Petitioners submitted similar information to Commerce and argued that “the Department should select Ujin as a mandatory respondent to ensure that the investigation covers a representative sample of Korean OCTG producers.” Petitioners’ Comments on Respondent Selection at 5, PD 57 (Aug. 6, 2013). ILJIN noted that Commerce had the authority to consider differences in product type and that using the sampling methodology under § 1677f-1(c)(2)(A) would allow Commerce to select producers of seamless and welded OCTG. ILJIN’s Comments on Respondent Selection at 5; ILJIN Case Brief at 6-7. IL-JIN also argued that Commerce could satisfy both statutory provisions by selecting ILJIN under subsection (A) to ensure that producers of both kinds of OCTG were represented and then choosing the largest producers or exporters under subsection (B). ILJIN Case Brief at 7.

Commerce provided the following explanation in the Respondent Selection Memo for its choice of mandatory respondents:

[T]he Department has the statutory discretion to choose respondents by either sampling or selecting the exporters or producers that account for the largest volume of exports of subject merchandise. In selecting respondents in this antidumping duty investigation, the Department finds that, given its limited resources, it is most appropriate to select the exporters or producers accounting for the largest volume of the subject merchandise that can reasonably be examined, pursuant to section 777A(c)(2)(B) of the Act.

Respondent Selection Memo at 7. Regarding the arguments raised by ILJIN and the petitioners, Commerce stated in a footnote that

[w]ith respect to ... ILJIN’s argument that we should select it because it is allegedly the only Korean producer of seamless OCTG, and petitioners’ proposed respondent selection methodology, we note that none of these suggestions for respondent selection are pertinent to the factors that we normally consider in selecting respondents under the two methodologies (ie., choosing a statistically valid sample or selecting the largest volume exporters and producers) permitted by the statute.

Id. at 7-8 n. 49. Commerce also stated that “[wjhile petitioner argues that ILJIN’s sales would be more representative, the statute allows for selection based upon the largest exporters.” Id. at 8. Later in the proceedings, when Commerce declined to investigate any voluntary , respondents, Commerce explained that

[i]n making our determination regarding mandatory respondent selection, the Department already took into account IL-JIN’s argument that it was the only producer of seamless OCTG in Korea. The scope of this investigation covers both welded and seamless OCTG, and, thus, seamless OCTG is of the same class or kind as welded OCTG.

Voluntary Respondent Memo at 5 n. 31. The I & D Memo did not address ILJIN’s representativeness argument at all.

Commerce has a general duty to explain the basis for its decisions. NMB Sing. Ltd. v. United States, 557 F.3d 1316, 1319-20 (Fed.Cir.2009). This includes addressing relevant arguments made by interested parties. Id. Even when an agency has discretion, “[a]n agency ‘must cogently explain why it has exercised its discretion in a given manner.’” Changzhou Hawd Flooring Co. v. United States, 44 F.Supp.3d 1376, 1390 (CIT 2015) (quoting Motor Vehicle Mfrs. Ass’n of the U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 48, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983)). Commerce failed to provide adequate reasoning for refusing to examine ILJIN as a mandatory respondent.

As is apparent from the quoted passages, Commerce essentially ignored ILJIN’s arguments regarding whether its participation was required in order for the examined respondents to be representative of the Korean market and that dumping margins based on producers who manufacture only welded OCTG would be unfair to ILJIN, which produces only seamless OCTG. Commerce similarly ignored the argument by petitioners that examination of ILJIN was necessary to ensure that the experiences of Korean seamless OCTG producers were included in the investigation, leaving an important type of subject merchandise, which petitioners successfully sought to have included in the investigation, completely unexamined. Commerce’s reasoning appears to be little more than it has discretion in choosing between the respondent selection methodologies. That is insufficient. See id. Nowhere in the agency record is there evidence that it exercised that discretion in a lawful way. 11

In its brief before the court, the government cites to Mid Continent Nail Corp. v. United States, 949 F.Supp.2d 1247 (CIT 2013), as supporting Commerce’s conclusion. Def.’s Resp. in Opp’n to Mots, for J. upon the Administrative R. 72-74, ECF No. 146 (“Gov. Br.”). The court in that case noted that “[njothing in the language of [19 U.S.C. § 1677f-l(e)(2)(B) ] even hints that the exporters and producers selected for individual review must be ‘representative’ ” and that nothing in the SAA suggests that Commerce’s selection of respondents based on volume is constrained by concerns about representativeness. 949 F.Supp.2d at 1271-72. The court did suggest, however, that representativeness was a concern when employing the sampling method in § 1677Í-1 (c)(2)(A). Id. at 1272. The government’s reliance on this case is unavailing. First, the court in Mid Continent Nail determined that the plaintiff in that case had failed to exhaust its administrative remedies, and thus this claim was barred. Id. at 1263. Thus, the discussion cited by the government likely is nothing more than dicta. Second, the plaintiff in that case did not challenge the government’s decision to rely solely on § 1677f-l(c)(2)(B) as the appropriate method for choosing respondents. The court specifically stated that

[n]othing herein should be understood to suggest that Commerce’s discretion to choose between the two methodologies specified in 19 U.S.C. § 1677f-l(c)(2) is wholly unfettered, or that “representativeness” could never constrain Commerce’s ability to rely on 19 U.S.C. § -1677f — 1(c)(2)(B) or affect a determination as to whether a specific number of exporters and producers is “reasonable” given the facts of a particular case. Those issues are not presented here.

Id. at 1274 n. 25. This is exactly the situation presented here. ILJIN argued that Commerce’s selection of mandatory respondents failed to consider that the two producers of welded OCTG that were selected were not representative of producers of seamless OCTG such as ILJIN, and Commerce failed to deal with the issue. 12

Accordingly, the court remands this issue for reconsideration. 13 In making its decision on remand, Commerce must consider record evidence that is probative of the difference between welded and seamless OCTG, including costs and pricing.

C. Voluntary Respondent Selection

Even when Commerce lawfully limits the number of respondents selected as mandatory respondents, the statute contemplates that exporters or producers can still obtain their own margin as a voluntary respondent. 19 U.S.C. § 1677m(a) provides:

In any investigation ... or a review ... in which the administering authority has, under section 1677f — 1(c)(2) of this title ..., limited the number of exporters or producers examined, or determined a single country-wide rate, the administering authority shall establish ... an individual weighted average dumping margin for any exporter or producer not initially selected for individual examination under such sections who submits to the administering authority the information requested from exporters or producers selected for examination, if—

(1) such information is so submitted by the date specified—

(A) for exporters and producers that were initially selected for examination, [and] ...

(2) the number of exporters or producers who have submitted such infor mation is not so large that individual examination of such exporters or producers would be unduly burdensome and inhibit the timely completion of the investigation. 14

Commerce declined to accept any voluntary respondents, claiming that doing so would be unduly burdensome and inhibit the timely completion of the review.

Husteel and ILJIN argue that Commerce has not shown that examination of additional respondents would have been “unduly burdensome.” Husteel Br. at 44-49; ILJIN Br. at 32-33. They argue that Commerce failed to cite any burden that would result from investigating the additional respondents that is different from the typical burdens of a thorough investigation, which they claim is insufficient to create an “undue burden.” See Husteel Br. at 45-47; ILJIN Br. at 33-35. They note that only three companies asked to be voluntarily reviewed and also argue that the investigation of each company would have been relatively straight-forward. Husteel Br. 47-49; ILJIN Br. 33. ILJIN also argues that Commerce could have chosen to review just a single additional company, rather than all three, and reemphasizes that it should have been examined because it was the only producer of seamless OCTG, which has different costs and sells at a higher price than welded OCTG. ILJIN Br. at 34-35.

The government argues that Commerce properly considered its limited resources, current and anticipated workload, and the complexities of the investigation, and reasonably limited its investigation to only the two mandatory respondents. Gov. Br. at 75-78. The government and U.S. Steel note that the investigations into just the two mandatory respondents was extensive and complex and that investigating additional companies would have required additional verifications, which are mandatory in investigations. Id.; U.S. Steel Resp. at 89-91. The government and U.S. Steel additionally highlight the shorter statutory deadlines in investigations compared to reviews, explaining that “Commerce has more work to do in less time.” U.S. Steel Resp. at 91; see also Gov. Br. at 76-77. On the facts of this case, the court agrees with the government.

Husteel and ILJIN rely heavily on Grobest & I-Mei Industrial (Vietnam) Co. v. United States, 815 F.Supp.2d 1342 (CIT 2012) (“Grobest I ”), and Grobest & I-Mei Industrial (Vietnam) Co. v. United States, 853 F.Supp.2d 1352 (CIT 2012) (“Grobest II”). The Grobest plaintiff challenged Commerce’s decision to limit individual examinations in the administrative review to only the two mandatory respondents initially chosen. See Grobest I, 815 F.Supp.2d at 1360-61 & n. 25. In Grobest I, the court remanded Commerce’s refusal to accept the plaintiffs request for review as a voluntary respondent because Commerce had.unlawfully treated its decision to limit the number of mandatory respondents under 19 U.S.C. § 1677f-l(c) as dis- positive of the issue as to whether it needed to review any voluntary respondents. Id. at 1362-64. The court noted that the two distinct standards listed in § 1677f-1(c)(2) and § 1677m(a) require two separate determinations, and concluded that § 1677m(a) “sets a higher threshold of agency burden before the requirement of individual review can be avoided.” Id. at 1363.

On remand, Commerce again refused to examine the plaintiff as a voluntary respondent. The court rejected the plaintiffs claim that Commerce’s determination violated the unambiguous language of the statute pursuant to step one of the Chevron analysis. Grobest II, 853 F.Supp.2d at 1363. The court noted that “the statute conditions consideration of ‘a number so large’ on whether review of such a number of respondents would be unduly burdensome and inhibit the timely completion of the review” and thus concluded that the statute does not require the number of voluntary respondents to reach “some arbitrary threshold of largeness,” as such an interpretation would fail to consider the relative burdens that may be caused by reviewing any one respondent. Id. The court concluded, however, that Commerce had failed to show an undue burden. Id. at 1364. The court determined that

the facts that Commerce put forward to support that conclusion do not distinguish this case from the paradigmatic review of an antidumping or countervailing duty order. Rather, the burdens Commerce names in the Remand Results are the same burdens that occur in every review. In this regard, Commerce’s decision that the burden in this case is undue sets the bar for undue burden too low because it would make individual review of voluntary respondents in any typical antidumping or countervailing duty review unduly burdensome, and such a determination renders § 1677m(a) meaningless.

Id. at 1364-65 (footnote omitted).

Husteel and ILJIN assert that because § 1677m(a) sets a higher bar than § 1677f-l(c)(2), Commerce could not limit its review to solely the two mandatory respondents. They reason that cases interpreting § 1677f-l(c)(2) as requiring individual examination of numbers as large as eight, see, e.g., Carpenter, 33 CIT at 1730, 662 F.Supp.2d at 1344, set the baseline for the total number of respondents that Commerce must review. They also note that many of the burdens cited by Commerce in this case, including the need to issue supplemental questionnaires, issues regarding affiliation, unfamiliarity with the respondents, and high workloads throughout Commerce, reflect the “typical” burdens cited by Commerce in Grobest II and rejected by the court. Compare Grobest II, 853 F.Supp.2d at 1365 n. 12, with Voluntary Respondent Memo at 4-7. Accordingly, they argue Commerce failed to show that reviewing any and/or all of the three firms that requested voluntary status would be unduly burdensome.

The court agrees with the analysis in Grobest II that § 1677m(a) does not set an arbitrary threshold as to the number of exporters or producers that must submit a request for voluntary review before Commerce may decline to individually examine each such exporter or producer. The court also agrees with the implicit conclusion in Grobest II that Commerce may in some cases refuse to review any voluntary respondents. As the court noted in that case, the term “not so large” is defined in relation to the burden additional examinations would place on the agency and its ability to timely complete the investigation. 853 F.Supp.2d at 1363. The SAA also contemplates that in certain eases, Commerce may decline to analyze such re sponses. See SAA, H.R. Doc. No. 103-316, vol. 1, at 873, 1994 U.S.C.C.A.N. at 4201 (“Although Commerce ... will not discourage voluntary responses and will endeavor to investigate all firms that voluntarily provide timely responses in the form required, in certain cases (including cases involving the same product from multiple countries) where the number of exporters or producers is particularly high, Commerce may decline to analyze voluntary responses because it would be unduly burdensome and would preclude the completion of timely investigations or reviews”).

The court does not agree, however, with Husteel and ILJIN’s interpretation of Grobest that § 1677m(a)’s “higher threshold” means that cases interpreting “large” in § 1677f — 1(c)(2) essentially set a minimum number of total respondents that must be examined, either as mandatory or voluntary respondents. The court notes that the analysis in § 1677f — 1 (c) (2) should be made without considering the resources available to Commerce, whereas the concept of “undue burden” contained in § 1677m(a) is predicated on Commerce’s ability to complete the investigation on time, which would seem to invite consideration of Commerce’s resources. Additionally, if the court were to conclude that § 1677m(a) sets a bar higher than § 1677f-l(c)(2), in the manner suggested by Husteel and ILJIN, then the court’s cases interpreting § 1677f-l(c)(2) essentially would set a baseline as to the number of respondents that Commerce must review in each case, which appears contrary to other parts of the analysis in Grobest II.

The court understands the problem in Grobest to be a concern that Commerce was interpreting § 1677m(a) in a manner that rendered that provision a nullity. See Grobest I, 815 F.Supp.2d at 1362 (concluding that Commerce’s interpretation “would mean that § 1677m(a) review of voluntary respondents is already curtailed once a § 1677f-l(c)(2) decision to limit the number of respondents is made” and would render § 1677m(a) “meaningless”); Grobest II, 853 F.Supp.2d at 1365 (holding that Commerce’s failure to show that the burden of reviewing a voluntary respondent would exceed that presented in a typical review rendered § 1677m(a) “meaningless” and thus its decision was an abuse of discretion). Commerce initially had treated its decision to limit the number of mandatory respondents as allowing it to ignore voluntary respondent requests, and then on remand relied on burdens that are present in almost every single case to justify its decision to limit the review to only two respondents. Thus, whether by accepting Commerce’s interpretation of the statute or its explanation of its burdens, respondents that were not chosen as mandatory respondents would have no hope of receiving an individual margin via § 1677m(a), which defeats the congressional intent reflected in the inclusion of that provision in the statute.

Viewed in this context, the “higher threshold” referenced in Grobest is better understood as a requirement that Commerce rely on something other than its initial decision to limit the number of mandatory respondents when analyzing requests for voluntary respondents. Commerce has the authority to limit the number of mandatory respondents to a “reasonable number.” 19 U.S.C. § 1677f-1(c)(2). • Once Commerce does that, it must show that it actually would be burdened by individually examining exporters or producers that request to be treated as voluntary respondents. 19 U.S.C. § 1677m(a). It cannot simply rely on the fact that it already chose to limit the number of respondents to a “reasonable number” pursuant to § 1677f-l(c)(2).' Commerce in this ease did not simply rely on the fact that it limited the number of man datory respondents pursuant to § 1677f-1(c)(2) in declining to review any voluntary respondents. Rather, it gave specific reasons for why examining any additional respondents “would be unduly burdensome and inhibit the timely completion of the investigation.” 19 U.S.C. § 1677m(a). The court therefore rejects the arguments that any baselines supposedly created in cases interpreting § 1677f-l(c)(2) should be transported into the 19 U.S.C. § 1677m(a) analysis on the basis that § 1677m(a) requires a “higher threshold” before Commerce may refuse to perform an individual examination.

The court’s understanding of the problem addressed by Grobest, namely the risk that Commerce effectively was eliminating 19 U.S.C. § 1677m(a) from the statute (either as a legal matter or a practical matter), similarly informs its analysis of the argument that the burdens cited by the agency in this case largely mirror the burdens rejected in Grobest II. Had the court accepted the agency’s arguments in Grobest II,' the standard for declining to review voluntary respondents would have been so low that Commerce would be able to justify its refusal to consider voluntary requests in nearly every single case. That is not the case here. Although many of the burdens cited by Commerce in this case mirror the burdens cited in Grobest II, there are two key distinctions.

First, this case involves an investigation. In investigations, Commerce’s statutory deadlines for completing the administrative proceedings are shortened. Compare 19 U.S.C. §§ 1673b, 1673d, with 19 U.S.C. § 1675. Commerce must initially familiarize itself with the product and respondents, and verification of all information relied upon is required, whereas verification in reviews is needed only under certain circumstances. 19 U.S.C. § 1677m(i). Commerce noted that accepting additional respondents would require additional on-site verifications in Korea and possibly the United States. See Voluntary Respondent Memo at 7. As Commerce explained, “[i]n investigations, we have less time in which to complete more work when considering the vast quantity of previously unknown information submitted to us.” Id. at 5. Thus, the burden placed on Commerce in this case is not typical of every administrative proceeding, although it may be typical of many investigations.

Second, this case was part of a number of investigations concurrently initiated regarding OCTG. Commerce noted that “the Department is currently handling 11 concurrent AD and CVD investigations on OCTG from various countries.” Id. at 6. The SAA specifically notes that

[although Commerce ... will not discourage voluntary responses and will endeavor to investigate all firms that voluntarily provide timely responses in the form required, in certain cases (including cases involving the same product from multiple countries) where the number of exporters or producers is particularly high, Commerce may decline to analyze voluntary responses because it would be unduly burdensome and would preclude the completion of timely investigations or reviews.

H.R. Doc. No. 103-316, vol. 1, at 873, 1994 U.S.C.C.A.N. at 4201 (emphasis added). The fact that Commerce was handling numerous OCTG investigations is a pertinent factor the court takes into consideration. Although Commerce limited the number of respondents examined in each case, the total number of respondents examined was large. There is no indication that Commerce faced a similar situation in Grobest.

Because of the concurrent investigations into the same product and the fact that Commerce is required to do more work in less time when conducting such investigations, Commerce has shown that the bur den of reviewing a voluntary respondent in this case would exceed the typical burden Commerce faces in other administrative proceedings. On the facts of this case, Commerce’s determination that it would be unduly burdensome to examine any additional respondents was supported by substantial evidence, and was otherwise in accordance with law. See Grobest II, 853 F.Supp.2d at 1365 (“When Commerce can show that the burden of reviewing a voluntary respondent would exceed that presented in the typical antidumping or countervailing duty review, the court will not second 'guess Commerce’s decision on how to allocate its resources.”). 15

II. Constructed Value Profit

Plaintiffs argue that Commerce committed a multitude of errors regarding Commerce’s calculation of CV profit, which can be distilled down to two general arguments. 16 First, Commerce should not have used the financial statement of Tenaris to calculate CV profit. See HYSCO Br. at 12-46; NEXTEEL Br. at 13-44; Husteel Br. at 16-33, 36-38. Second, assuming that Commerce could use Tenaris’s financial statement for the purposes of CV profit, Commerce erred in failing to apply a profit cap. See HYSCO Br. at 46-50; NEXTEEL Br. at 44 — 49; Husteel Br. at 33-36. These arguments have merit.

When using constructed value to calculate the normal value, the constructed value is to include “the actual amounts incurred and realized by the specific exporter or producer being examined ... for selling, general, and administrative expenses, and for profits, in connection with the production and sale of a foreign like product, in the ordinary course of trade, for consumption in the foreign country.” 19 U.S.C. § 1677b(e)(2)(A). If such data is unavailable, however, Commerce must resort to one of three alternatives for calculating an appropriate amount for selling, general, and administrative expenses, and profits:

(i) the actual amounts incurred and realized by the specific exporter or producer being examined in the investigation or review for selling, general, and administrative expenses, and for profits, in connection with the production and sale, for consumption in the foreign country, of merchandise that is in the same general category of products as the subject merchandise,

(ii) the weighted average of the actual amounts incurred and realized by exporters or producers that are subject to the investigation or review (other than the exporter or producer described in clause (i)) for selling, general, and administrative expenses, and for profits, in connection with the production and sale of a foreign like product, in the ordinary course of trade, for consumption in the foreign country, or

(Hi) the amounts incurred and realized for selling, general, and administrative expenses, and for profits, based on any other reasonable method, except that the amount allowed for profit may not exceed the amount normally realized by exporters or producers (other than the exporter or producer described in clause (i)) in connection with the sale, for consumption in the foreign country, of merchandise that is in the same general category of products as the subject merchandise, [i.e., what is commonly referred to as the “profit cap.”]

19 U.S.C. § 1677b(e)(2)(B). The court will refer to these alternatives as “alternative (i),” “alternative (ii),” and “alternative (iii),” respectively. In this case, Commerce determined that the data to calculate a profit figure under § 1677b(e)(2)(A) was unavailable and therefore it had to rely on one of the alternatives listed in § 1677b(e)(2)(B). 17 I & D Memo at 14.

For the Preliminary Determination, Commerce considered three possible options for CV profit: “[ (1) ] the profit reflected in the audited financial statements for seven Korean OCTG producers, [ (2) ] the profit earned by HYSCO on its home market sales of non-OCTG pipe products, and [ (3) ] the profit for Tenaris, SA (Tenaris), an Argentinian global producer and seller of OCTG,” as described in a research paper prepared by a student at the University of Iowa School of Management. Preliminary I & D Memo at 22. Commerce noted that “all three options have their limitations.” Id. For the profit on HYSCO’s home market sales of non-OCTG pipes, Commerce noted that this profit “reflect[ed] the profit on pipe products typically used in the construction industry, as opposed to the OCTG products used in the specialized oil and gas industry.” Id. “Likewise, the profit reflected in the Korean OCTG producers’ financial statements reflect the profits on the same non-OCTG pipe products, as well as the profits on OCTG sales predominantly to the United States.” Id. Regarding the Tenaris profit information, Commerce explained that although the information reflected predominantly OCTG sales, “it represents neither production nor sales in the market under consideration” and “is based on a research paper containing a disclaimer statement regarding its accuracy.” Id.

After considering the relative strengths and weaknesses of the various profit sources, Commerce preliminarily decided to base HYSCO’s CV profit on HYSCO’s profit on home market sales of non-OCTG pipe, pursuant to alternative (i). Id. For NEXTEEL, Commerce preliminarily decided to base CV profit on the profit earned by six Korean OCTG producers that earned a profit, pursuant to alternative (iii). Id. Commerce noted that “after the preliminary determination, we intend to continue to explore other possible options for CV profit for both respondents.” Id.

Following the Preliminary Determination, Commerce issued a supplemental questionnaire to NEXTEEL, requesting a breakdown of its costs and sales figures by product type (e.g., standard pipe, line pipe, OCTG) and by country to which it sold its products (e.g., U.S., Korea, Canada). See NEXTEEL’s Third Suppl. Section D Questionnaire Resp., CD 264 (Mar. 6, 2014). On March 21, 2014, U.S. Steel submitted a large amount of new factual information under 19 C.F.R. § 351.301(c)(l)(v), purporting to “rebut, clarify, or correct” evidence that was submitted by NEXTEEL in response to Commerce’s questionnaire. See U.S. Steel’s Comments re: NEXTEEL’s Third Suppl. Section D Questionnaire Resp., CD 303 (Mar. 21, 2014); U.S. Steel Resp. to Obj. of NEXTEEL at 1-2 & n. 1, PD 366 (Apr. 2, 2014). Included in U.S. Steel’s submission was Tenaris’s 2012 financial statement. See id. at Ex. P. NEXTEEL promptly requested that Commerce reject the information as untimely on the grounds that the information did not rebut, clarify, or correct the information contained in NEXTEEL’s response. NEXTEEL’s Req. to Reject Untimely New Factual Information at 1-2, PD 354 (Mar. 27, 2014) (“Req. to Reject Untimely Information”).

For the Final Determination, Commerce relied on the profit contained in Tenaris’s 2012 financial statement to calculate CV profit for both NEXTEEL and HYSCO pursuant to alternative (iii). I & D Memo at 14. Commerce rejected NEXTEEL’s claim that the information was untimely new factual information rather than rebuttal information. Id. at 29. Commerce concluded that U.S. Steel’s submission was rebuttal evidence, because NEXTEEL’s data could be used for purposes of calculating CV profit, and the information submitted by U.S. Steel was for the same purpose. Id. Commerce also explained that it has discretion to relax its regulations regarding the timely submission of information as long as parties are not substantially prejudiced, and it concluded that there was no prejudice because “NEXTEEL and HYSCO had an opportunity to submit rebuttal information ... had they chosen to do so.” Id. at 29-30.

Commerce determined that it could not rely upon alternative (i) for HYSCO, as it had in the Preliminary Determination, because HYSCO’s non-OCTG pipe products, such as line pipe and standard pipe, did not fall within the “same general category of products” as required to apply alternative (i). See I & D Memo at 18-19. Commerce highlighted the fact that OCTG are used in down-hole applications requiring that they withstand harsh conditions and are sold to the oil and gas exploration industry, which had seen an uptick in activity and demand. Id. at 17-19. Line pipe and standard pipe, however, are not used in down-hole applications, and the Korean producers sold their non-OCTG pipe products primarily to the Korean construction industry, which generally is unable and unwilling to pay the price premium paid in the oil and gas industry and which had seen sluggish activity during the period of investigation (“POI”). Id. at 17-18. Commerce also noted that OCTG require different grades of steel, are subjected to different testing and certification requirements, and are generally connected in ways that are different from non-OCTG products. See id. Commerce therefore had to resort to alternative (iii) for calculating a CV profit for both mandatory respondents. 18

In considering the various alternatives for calculating CV profit pursuant to alternative (iii), Commerce determined that the profit reflected in Tenaris’s financial statement represented the best information available. See id. at 23. Commerce rejected the respondents’ arguments that it should rely on the profit reflected in the financial statements of the various Korean OCTG producers, because the majority of their sales were of non-OCTG pipe outside of the same general category of products and the sales of OCTG imbedded in those statements were primarily the allegedly dumped sales,to the United States. 19 See id. at 20. Commercé explained that “[a]s OCTG is a very specialized premium product used exclusively in the oil and gas exploration industry with significant quality differences, different end uses, different end customers, and different demand patterns than those of non-OCTG pipe, it is important that we rely on a source that closely reflects such product.” Id. at 20-21 (footnote omitted). Commerce preferred the financial statement of Tenaris, because its sales consisted primarily of OCTG and the majority of its OCTG sales were to non-U.S. customers. Id. at 19, 21. Commerce further reasoned that “[b]e-cause Tenaris is an OCTG producer that sells OCTG in significant quantities, and in virtually every market in which OCTG is sold, we find its average profit experience is representative of sales of OCTG across a broad range of different geographic markets.” Id. at 21.

Commerce also determined that it was unable to calculate and apply a profit cap under alternative (iii), because Commerce did “not have home market profit data for other exporters and producers in Korea of the same general category of products.” Id. Whereas the six Korean OCTG producers used to calculate NEXTEEL’s CV profit for the Preliminary Determination had an average profit margin of 5.30% and the revised CV profit rates calculated by the petitioners for the petition were between 7.19% and 7.22%, Tenaris’s profit .rate was 26.11%. Compare Preliminary 'Constructed Value Calculation Adjustments for NEXTEEL at 2-3, CD 234 (Feb. 14, 2014), and Petitioner’s Resp. to July 8, 2013 Questionnaire re: Volume IV of the Petition at Ex. IV-34, Attach. Suppl. F, PD 14-16 (July 12, 2013), with I & D Memo at 7.

B. Use of Tenaris’s Financial Statement Under Alternative (iii)

HYSCO, NEXTEEL, and Husteel argue that Commerce’s use of Tenaris’s profit to calculate CV profit was unsupported by substantial evidence and unlawful. They contend that the profit data used by Commerce was untimely and should have been rejected. . See HYSCO Br. at 43-46; NEXTEEL Br. at 44; Husteel Br. at 11, 18 n. 5. They further contend that even if it was properly allowed on the record, Commerce should have used either the profit earned by the mandatory respondents’ on their home market sales of OCTG and/or non-OCTG pipe products or the average profit earned by the Korean OCTG producers. They assert that Commerce’s reasoning for declining to use this data, namely that the line pipe and standard pipe sold by the Korean producers in the Korean market were not in the same general category of products as OCTG, was unsupported by substantial evidence and contrary to prior Commerce decisions. See HYSCO Br. at 16-28; NEXTEEL Br. at 17-29; Husteel Br. at 20-24. They also argue that Commerce was required to use this data, which was based on production and sales in Korea, over the Tenaris profit data, which did not reflect production or sales in Korea. See HYSCO Br. at 12-15, 28-30; NEXTEEL Br. at 13-17, 29-31; Husteel Br. at 25-29. HYSCO, NEXTEEL, and Husteel further highlight certain features of Tenaris’s OCTG products and business operations that distinguish it from the Korean OCTG producers and that render Tenaris’s profit rate aberrational and unrepresentative of what the Korean respondents could expect in selling to the Korean market; according to plaintiffs, the profit earned on sales of pipe products in Korea by Korean producers is more representative of the respondents’ commercial experiences. HYSCO Br. at 30-43; NEXTEEL Br. at 31-44; Husteel Br. at 29-33.

i Timeliness of the Tenaris Financial Statement

Commerce’s regulations provide that a party may submit “factual information to rebut, clarify, or correct factual information contained in [a supplemental] questionnaire response” within ten days. 19 C.F.R. § 351.301(c)(l)(v). Thus, if U.S. Steel’s information rebutted, clarified, or corrected factual information contained in NEXTEEL’s questionnaire response, it was filed on time. If- U.S. Steel’s factual information did not fall within this category of rebuttal information, however, it should have been considered factual information not otherwise specifically accounted for in § 351.301(c), and it should have been filed at least thirty days before the Preliminary Determination. See 19 C.F.R. § 351.301(c)(5).

The government and petitioners assert that Commerce reasonably concluded that the factual information submitted by U.S. Steel, including the Tenaris financial statement, rebutted clarified, or corrected information submitted by NEXTEEL in its Third Supplemental Section D Questionnaire Response. Gov. Br. at 55-56; U.S. Steel Resp. at 64-66; Def.-Intvnr. Maverick Tube Corp.’s Resp. to Pis.’ Brs. in Supp. of Their Mots, for J. upon the Agency R. 43-44, ECF No. 153 (“Maverick Resp.”). They contend that the data contained in NEXTEEL’s response could be used to calculate a CV profit margin and they point to Commerce’s statement in the Preliminary I & D Memo that it would continue to explore other possible options for CV profit as supporting U.S. Steel’s belief that the supplemental questionnaire was aimed at obtaining information to be used to calculate CV profit. Gov. Br. at 55-56; U.S. Steel Resp. at 64-66. Because the information contained in NEXTEEL’s response was pertinent to CV profit, the information submitted by U.S. Steel was properly considered rebuttal evidence, as this information was also pertinent to calculating a CV profit margin. Gov. Br. at 55-56; U.S. Steel Resp. at 64-66; Maverick Resp. at 43-44. The court disagrees.

The regulations do- not define “factual information to rebut, clarify, or correct,” and thus Commerce’s interpretation is given deference as long as it is reasonable. See Baroque Timber Indus. (Zhongshan) Co. v. United States, 925 F.Supp.2d 1332, 1350 (CIT 2013). “Rebuttal evidence” is generally understood to be “evidence offered to disprove or contradict the evidence presented by an opposing party.” Black’s Law Dictionary (10th ed. 2014). The information submitted by U.S. Steel does not appear to satisfy this general understanding. NEXTEEL was asked to break down its costs and sales by country of sale and product type. Little if anything in U.S. Steel’s factual submission, and especially the evidence in Tenaris’s 2012 financial statement, disproves or contradicts NEXTEEL’s answers to those questions. Rather, U.S. Steel’s submission constituted a substitute data source that Commerce could use to calculate CV profit.

That such evidence should not be considered rebuttal evidence is supported by Commerce’s treatment of similar data in other cases. In the non-market economy (“NME”) context, Commerce must use surrogates to value the respondent’s factors of production, including an amount for profit. See 19 U.S.C. § 1677b(c)(l). The use of a surrogate profit margin in the NME context and a CV profit margin based on another company’s profit serve a very similar purpose, namely to calculate an amount for profit that the respondent could have been expected to earn if it had reliable home market sales data. In the NME context, however, substitute surrogate information is not considered rebuttal evidence. Commerce’s regulations specifically provide that “[a]n interested party may not submit additional, previously absent-from-the-record alternative surrogate value information” as rebuttal information in order to value factors of production. 19 C.F.R. § 351.301(c)(3)(iv); see also Definition of Factual Information and Time Limits for Submission of Factual Information, 78 Fed.Reg. 21,246, 21,248 (Dep’t Commerce Apr. 10, 2013) (“Definition of Factual Information ”) (“We also note that all interested parties may submit factual information to rebut, clarify, or correct factual information to value factors, as long as that information is submitted solely for rebuttal and not for purposes of establishing new surrogate values.” (emphasis added)). Prior to Commerce explicitly including this limitation on rebuttal evidence in the regulations, Commerce’s position had been that the regulation for submitting rebuttal evidence did not allow a party to submit substitute surrogate value information. See Baroque Timber, 925 F.Supp.2d at 1350. The court upheld this interpretation, explaining that “Commerce’s interpretation is ... consistent with the purpose of the subsection, which is to respond to factual information that has been placed on the record, not to expand the scope of the record.” Id. The court also explained that excluding new surrogate value data “prevents Commerce from facing a scenario in which either a party has no opportunity to rebut, clarify, or correct new surrogate values submitted in a rebuttal, or Commerce must accede to rolling rebuttals while also complying with the statutory deadlines for completing investigations and reviews.” Id. Commerce mentioned similar concerns when it revised its regulations to explicitly prevent new surrogate value information from being submitted as rebuttal information. See Definition of Factual Information, 78 Fed.Reg. at 21,248.

The same concerns and reasoning apply in this case. The Tenaris profit data did not cast doubt on NEXTEEL’s answers. Rather, U.S. Steel offered a new alternative for valuing NEXTEEL’s profit. Commerce’s determination that this constituted rebuttal information does not accord with the general understanding of “rebuttal evidence,” is contradicted by its regulations and practice when facing similar situations, and effectively means there is no distinction between rebuttal evidence and any new factual information. The court therefore holds that Commerce’s determination that Tenaris’s financial statement constituted rebuttal evidence was unreasonable.

The court’s conclusion is bolstered by the fact that treating the submission of substitute CV profit valuation information as rebuttal evidence is also inconsistent with the limits Commerce places on responding to rebuttal evidence. Under Commerce’s regulations, only the party that submitted the original information may respond to the rebuttal information. 19 C.F.R. § 351.301(c)(l)(v). The submission of Tenaris’s financial statement as a possible CV profit source, however, was relevant to all Korean producers, and fairness suggests that all parties should have an opportunity to respond to this kind of information. When rebuttal information is limited to submitting data that actually rebuts, clarifies, or corrects data submitted by a party, the original submitter will be in the best position to respond to the rebuttal evidence, and this fairness concern is mitigated. And if information such as the Tenaris financial statement is properly treated as factual information covered by § 351.301(c)(5), all parties are given a chance to respond to it. See 19 C.F.R. § 351.301(c)(5).

Furthermore, the court notes that earlier in the proceedings, U.S. Steel treated extremely similar information as falling under § 351.301(c)(5). Prior to the Preliminary Determination, U.S. Steel submitted a research paper regarding Tenaris, which included a profit margin, to be used as a possible source for CV profit data. U.S. Steel’s Jan. 16, 2014 Submission of Factual Information at Ex. J, PD 227 (Jan. 16, 2014). U.S. Steel submitted this information pursuant to § 351.301(c)(5) as information not covered by the other provisions of the regulation, including the provision for submitting rebuttal information. Id. at 1-2. U.S. Steel did so despite that fact that NEXTEEL and HYSCO both had submitted initial and supplemental Section D questionnaire responses containing information that Commerce could use to calculate a CV profit rate. NEXTEEL’s Sections C-D Questionnaire Resp., PD 152-153 (Nov. 5, 2013); NEXTEEL’s Suppl. Section D Questionnaire Resp., PD 189 (Dec. 24, 2013); HYSCO’s Sections C-D Questionnaire Resp., PD 154-157 (Nov. 5, 2013); HYSCO’s Suppl. Sections A, C & D Questionnaire Resp., PD 206-208 (Jan. 6, 2014); see also Enforcement and Compliance Antidumping Manual, Ch. 4 at 7-8, (Mar. 16, 2015), available at http://enforcement. trade.gov/admanual/2015/Chapter% 2004% 20Questionnaires.pdf (last visited Aug. 27, 2015) (“In market economy cases we request a response to section D if CV is, or is likely to be, used as [normal value] and/or if we decide to investigate whether foreign market sales are made at prices below the COP.”). U.S. Steel’s actions lend further confirmation of the court’s conclusion.

The court holds that Commerce unreasonably concluded that Tenaris’s financial statement constituted evidence rebutting, clarifying, or correcting the information NEXTEEL submitted in its supplemental questionnaire response, and under the applicable regulation this evidence should have been rejected as untimely.

ii. Prejudice

The government and petitioners argue that even if the evidence submitted by U.S. Steel were untimély, plaintiffs did not show that they were substantially prejudiced by Commerce allowing the information to remain on the record. Gov. Br. at 57-59; U.S. Steel Resp. at 67-70; Maverick Resp. at 45. They contend that respondents were not substantially prejudiced because they were on notice that Commerce would seek additional CV information, they had an opportunity to submit further rebuttal evidence but chose not to do so, and they had an opportunity to argue against the use of Tenaris’s financial statement in their case briefs. Gov. Br. at 57-59; U.S. Steel Resp. at 67-70; Maverick Resp. at 45. The court rejects this contention.

The fact that Commerce indicated that it intended to explore other possible options for CV profit says nothing as to whether the Tenaris financial statement was properly placed on the record and/or whether plaintiffs were prejudiced by Commerce allowing the submission in violation of its regulations. Commerce did not invite parties to submit alternative CV profit sources. Had it done so, Commerce’s proclamation that it intended to explore alternative CV profit sources might have some relevance.

The court also rejects the contention that plaintiffs had an adequate opportunity to submit additional evidence to either undercut or serve as an alternative to the Tenaris data. First, Commerce’s regulation only permits the original submitter of the information sought to be “rebutted,” in this instance NEXTEEL, to respond to rebuttal evidence. 19 C.F.R. § 351.301(c)(l)(v). HYSCO and the other plaintiffs did not have such an opportunity. Furthermore, NEXTEEL only had seven days to offer any such evidence. Id, Expecting NEXTEEL to adequately respond to the amount of information U.S. Steel filed within a week is unreasonable, especially when that information did not directly pertain to NEXTEEL’s own data.

Apparently recognizing the obvious shortcomings of this supposed opportunity to submit rebuttal information pursuant to 19 C.F.R. § 351.301(e)(l)(v), the government and petitioners contend that plaintiffs could and should have asked Commerce for an opportunity to place additional information on the record and/or for an extension of time to submit any additional evidence it wished to present. Gov. Br. at 57-59; U.S. Steel Resp. at 68-70; Maverick Resp. at 45. 19 C.F.R. § 351.301(a) states that Commerce “may ... provide additional opportunities to submit factual information.” And 19 C.F.R. § 351.302(b) provides that Commerce “may, for good cause, extend any time limit established” in Commerce’s regulations, unless precluded by statute. Here, theoretical options, however, are insufficient to defeat plaintiffs’ objection.

First, the court notes that these are the avenues that U.S. Steel should have pursued if it wanted to submit its new factual information, including the Tenaris financial statement. Commerce’s decision to allow this untimely evidence onto the record should not force parties that had complied with Commerce’s deadlines to seek discretionary relief from Commerce to file additional evidence to rebut the untimely evidence. Second, and more importantly, it was not clear whether plaintiffs needed to avail themselves of this option. On March 27, 2014, less than a week after U.S. Steel filed the untimely submission, NEXTEEL filed a request with Commerce that the information be rejected. Req. to Reject Untimely Information at 1-2. Commerce failed to respond to this request until the Final Determination. As explained, the information should have been rejected, but had a request for an extension of time been made by U.S. Steel and accepted, and had the information been submitted and accepted under the proper provision, 19 C.F.R. § 351.301(c)(5), Commerce would have issued a schedule providing deadlines for the submission of information to rebut, clarify, or correct it. See 19 C.F.R. § 351.301(c)(5)(ii). Under this scenario, all parties would have been notified that the information was on the record and would have known that they needed to respond to it. In this case, however, because the information was improperly submitted as rebuttal evidence and Commerce failed to rule on the request to reject it, plaintiffs were left to guess whether Commerce would accept or reject the information and speculate as to how Commerce would use the evidence (for example, whether Commerce would consider it solely to cast doubt on the use of NEXTEEL’s data for CV profit, or whether Commerce would consider it as a substitute surrogate for CV profit). By waiting until the Final Determination to rule on whether the evidence was properly submitted, Commerce placed plaintiffs in a difficult and undesirable situation, where the scope of the record was unclear. If Commerce had made clear that it considered the information timely filed earlier in the proceedings, the government and petitioners’ argument might have carried more weight.

Finally, the fact that plaintiffs were able to comment on the Tenaris financial statement does not eliminate the prejudice they suffered by Commerce allowing the information onto the record. Plaintiffs did not have a sufficient opportunity to submit evidence that would have either undermined the information contained in U.S. Steel’s submission or acted as an alternative CV profit source. The arguments that plaintiffs could make in their case briefs thus were limited. Given the proper notice and opportunity to respond to the information, plaintiffs could have conducted a more robust attack on its suitability to serve as the CV profit source in this case.

This case is distinguishable from the cases cited by the government in support of its contention that there was no prejudice. In the cases cited by the government, there was a technical deficiency in the proceedings, but all parties had a full opportunity to respond. See Am. Farm Lines v. Black Ball Freight Serv., 397 U.S. 532, 537-38, 90 S.Ct. 1288, 25 L.Ed.2d 547 (1970) (application for temporary operating authority submitted to Interstate Commerce Commission failed to contain certain specific pieces of information, but parties opposing grant of authority able to make precise and informed objections to the application); PAM, S.p.A. v. United States, 463 F.3d 1345, 1346-47, 1349 (Fed.Cir. 2006) (failure to serve foreign producer with request for review, but producer received actual and constructive notice of the proceedings and received multiple extensions of time during the proceedings to respond); Kemira Fibres Oy v. United States, 61 F.3d 866, 871, 875 (Fed.Cir.1995) (Commerce failed to follow regulations in seeking comment from domestic producers as to whether order should be revoked, but foreign producer not prejudiced by delay except to extent that dumping order remained in effect). In those cases, the complaining parties also asserted that these technical errors rendered the entirety of the agency’s actions void, thus leaving the agency unable to act. See Am. Farm Lines, 397 U.S. at 536, 90 S.Ct. 1288; PAM, S.p.A., 463 F.3d at 1347; Kemira Fibres, 61 F.3d at 871.

In this case, the error affected a key component of the respondents’ dumping margins. Tenaris’s profit margin of 26.11% was far greater than the Korean profit margins relied upon by Commerce in the Preliminary Determination, which averaged 5.30%, and was the main factor in the increase in the respondents’ dumping margins from zero in the Preliminary Determination to 17.75% and 9.89% in the Final Determination. As explained, plaintiffs were prejudiced because their need and ability to respond to the untimely filed information was murky at best. Additionally, the public “harm” that would result from enforcing Commerce’s regulations seems negligible. Commerce’s ability to remedy any unfair trade practices would not be impeded. Rather, Commerce simply would have been left with fewer options for the purposes of calculating a CV profit. And although petitioners obviously would prefer that the Tenaris data be used because it would result in higher dumping margins, it appears that petitioners could have submitted the same information by the regulatory deadlines, and thus the prejudice to them caused by enforcing Commerce’s deadlines for submitting factual information would be of their own making. 20

In conclusion, the court determines that this was not a simple technical violation that can be overlooked, but rather plaintiffs were substantially prejudiced by Commerce’s acceptance and use of U.S. Steel’s untimely submitted new factual information. On remand, Commerce may simply remove this information from the record and reconsider its CV profit determination based on the information that was submitted in accordance with the regulatory deadlines. Alternatively, Commerce must determine if and how, at this late date, the prejudice caused by accepting the Tenaris financial statement in violation of the regulations can be rectified.

iii. Commerce’s Choice of Tenaris’s Data Over the Korean Producers’ Data

Because the court is remanding for Commerce to either remove the Tenaris financial statement or to otherwise counter the prejudice to plaintiffs, the court deems it unnecessary to decide the bulk of the other arguments raised by plaintiffs regarding why the various sources of Korean data should have been used instead of the Tenaris data. These arguments may be rendered moot following remand, and many of the arguments presented in the briefs likely would have been made more forcefully if plaintiffs had been given a fully adequate' opportunity before the agency to cast doubt on the Tenaris data’s suitability as a CV profit source.

The court does hold, however, that Commerce must readdress its “same general category of products” determination on remand, as certain aspects of its reasoning suggest that it has impermissibly interpreted that term in making its determination. Specifically, although Commerce’s explanation that Korean non-OCTG pipe products were sold to the construction industry (in other words, they have different users and uses than OCTG) appears to be one factor Commerce reasonably could consider in determining that nonrOCTG pipe products are not in the same general category of products, the specific market conditions within those industries seems to be an irrelevant consideration. See I & D Memo at 17-19. Commerce’s reasoning appears to suggest that the weak demand in the construction industry coupled with the strong demand in the oil and gas industry was an important factor in its consideration. See id. Such logic, however, suggests that if demand in the construction industry had been strong during the POI while demand in the oil and gas industry had been weak, Commerce’s conclusion regarding the same general category of products might have been different. This would suggest a product might be within the same general category of products one year, but outside of that category the next year because of general market conditions in the industry. The court has strong doubts that the general category of products can be defined by such temporary factors.

The court additionally has doubts regarding Commerce’s reliance on the testing and certification requirements for OCTG. See id. at 18. If non-OCTG pipe products met those testing and certification requirements, it seems that they would be classified as OCTG. The SAA indicates that the “same general category of products” “encompasses a category of merchandise broader than the ‘foreign like product.’ ” SAA, H.R. Doc. No. 103-316, vol. 1, at 840, 1994 U.S.C.C.A.N. at 4176. Commerce’s reasoning seems to suggest that because non-OCTG pipe cannot be classified as OCTG, then it cannot be within the same general category of products. Commerce thus appears to have limited the same general category of products to the foreign like product.

On remand, Commerce must either omit these considerations from its analysis or provide an adequate explanation as to why these are appropriate factors for it to consider in determining what products fall within the same general category of products as OCTG.

C. Profit Cap

HYSCO, NEXTEEL, and Husteel argue that Commerce also erred by failing to apply a profit cap when it relied upon the profit from Tenaris’s financial statement for CV profit pursuant to alternative (in). They assert that Commerce’s reasoning for declining to apply the cap, namely that there was no data on the record regarding the profits normally earned by Korean producers on sales of merchandise in the same general category of products, was unsupported by substantial evidence. HYSCO Br. at 47, 50; NEXTEEL Br. at 45-46, 48; Husteel Br. at 34-36. They further assert that even if Commerce reasonably concluded that non-OCTG pipe does not fall within the same general category of products, Commerce was still required to attempt to apply a profit cap on the basis of the facts available. HYSCO Br. at 47-50; NEXTEEL Br. at 46, 48-49; Husteel Br. at 34-35. HYSCO, NEXTEEL, and Husteel emphasize the importance of applying a profit cap in this case because of the allegedly aberrational profit margin Tenaris earned, which was not based on production or sales of OCTG in Korea. HYSCO Br. at 48-49; NEXTEEL Br. at 46^18; Husteel Br. at 36.

Because the court is remanding the use of Tenaris’s financial statement to value a CV profit figure for the mandatory respondents, it need not discuss this issue in great detail. On remand, Commerce may decide to rely on the data plaintiffs suggest should be used as a profit cap in order to calculate CV profit. On the other hand, the court finds it appropriate to give Commerce some guidance on this issue should it continue to rely on the Tenaris financial statement pursuant to alternative (in) and continue to find that the nonOCTG products sold by the Korean producers in Korea do not fall within the same general category of products as the subject merchandise.

When Commerce used Tenaris’s financial statement to calculate a CV profit amount, it relied on alternative (in), which provides that Commerce may use

the amounts incurred and realized for selling, general, and administrative expenses, and for profits, based on any other reasonable method, except that the amount allowed for profit may not exceed the amount normally realized by exporters or producers (other than the exporter or producer described in clause (i)) in connection with the sale, for consumption in the foreign country, of merchandise that is in the same general category of products as the subject merchandise ....

19 U.S.C. § 1677b(e)(2)(B)(iii) (emphasis added). The government and petitioners argue that Commerce lacked record evidence regarding “the amount normally realized by exporters or producers ... in connection with the sale, for consumption in the foreign country, of merchandise that is in the same general category of products as the subject merchandise.” Gov. Br. at 48-49; Maverick Resp.’at 42; U.S. Steel Resp. at 56-57. In support of Commerce’s decision to apply alternative (iii) without a profit cap, they cite to the following passage in the SAA:

.The Administration also recognizes that where, due to the absence of data, Commerce cannot determine amounts for profit under alternatives (1) and (2) or a “profit cap” under alternative (3), it might have to apply alternative (3) on the basis of “the facts available.” This ensures that Commerce can use alternative (3) when it cannot calculate the profit normally realized by other companies on sales of the same general category of products.

SAA, H.R. Doc. No. 103-316, vol. 1, at 841, 1994 U.S.C.C.A.N. at 4177. According to the government and petitioners, this passage indicates that Commerce had the authority to apply alternative (iii) without calculating a profit cap. Gov. Br. at 48; Maverick Resp. at 41-42; U.S. Steel Resp. at 56-57, 60.

Even assuming that Commerce reasonably concluded that the record lacked data regarding the profit normally realized by Korean producers on Korean sales of merchandise in the same generally category of products, Commerce still was required to attempt to apply a profit cap on the basis of the facts available. As explained in Geum Poong Corp. v. United States, “[i]f Alternative Three without the profit cap may be used as ‘facts available,’ it would seem a ‘facts available’ profit cap may also be used.” 25 CIT 1089, 1097, 163 F.Supp.2d 669, 679 (2001). “Because the statute mandates the application of a profit cap, Commerce cannot sidestep the requirement without giving adequate explanation even in a facts available scenario.” Id.; accord Atar, S.r.l. v. United States, 34 CIT 465, 470, 703 F.Supp.2d 1359, 1364 (2010) (“But even the exception for absence of record data does not allow Commerce to ignore the profit cap requirement entirely when determining constructed value profit. Where the record lacks data on profit normally realized by other companies on sales of the same general category of products, Commerce still must'attempt to comply with the profit cap requirement through the use of facts otherwise available.”), rev’d on other grounds, 730 F.3d 1320 (Fed.Cir.2013). Even when the record evidence is deficient for the purposes of calculating the profit cap, Commerce must attempt to calculate a profit cap based on the facts otherwise available, and it may dispense with the profit cap entirely only if it provides an adequate explanation as to why the available data would render any cap based on facts available unrepresentative or inaccurate. See Geum Poong Corp. v. United States, 26 CIT 322, 324, 193 F.Supp.2d 1363, 1367 (2002) (“Geum Poong II ”).

Contrary to the claims of the government and U.S. Steel, Commerce failed to provide an adequate explanation as to why it dispensed with the profit cap. See Gov. Br. at 50-52; U.S. Steel Resp. at 60. The entirety of Commerce’s discussion regarding the profit cap was limited to a single sentence. See I & D Memo at 21 (“Lastly, we are unable to calculate a profit cap for Korea under section (iii) because we do not have home market profit data for other exporters and producers in Korea of the same general category of prod ucts.”). This explanation falls far short of the standard expressed in the court’s prior cases. As best the court can determine, Commerce completely failed to consider the possibility of applying a facts available profit cap, based on an erroneous legal conclusion. Commerce certainly did not explain why the use of such a profit cap would render the CV profit rate unreasonable and unrepresentative for HYSCO and NEXTEEL.

The use of an appropriate profit cap seems especially important in this case. The goal in calculating CV profit is to approximate the home market profit experience of the respondents. See Geum Poong II, 26 CIT at 327, 193 F.Supp.2d at 1370. The profit data imbedded in Tenaris’s financial statement does not appear to be based on any sales or production in Korea. It therefore appears to be a relatively poor surrogate for the home market experience. Additionally, record evidence suggests that Tenaris is a massive producer of OCTG with production and associated services around the world. See, e.g., U.S. Steel’s Comments re: NEXTEEL’s Third Suppl. Section D Questionnaire Resp. at Ex. P 6-11, 14. Record evidence also suggests that Tenaris’s profits are among the highest in the world and that this profit figure is due in large part to Tenaris’s sales of unique, high-end OCTG products and global services. See id. at 19-20, 27; U.S. Steel’s Jan. 16, 2014 Submission of Factual Information at Ex. J 1-3, 5-6. The Korean producers, on the other hand, appear to be rather modest in comparison, both in the size of their operations and in the products and services they offer. See Husteel Br. at 30-32. As Commerce recognized in the preamble to its own regulations, “the sales used as the basis for CV profit should not lead to irrational or unrepresentative results.” Antidumping Duties; Countervailing Duties, 62 Fed. Reg. 27,296, 27,360 (Dep’t Commerce May 19, 1997); see also Thai I-Mei Frozen Foods Co. v. United States, 32 CIT 865, 883, 572 F.Supp.2d 1353, 1368 (2008) (“An unreasonably high profit estimate will defeat the fundamental statutory purpose of achieving a fair comparison between normal value and export price.”), rev’d on other grounds, 616 F.3d 1300 (Fed.Cir. 2010). It appears that dispensing with the profit cap requirement entirely in this case could run the risk that the CV profit rate will be unrepresentative of the respondents’ expected home market experience.

On remand, if Commerce calculates CV profit pursuant to alternative (iii), Commerce must either apply a profit cap or provide an adequate explanation as to why data on the record cannot be used to calculate a facts available profit cap. This is especially so should Commerce find adequate reason, heretofore absent, to use Tenaris’s 2012 financial statement to establish CV profit.

III. NEXTEEL’s Affiliation with POS-CO

In the Final Determination, Commerce determined that NEXTEEL was affiliated with POSCO, one of its hot-rolled coil suppliers, due to a “close supplier relationship.” See I & D Memo at 72. Commerce claimed that POSCO’s involvement in the production and sales sides of business put POSCO in a position to potentially exercise restraint or direction over NEXTEEL. 21 Id. at 72-73. As a result of the affiliation finding, Commerce applied the major input rule and adjusted NEXTEEL’s purchase prices for hot-rolled coil sourced from POSCO. Id. at 73-74. In a related finding, Commerce determined that NEXTEEL was affiliated with one of its customers and disregarded NEXTEEL’s sales data, opting to use different sales and expense data as the basis for its export and constructed export price calculations. See id. at 89.

NEXTEEL argues that Commerce erred in determining that it was affiliated with POSCO. 22 NEXTEEL Br. at 49-53. NEXTEEL asserts that it was not reliant on POSCO for its hot-rolled coil and that Commerce failed to consider certain temporal aspects of the relationship pertaining to POSCO’s involvement with NEXTEEL’s sales suggesting that POSCO did not control NEXTEEL. Id. NEXTEEL’s argument lacks merit.

The AD duty statute states “affiliated persons” include “any person who controls any other person and such other person,” and that “a person shall be considered to control another person if the person is legally or operationally in a position to exercise restraint or direction over the other person.” 19 U.S.C. § 1677(33)(G). Commerce considers “close supplier relationships” among other factors when determining whether control exists. 19 C.F.R. § 351.102(b)(3). Control will not be found to exist unless the relationship in question “has the potential to impact decisions concerning the production, pricing, or cost of the subject merchandise or foreign like product.” Id. The temporal aspect of a relationship is also considered. Id.

NEXTEEL disregards the text of the statute in arguing that POSCO did not exercise restraint or control over NEXTEEL. The statute does not require that Commerce find that POSCO actually controlled or restrained NEXTEEL, but rather it only requires that Commerce evaluate whether POSCO was in a position from which it could exercise such control. See 19 U.S.C. § 1677(33)(G). Moreover, NEXTEEL fails to recognize that, although Commerce may look to whether one of the parties has become reliant on the other, the presence or lack of reliance is not necessarily dispositive. See SAA, H.R. Doc. No. 103-316, vol. 1, at 838, 1994 U.S.C.C.A.N. at 4174-75. The record reveals that POSCO and NEXTEEL shared technology and marketing information, and POSCO had a very significant role on both the production and sales sides of NEXTEEL’s OCTG operations during the POI. 23 See NEXTEEL Affiliation Memorandum at 2-5, CD 443 (July 10, 2014). Based on the record information, Commerce reasonably concluded that POSCO had the ability to impact NEXTEEL’s decisions in most, if not all, of these areas. Accordingly, Commerce did not err in con eluding that NEXTEEL and POSCO were affiliated.

IY. NEXTEEL’s Warranty Expenses

Commerce’s initial questionnaire asked NEXTEEL to report its warranty expenses for its reported U.S. sales and for its annual warranty expenses during the three most recent fiscal years — 2010, 2011, and 2012. 1 & D Memo at 80. NEXTEEL reported it had incurred no warranty expenses for its reported U.S. sales, but it did not address its warranty expenses for the three most recent fiscal years. Id. Commerce then reiterated its request in a supplemental questionnaire, noting that there appeared to be discrepancies between NEXTEEL’s claims about its lack of warranty expenses during the POI and other information provided in its questionnaire response. Id. NEXTEEL responded with the three previous years’ warranty expenses, insisted that there were no discrepancies, and stressed that it had not incurred any warranty expenses during the POI. Id. at 80-81. NEXTEEL also made statements that appeared to imply that it had not received any warranty claims during the POI. See id. At verification, however, NEXTEEL revealed that it had outstanding unresolved warranty claims for 2012 and 2013. Id. at 81.

Though Commerce acknowledged that NEXTEEL “may have been less than candid in its questionnaire responses,” Commerce chose not to apply adverse facts available (“AFA”). Id. at 81. Commerce recognized that NEXTEEL did not appear to have incurred any warranty expenses during the POI, as NEXTEEL maintained throughout the investigation and in its questionnaire responses. Id. Commerce also noted that NEXTEEL supplied its three year warranty expense data despite its initial failure to do so. Id. For NEXTEEL’s warranty expenses, Commerce relied on NEXTEEL’s historical average, but excluded the year 2012 because there remained unresolved claims for that year. Id. Commerce similarly rejected using NEXTEEL’s POI warranty expenses because of outstanding claims. Id. Commerce further explained that “[u]se of all of the outstanding balances of NEXTEEL’s customer to determine NEXTEEL’s expenses as facts available [as petitioners had suggested] may yield an excessive estimate, given it is not evident that the outstanding balances are all due to warranty claims, nor is it obvious that all claims would result in actual warranty expenses.” Id.

B. Adverse Facts Available

Maverick and U.S. Steel argue that Commerce erred in deciding not to apply AFA to NEXTEEL because NEXTEEL misled Commerce regarding its warranty experience and did not fully cooperate during Commerce’s investigation. Maverick Br. at 31-35; U.S. Steel Br. at 12-22. Maverick and U.S. argue that Commerce’s conclusion that NEXTEEL acted to the best of its abilities was erroneous, and that the failure to apply AFA was an arbitrary departure from past practice. Maverick Br. at 31-35; U.S. Steel Br. at 19-22. This argument lacks merit.

According to the statute, Commerce shall use facts otherwise available if a party (i) withholds requested information, (ii) fails to provide such information by the appropriate deadlines or in the form and manner requested, (iii) significantly impedes a proceeding, or (iv) provides the requested information, but the information is incapable of being verified. 19 U.S.C. § 1677e(a)(2). Commerce may apply an adverse inference in selecting from the facts otherwise available if the party “has failed to cooperate by not acting to the best of its ability to comply with a request for information.” 19 U.S.C. § 1677e(b). Commerce has discretion over whether to apply or not apply AFA. See AK Steel Corp. v. United States, 28 CIT 1408, 1416-17, 346 F.Supp.2d 1348, 1355 (2004). Commerce is. not required “to prove that an importer cooperated to the best of its ability every time that the agency decides not to apply adverse facts available.” Id. at 1417, 346 F.Supp.2d at 1355. The issue of whether a respondent has acted to the best of its ability and whether AFA is appropriate “amounts to a line-drawing exercise that is precisely the type of discretion left within the agency’s domain.” Ta Chen Stainless Steel Pipe Co. v. United States, 31 CIT 794, 812, 2007 WL 1573920 (2007) (internal quotation marks and brackets omitted).

Maverick and U.S. Steel argue that AFA should have been applied, yet such arguments mischaracterize Commerce’s ability to apply AFA as an obligation to apply AFA. The statute limits Commerce’s ability to apply AFA to situations in which Commerce finds that a party has failed to act to the best of its ability. See Kawasaki Steel Corp. v. United States, 24 CIT 684, 689, 110 F.Supp.2d 1029, 1034 (2000). There is nothing in the statute, however, that requires Commerce to apply an adverse inference upon such a finding. See 19 U.S.C. § 1677e(b) (“If [Commerce] ... finds that an interested party has failed to cooperate by not acting to the best of its ability ..., [Commerce] ... may use an inference that is adverse to the interests of that party .... ” (emphasis added)).

Maverick and U.S. Steel also overstate the events that transpired while NEXTEEL was being investigated, exaggerating the possible effects on Commerce’s investigation. NEXTEEL did provide information that suggested it had not received any warranty claims during the POI, but Commerce later concluded that NEXTEEL’s repeated assertions about its warranty expenses (as opposed to unresolved claims) during the POI appeared to be true'. I & D Memo at 81. Providing information capable of misinterpretation is not necessarily emblematic of “gamesmanship” or attempts to obtain an “inaccurately low dumping margin,” particularly when such information is capable of verification. See Maverick Br. at 31. NEXTEEL provided the information Commerce requested, and ultimately relied upon, in plenty of time for the information to be verified and considered by Commerce. Although U.S. Steel additionally argues that NEXTEEL made false claims when it asserted that third parties were not involved in settling warranty expenses, this is actually not a false claim. See U.S. Steel Br. at 19. NEXTEEL stated that “[n]o third party acts on NEXTEEL’s behalf to cover warranty expenses,” or, in other words, no party but NEXTEEL ultimately pays for any warranty expenses. NEXTEEL’s Suppl. Sections A & C Questionnaire Resp. at 25, PD 193 (Dec. 30, 2013). This is not the same as claiming no third party played a role in settling these expenses, and it is a weak basis upon which to assert that NEXTEEL provided false information.

Maverick and U.S. Steel’s arguments that Commerce’s decision not to apply AFA is a departure from its past practice are unpersuasive, as the petitioners rely on cases distinguishable from the present case. For example, petitioners cite Mukand, Ltd. v. United States to argue that AFA is appropriate when a respondent “evade[s] providing a direct response to Commerce’s specific questions” and effectively “sit[s] out the preliminary phase of the investigation.” 767 F.3d 1300, 1307 (Fed.Cir.2014). But in that case, the respondent, Mukand, repeatedly evaded Commerce’s requests. Id. at 1303. Unsatisfied with Mukand’s responses in its fourth supplemental questionnaire, Com merce went as far as to create a sample chart for Mukand to complete that clearly identified the information Commerce was requesting and the manner in which it should be recorded. Id. Here, it did not take Commerce four attempts to secure the information it had initially requested from NEXTEEL, and NEXTEEL was demonstrably more cooperative and forthcoming than Mukand. Similarly, in Shandong Huarong Machinery Co. v. United States, the court held that Commerce’s need to resort to several supplemental questionnaires to obtain information from an importer “surely significantly impeded Commerce’s investigation.” 30 CIT 1269, 1277, 435 F.Supp.2d 1261, 1269 (2006). There is nothing indicating that Commerce’s investigation was similarly hindered by NEXTEEL. And unlike Essar Steel Ltd. v. United States, where the court held that providing false information and failing to produce key documents demonstrated a respondent did not put forth its maximum effort, NEXTEEL’s pertinent responses were never found to contain false information. See 678 F.3d 1268, 1275-76 (Fed.Cir.2012).

Moreover, Maverick cites to a number of prior issues and decision memoranda that concern behavior much more problematic than NEXTEEL’s. See Maverick Br. at 34-35. For example, in Certain ColdrRolled Carbon Steel Flat Products from Brazil, Commerce applied partial AFA where a respondent failed to report U.S. sales that were discovered at verification. See Issues and Decision Memorandum for the Final Determination of Sales at Less Than Fair Value: Certain Cold-Rolled Carbon' Steel Flat Products from Brazil at 7-8, A-351-834, (Sept. 23, 2002), available at http://enforcement.trade.gov/frn/ summary/brazil/02-24800-l.pdf (last visited Aug. 27, 2015). And in Prestressed Concrete Steel Wire Strand from Mexico, Commerce applied AFA when the respondent misreported its movement expenses and failed to correct them despite numerous opportunities to do so. Issues and Decision Memorandum for the Final Determination of the Investigation of Prestressed Concrete Steel Wire Strand from Mexico at 3-4, A-201-831, (Dec. 1, 2003), available at http://enforcement.trade.gov/ frn/summary/mexico/03-30384-1.pdf (last visited Aug. 27, 2015). Such situations evince behavior that is more egregious, deceitful, and deserving of AFA than NEXTEEL’s. Unlike the above mentioned respondents, NEXTEEL voluntarily revealed information about its warranty expenses, and NEXTEEL did not misreport and leave uncorrected any information it provided to Commerce. Commerce has not arbitrarily refused to apply AFA or unreasonably departed from its past practice.

Finally, Commerce is not required to apply an AFA expense that it determines is likely to be unreflective of the respondent’s actual expenses. Commerce explained that reliance on NEXTEEL’s customer’s outstanding balances as facts available “may yield an excessive estimate, given it is not evident that the outstanding balances are all due to warranty claims, nor is it obvious that all claims would result in actual warranty expenses.” I & D Memo at 81. • Commerce’s goal is to calculate dumping margins that are as accurate as possible. Parkdale, 475 F.3d at 1380 (citing Rhone Poulenc, 899 F.2d at 1191). Commerce acted reasonably in relying upon NEXTEEL’s historical warranty experiences rather than using the distortive amounts suggested by petitioners.

The court holds that Commerce’s decision to rely on NEXTEEL’s historical experience regarding warranty expenses rather than applying AFA was reasonable, supported by substantial evidence, and in accordance-with law.

C. Warranty Claims in the Warranty Expense Calculation

U.S. Steel argues that Commerce erred when it excluded outstanding payments withheld by NEXTEEL’s U.S. Customer for warranty claims filed during the POI. U.S. Steel Br. at 22-24. U.S. Steel asserts that even if the court upholds Commerce’s determination regarding AFA, the court should at least remand Commerce’s calculation of NEXTEEL’s warranty expenses and direct Commerce to include the amount of the outstanding balances unpaid by NEXTEEL’s U.S. customer due to warranty claims. Id. U.S. Steel’s argument lacks merit.

Generally, an entity’s total amount of warranty expenses is unknown at the time of sale, and, because of this, Commerce has developed a practice of relying on a company’s warranty expenses during the POI. I & D Memo at 80. If these warranty expenses are found distortive, Commerce uses a company’s three-year historical warranty expenses regardless of the particular periods during which the relevant sales occurred. Id.; see, e.g., Issues and Decision Memorandum for the Final Determination in the Antidumping Duty Investigation of Crystalline Silicon Photovoltaic Cells, Whether or Not Assembled into Modules, from the People’s Republic of China at 80, A-570-979, (Oct. 9, 2012), available at http://enforcement.trade.gov/ frn/summary/prc/2012-25580-1.pdf (last visited Aug. 27, 2015).

Here, Commerce determined that relying on NEXTEEL’s POI warranty expenses would be distortive, as NEXTEEL had outstanding warranty claims and its reported expenses were not representative of its historical experience. See I & D Memo at 80-81. Commerce adjusted NEXTEEL’s export price by a historical average of its 2010 to 2011 warranty expenses. Id. at 81. Commerce ' excluded the 2012 warranty expenses from its calculation, finding that including unsettled warranty claims for that year also would be distortive. Id.

U.S. Steel’s argument that NEXTEEL’s outstanding warranty claims are the best measure of NEXTEEL’s warranty expenses for the POI is unpersuasive. The warranty claims U.S. Steel urges Commerce to include in NEXTEEL’s warranty expense calculation were pending claims. See U.S. Steel Br. at 24. It was reasonable for Commerce to exclude claims of uncertain amounts from its calculation and to conclude that such claims could be distortive. For example, it is possible that a customer could make a warranty claim for an amount that, once investigated, is determined to be incorrect and overestimated, or the alleged defect might have been caused by a party other than NEXTEEL. See I & D Memo at 81. These examples illustrate that a claimed amount will not necessarily equal the amount NEXTEEL ultimately pays, and thus including unsettled claims in the margin calculation is likely to lead to inaccurate results. Commerce’s decision to use NEXTEEL’s historical average for its warranty expenses rather than the full amount of the unsettled pending claims was reasonable, and the court therefore upholds Commerce’s warranty expense calculation for NEXTEEL. 24

V. NEXTEEL’s Warehousing Expenses

Commerce is required to include general and administrative (“G & A”) expenses in the CV calculation. See 19 U.S.C. § 1677b(e)(2). G & A expenses are costs associated with the day-to-day operation of a business, such as rent, electricity, and executive salaries. See Ass’n of Am. Sch. Paper Suppliers v. United States, 33 CIT 1742, 1745, 1752 (2009) affd, 410 Fed. Appx. 320 (Fed.Cir.2010). It is Commerce’s practice to use the financial statements from the full fiscal year that most closely corresponds to the POI in calculating these G & A expenses. Issues and Decision Memorandum for the Final Affirmative Determination in the Less-Than-Fair-Value-Investigation of Grain-Oriented Electrical Steel from the Republic of Korea at 24, A-580-871, (Sept. 24, 2014), available at http://enforcement.trade.gov/ frn/summary/korea-south/2014-23393-1. pdf (last visited Aug. 27, 2015). When the POI is divided across two fiscal years, Commerce uses the financial statements from the most recently completed fiscal year. Id.

Commerce’s initial questionnaires requested that NEXTEEL identify each of its affiliated entities and report its domestic warehousing expenses for its sales of OCTG to the United States. NEXTEEL responded that it did not have affiliates beyond NEXTEEL America and NEXTEEL QNT Co., Ltd., and that it had no such warehousing expenses. .NEXTEEL’s Section A Questionnaire Resp. at A-8 — A-9, PD 121 (Sept. 18, 2013); NEXTEEL’s Sections C-D Questionnaire Resp. at C-23, PD 152-153 (Nov. 5, 2013). Commerce requested that NEXTEEL further explain how it reported expenses related to transporting OCTG from its plants to the storage yards at ports or other intermediate locations. NEXTEEL’s Suppl. Sections A & C Questionnaire Resp. at 19. NEXTEEL again maintained that it did not have any warehousing expenses and that it did not transport the OCTG to an intermediate distribution warehouse. Id. at 19-21.

Soon after the preliminary determination, however, NEXTEEL reported that it had a previously unreported affiliate: NEXTOGY.- NEXTEEL’s Second Suppl. Sections A & C Questionnaire Resp. at 9-10, CD 256 (Feb. 18, 2014). NEXTEEL acknowledged that it incurred warehousing expenses during the POI for services provided by NEXTOGY and claimed that these expenses were reported to Commerce as a part of NEXTEEL’s G & A expenses. Id. NEXTEEL submitted lease contracts between itself and an unaffiliated party 25 to demonstrate that NEXTEEL paid the same amount for warehousing services to an unaffiliated party as it did to NEXTOGY, supposedly indicating that its transactions with NEXTOGY were conducted on an arm’s-length basis. Id. at 10.

For the Final Determination, Commerce declined to apply AFA when calculating NEXTEEL’s warehousing expenses, despite its recognition that NEXTEEL’s belated disclosure was too late for Commerce to consider any expenses associated with NEXTOGY in its preliminary margin calculations, and despite Commerce’s conclusion that it was implausible NEXTEEL was unaware of the NEXTOGY facility. I & D Memo at 85. Commerce explained that, consistent with the Preliminary Determination, it was basing its CV selling ratios,, which includes G & A expenses, on NEXTEEL’s 2012 data. Id. Commerce verified that NEXTEEL had not made warehousing payments to NEXTOGY until 2013, and thus NEXTEEL had no expenses it neglected to include in its 2012 G & A expenses. Id. Because Commerce was relying on the 2012 data, any expenses incurred in 2013 were irrelevant to the margin calculation. Id.

Maverick and U.S. Steel argue that Commerce erroneously failed to apply AFA with regard to NEXTEEL’s warehousing expenses. Petitioners claim that NEXTEEL failed to cooperate with Commerce’s investigation as it related to reporting the warehousing expenses it incurred during the POI and disclosing the fact that it purchased warehousing services from an affiliate, NEXTOGY. Maverick Br. at 36-41; U.S. Steel Br. at 25-32. In addition to NEXTEEL’s failure to supply this information when initially asked, they point to alleged errors in the information that NEXTEEL ultimately submitted and challenge NEXTEEL’s assertion that it included any relevant warehousing expenses in its G & A expenses. Maverick Br. at 36-41; U.S. Steel Br. at 25-32. Petitioners argue that Commerce acted contrary to law and that its determination was unsupported by substantial evidence. These arguments lack merit.

First, as the government points out, although NEXTEEL’s initial responses regarding its affiliates and its warehousing expenses during the POI were incorrect, NEXTEEL corrected this information before verification. Gov. Br. at 95. The government also notes that NEXTEEL provided over two thousand pages of initial and supplemental questionnaire responses and cooperated fully during the verification process. Id. at 94-95. The “best of its ability” standard “does not require perfection and recognizes that mistakes sometimes occur.” Nippon Steel Corp. v. United States, 337 F.3d 1373, 1382 (Fed.Cir.2003). The court is disinclined to second guess Commerce’s line-drawing when it determines that a party has acted to the best of its ability. See Ta Chen, 31 CIT at 812.

Second, even if NEXTEEL’s ultimate disclosure of NEXTOGY and warehousing expenses was untimely and in some waydeficient, Commerce is not required to apply an adverse inference, even if a party has not acted to the best of its ability. See AK Steel Corp., 28 CIT at 1416-17, 346 F.Supp.2d at 1355. Consistent with its practice, Commerce relied on NEXTEEL’s 2012 data to calculate G & A expenses. This information was promptly submitted and verified. NEXTEEL did not purchase warehousing services from NEXTOGY until 2013 and thus any errors or omissions related to the NEXTOGY expenses did not affect NEXTEEL’s margin calculation. Thus, even if Commerce could have concluded that NEXTEEL did not act to the best of its abilities, Commerce did not abuse its discretion by choosing to rely on NEXTEEL’s timely submitted and verified 2012 expenses instead of applying AFA because of supposed deficiencies regarding information that Commerce ultimately deemed irrelevant when calculating NEXTEEL’s dumping margin.

VI. Excluding a Loss from NEXTEEL’s G & A Expense Calculation

U.S. Steel argues that Commerce improperly excluded a miscellaneous loss in calculating NEXTEEL’s G & A expense. U.S. Steel Br. at 32-34. The loss at issue was incurred in connection with a payment guarantee made for one of NEXTEEL’s domestic standard pipe customers when the customer 'defaulted on a loan. 1 & D Memo at 101. Commerce included the loss when calculating NEXTEEL’s G & A expense ratio in the Preliminary Determination, but excluded the loss from the G & A expense ratio calculated for the Final Determination. Id. Commerce based its final conclusion on NEXTEEL’s argument that the loss was “akin to a bad debt expense incurred in connection to the sales of.standard pipes in the domestic market” and thus a selling expense associated with non-subject merchandise. Id. U.S. Steel claims that the loss at issue occurred as part of NEXTEEL’s general operations and thus should have been included as a G & A expense. U.S. Steel Br. at 32. U.S. Steel’s arguments lack merit.

In calculating G & A expenses, it is Commerce’s practice to include those expenses “which relate to the activities of the company as a whole rather than to the production process.” Rautaruukki Oy v. United States, 19 CIT 438, 444, 1995 WL 170399 (1995). Commerce typically excludes expenses from the G & A rate calculation “only when the expenses are both: (1) unusual; and (2) infrequent in nature.” Torrington Co. v. United States, 25 CIT 395, 431, 146 F.Supp.2d 845, 886 (2001); see also Thai Plastic Bags Indus. Co. v. United States, 904 F.Supp.2d 1326, 1331-32 (CIT 2013) (recognizing that losses that are not related to a company’s normal production-related business operations are excluded from the G & A expense calculation).

Commerce reasonably concluded that the loss should not be included in NEXTEEL’s G & A expense, as this loss was not related to NEXTEEL’s general operations. Although it is true Commerce did not provide a thorough explanation of how it determined the loss was “akin to a bad debt expense” in the I & D Memo, NEXTEEL is correct that what matters is not whether the loss is properly classified as a bad debt, but rather whether the loss is related to NEXTEEL’s general operations. See I & D Memo at 101; NEXTEEL’s Resp. in Opp’n to Consol. Pis. Maverick and U.S. Steel Corp.’s Rule 56.2 Mots, for J. on the Agency R. 45, ECF No. 157 (“NEXTEEL Resp.”). NEXTEEL primarily functions as a manufacturer and seller of tubular products, and U.S. Steel has not pointed to anything in the record showing that guaranteeing loans for customers is an ordinary aspect of NEXTEEL’s business operations. Commerce therefore reasonably concluded that NEXTEEL’s providing the loan guarantee to its customer of non-subject merchandise was not part of NEXTEEL’s ordinary or general business operations, but rather was more properly characterized as a selling expense related to non-subject merchandise. To the extent that U.S. Steel contests Commerce’s decision to reverse course on this issue in the Final Determination without having received any new evidence following the Preliminary Determination, the court rejects this contention. “[Preliminary determinations are ‘preliminary’ precisely because they are subject to change.” NTN Bearing Corp. v. United States, 74 F.3d 1204, 1208 (Fed.Cir.1995). Commerce was not prohibited from reconsidering its analysis of the evidence. The exclusion of this expense from NEXTEEL’s G & A calculation was supported by substantial evidence and in accordance with law.

VII. Valuation of Hot-Rolled Steel Coil for NEXTEEL’s Constructed Value Calculation Using Weighted-Average Prices

Because Commerce determined that NEXTEEL and POSCO were affiliated within the meaning of 19 U.S.C. § 1677(33)(G), Commerce applied the major input rule to NEXTEEL’s purchases of hot-rolled coil from POSCO. I & D Memo at 72-74. The major input rule is applied when there is a transaction between affiliated parties involving one party’s production of a major input needed for the production of the subject merchandise, as such a situation presents reasonable grounds for Commerce to suspect that “an amount represented as the value of such input is less than the cost of production of such input.” 19 U.S.C. § 1677b(f)(3). Commerce normally calculates the major input’s value using the higher of (1) the transfer price the respondent paid the affiliate for the input, (2) the amount usually-reflected in sales of the input in the market under consideration, or (3) the costs the affiliate incurs in producing the input. 19 C.F.R. § 351.407(b) (2014). Here, Commerce calculated the market price using the weighted average of POSCO’s sales to all of its unaffiliated customers. See I & D Memo at 74. Commerce used the transfer price to value certain grades of coil and used the market price for other grades because the market price exceeded the transfer price. Constructed Value Calculation Adjustments for the Final Determination — NEXTEEL at 3, CD 431 (July 10, 2014). The transfer price for each grade was higher than the cost of production. Id.

Maverick argues Commerce erred in using a weighted average of the prices at which POSCO sold hot-rolled steel coil to all of its unaffiliated customers as the market price for the major input rule comparison. Maverick Br. at 21-28. Maverick claims Commerce improperly disregarded evidence demonstrating that POSCO’s prices to its unaffiliated customers were distorted and unreliable. See id. at 24-28. According to Maverick, the price paid by Company A 26 is the best representation of the actual market price, because Company A is a larger producer than some of the other unaffiliated producers and was the only unaffiliated producer with its own alternative supply of hot-rolled coil. Id. at 24-25. Maverick argues Commerce should have found that the POSCO-Company A price was the only price on the record that avoided the “aberrantly low weighted-average prices.” 27 Id. at 22. Maverick’s arguments lack merit.

It was not irrational or arbitrary for Commerce to reason that using POSCO’s sales to all of its unaffiliated customers to calculate a market price would better demonstrate the price usually reflected in sales of the major input in Korea than would the price paid by a single entity, Company A. As the government has explained, nothing in the applicable regulation, 19 C.F.R. § 351.407(b)(2), requires Commerce to focus on only the largest producers in the market or customers with independent sources for the input. Gov. Br. at 82. The regulation only specifies that Commerce should use the market price that is “usually” reflected in the sales of the input in the relevant market, and thus it was within Commerce’s discretion to determine that the weighted average of the prices that all of POSCO’s unaffiliated customers^paid for the input would provide the most comprehensive overview of market conditions. 28 19 C.F.R. § 351.407(b). Commerce’s deci sion demonstrates a “rational connection between the facts found and the choice made.” See Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962).

Further, Maverick’s argument that Commerce failed to consider evidence of a “silent agreement” between POSCO and Korean OCTG and line pipe producers is highly speculative and unpersuasive. Maverick Br. at 24-25, 28. Maverick’s argument relies on the assumption that the alleged agreement influenced prices for some unaffiliated OCTG producers, but not for other unaffiliated Korean OCTG producers. But the affidavit cited by Maverick supporting its argument does not mention any specific Korean OCTG producers, nor does it specifically discuss pricing practices in the Korean market. 29 U.S. Steel’s Comments re: POSCO’s Questionnaire Resp. at Ex. 3, CD 304 (Mar. 21, 2014). Maverick takes issue with Commerce’s failure to address the affidavit when conducting its major input rule analysis. Although an agency “must address significant arguments and evidence which seriously undermines its reasoning and conclusions,” an agency need not address every argument and piece of evidence. Altx, Inc. v. United States, 25 CIT 1100, 1117-18, 167 F.Supp.2d 1353, 1374 (2001). Because this argument and accompanying evidence were not significant, Commerce did not err in failing to specifically address them.

Maverick’s additional argument that the prices to the other unaffiliated producers should have been rejected because they were only slightly higher than the prices POSCO charged its affiliates is also without merit. See Maverick Br. at 26-27. The major input rule compares the transfer price to the market price (along with the cost of production), and Commerce will use the higher of these prices. 19 C.F.R. § 351.407(b). Implicit is the possibility that the transfer price between affiliates might be equal to or higher than the market price. The government correctly notes that Maverick’s argument essentially requires a comparison of market prices to the transfer price in order to determine if Commerce can properly rely on the market price when comparing it to transfer price, which seems to defeat the whole point of conducting the comparison in the first place. Gov. Br. at 84. Accordingly, Commerce’s application of the major input rule to NEXTEEL’s purchase of hot-rolled coil from POSCO is sustained. 30

VIII. Costs Associated with HYSCO’s Affiliated Service Providers

In making an arm’s-length determination, Commerce typically compares the transfer price a party pays an affiliate to the market price for the particular good. Where a market price is unavailable, Commerce will use the affiliate’s cost of production of the relevant input or service as a proxy for the market price. See I & D Memo at 44; see also Issues and Decision Memorandum for the Antidumping Duty Investigation of Large Residential Washers from Mexico at 12, A-201-842, (Dec. 18, 2012), available at http://enforeement. trade.gov/frn/summary/mexico/201231077-l.pdf (last visited Aug. 27, 2015). During the investigation, Commerce had first instructed HYSCO to provide the per-unit price HYSCO paid to each affiliate and the affiliates’ per-unit costs of production (“COP”) data so that Commerce could determine whether the services were obtained through arm’s-length transactions. See HYSCO’s Suppl. Sections A, C & D Questionnaire Resp. at SD-5. HYS-CO provided the per-unit price it paid to each affiliate, but it claimed it could not provide its affiliates’ COP data, as its affiliates considered this data highly confidential. Id. Instead, HYSCO provided an estimate of its affiliates’ COP data using their financial statements. Id. Based on this estimated data, for the Preliminary Determination, Commerce made an upward adjustment to the reported service costs so that they reflected the costs of arm’s-length transactions. Preliminary CV Calculation Memorandum for HYSCO at 1-2, CD 244 (Feb. 14, 2014). At verification, however, HYSCO backed away from its estimated COP data, arguing that its affiliates were overstating costs and revenues in their financial statements and that these costs and revenues should be reduced before calculating each affiliate’s per-unit COP. Cost Verification Report for HYSCO at 19-20, PD 419 (May 20, 2014). According to HYSCO, its affiliates were recording revenue and expenses related to costs that were actually paid by HYSCO. Id. For the Final Determination, Commerce accepted HYSCO’s assertions regarding overstated costs and recalculated each affiliate’s per-unit COP. See I & D Memo at 45.

Maverick and U.S. Steel argue that Commerce erred in failing to apply AFA to the service costs HYSCO paid to affiliated tolling service providers. Maverick Br. at 41-49; U.S. Steel Br. at 34-42. Maverick and U.S. Steel claim that HYS-CO did not act to the best of its ability in complying with Commerce’s request that it obtain and report COP data for its affiliated service providers. Maverick Br. at 41-49; U.S. Steel Br. at 34-42. Maverick and U.S. Steel argue that Commerce erred in concluding that HYSCO could not compel its affiliated service providers to provide their COP data. Maverick Br. at 43-47; U.S. Steel Br. at 38-42. U.S. Steel additionally contends that even if AFA was not warranted, Commerce erred when it concluded that the service fees HYSCO paid to its affiliates were arm’s-length transactions. U.S. Steel Br. at 42-46. Petitioners’ arguments lack merit.

Maverick and U.S. Steel assert that HYSCO did not cooperate with Commerce’s investigation to the best of its ability. Although evidence on the record might suggest that HYSCO was in a relatively strong position to command its affiliates’ data, the court cannot say that Commerce’s decision was without substantial evidence. See Consolo v. Fed. Mar. Comm’n, 383 U.S. 607, 620, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1966) (“[Substantial evidence] is something less than the weight of the evidence, and the possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency’s finding from being supported by substantial evidence.”). As Commerce recognized, HYSCO maintained only “small equity ownership in each of its affiliated service providers.” I & D Memo at 50. HYSCO’s ownership in each affiliate was less than 15%. Resp. to Ct.’s Req. Re: Confidentiality of Certain of Hyundai HYSCO’s Info. Contained in the Parties’ Brs. 3, ECF. No. 214. Maverick and U.S. Steel challenge Commerce’s analysis, pointing to the unique and interconnected nature of companies operating within the structure of larger Korean chaebols, yet HYSCO’s small ownership shares in its affiliates is significant considering Commerce had previously considered small equity ownership consistent with a party’s inability to compel the COP data of their affiliates. See Certain Cutr-To-Length Carbon Steel Plate From Brazil: Final Results of Antidumping Duty Administrative Review, 63 Fed.Reg. 12,744, 12,751 (Dep’t Commerce Mar. 16, 1998).

HYSCO’s situation is also distinguishable from the precedent Maverick and U.S. Steel rely on to argue that HYSCO did not cooperate to the best of its ability. In Kawasaki, the court sustained Commerce’s determination that the respondent’s letters and oral requests for information from its affiliate did not demonstrate that the respondent had acted to the best of its ability. 24 CIT at 694, 110 F.Supp.2d at 1039. Although HYSCO’s telephonic and written requests also do not appear to indicate the company expended a great degree of effort in obtaining the requested COP data, further such effort likely would have been futile and HYSCO did not exhibit the same “hands-off’ approach that led Commerce to apply AFA to the respondent in Kawasaki. Id. at 689-90, 110 F.Supp.2d at 1034-35. HYSCO calculated its own derived data and reconciled its affiliates’ sales revenues listed in each company’s 2012 financial statement with the transfer prices HYSCO reportedly paid to each affiliate. I & D Memo at 44-45. Unlike the respondent in Kawasaki, who requested to be excused from providing the data and did not suggest any alternative method of providing the requested information, see 24 CIT at 686, 110 F.Supp.2d at 1032, HYSCO made an effort to provide its best estimate of the information Commerce had asked HYSCO to report.

HYSCO’s situation is similarly distinguishable from many of Commerce’s determinations petitioners cite to for the same reason. See, e.g., Notice of Preliminary Determination of Sales at Less Than Fair Value, Postponement of Final Determination, and Affirmative Preliminary Critical Circumstances Determination: Certain Orange Juice from Brazil, 70 Fed. Reg. 49,557, 49,564 (Dep’t Commerce Aug. 24, 2005) (applying AFA where respondent completely failed to provide COP information for an affiliate’s facility); Stainless Steel Wire Rods from India: Preliminary Results and Partial Rescission of Anti-dumping Duty Administrative Review, 68 Fed.Reg. 70,765, 70,768-69 (Dep’t Commerce Dec. 19, 2003) (applying AFA on account of respondent’s repeated failure to provide affiliate’s COP data without providing explanation for failure to comply). Failing to provide data requested by Commerce is not the same as being unable to provide the requested data and providing a reasonable alternative. The court holds that Commerce’s decision to accept the estimated COP data rather than applying AFA was supported by substantial evidence and in accordance with law.

The court also concludes Commerce acted reasonably in adjusting the COP data to exclude costs HYSCO’s affiliates recorded as both revenue and expenses once Commerce learned that these costs actually were paid by HYSCO. See Constructed Value Calculation Adjustments for the Final Determination — HYSCO at 2-3 and Attach. 4, CD 433 (July 10, 2014). Once HYSCO brought this discrepancy to Commerce’s attention, Commerce reviewed and tested the reconciled information, deemed it acceptable, and recalculated the COP accordingly. I & D Memo at 45. HYSCO was able to sufficiently show that its affiliates were treating the payment of certain costs by HYSCO as revenue, and Commerce could reasonably infer that the affil iates likewise were treating those costs as if they were the affiliates’ own costs. Commerce’s determination was not based upon mere speculation without any support in the record. Rather, Commerce’s decision to adjust the data demonstrated a “rational connection between the facts found and the choice made.” See Burlington Truck Lines, 371 U.S. at 168, 83 S.Ct. 239.

IX. HYSCO’s Warranty Expenses

Initially, HYSCO reported that it had incurred no warranty expenses related to its U.S. sales during the POI, yet later it claimed to have made an error, stating that it had incurred warranty expenses. See HYSCO’s Sections C-D Questionnaire Resp. at C-29; HYSCO’s Suppl. Sections A, C & D Questionnaire Resp. at SC-21. Commerce also discovered three previously unreported warranty claims regarding HYSCO merchandise at the verification of HYSCO’s U.S. customer. I & D Memo at 58. Before Commerce, U.S. Steel argued that HYSCO and its U.S. affiliate, Hyundai HYSCO USA, Inc. (“HHU”), absorbed losses incurred in shipping defective pipe, but that these expenses were not reported as part of HYSCO’s warranty expenses or elsewhere in HYSCO’s data. Id. at 57. Commerce, however, did not revise HYS-CO’s warranty expenses to include movement expenses related to defective pipe. Id. at 58. Commerce found the record was unclear whether these expenses were accounted for elsewhere in HYSCO’s costs, and Commerce did not want to risk double counting these expenses. Id.; Gov. Br. at 122-23. Commerce also did not revise HYSCO’s warranty expenses to include any of the three warranty claims. I & D Memo at 58. Commerce determined that one of the three claims was dated after the POI, and it found no evidence that the other two claims actually were paid and settled during the POI. Id.

U.S. Steel argues that Commerce erred when it failed to adjust HYSCO’s reported warranty expenses to include certain movement expenses and warranty claims HYSCO had omitted from its calculation. U.S. Steel Br. at 46-50. U.S. Steel’s arguments lack merit.

A. Movement Expenses

U.S. Steel argues that Commerce erred in refusing to revise HYSCO’s warranty expenses to account for the costs it incurred in shipping defective pipe. U.S. Steel contends that the record is clear that the data HYSCO reported only captured the movement costs for non-defective pipe. See Reply Br. in Supp. of PI. United States Steel Corp.’s Mot. for J. on the Agency R. Under Rule 56.2 38, ECF No. 190 (“U.S. Steel Reply”). The court has reviewed the documents cited by U.S. Steel in support of this contention, and the court cannot determine with any degree of certainty whether such costs were included or excluded. The court therefore holds that Commerce’s determination that the record was unclear as to whether these costs were already captured elsewhere is supported by substantial evidence. Furthermore, although U.S. Steel does not appear to independently challenge Commerce’s decision to employ a methodology that avoids the risk of double counting, the court holds that this decision was reasonable.

B. Warranty Claims Discovered at Verification

U.S. Steel argues that the three warranty claims discovered at verification were all “incurred” during the POI. According to U.S. Steel, Commerce’s practice is to deduct expenses incurred during the POI, but Commerce arbitrarily departed from this practice. U.S. Steel Reply at 39-40. Commerce’s practice, however, is to include only warranty claims paid within the POI in its warranty expense calculation, regardless of whether the sale or the initial claim was made during the POI. See I & D Memo at 58, 80; Gov. Br. at 124. This practices developed because “the total actual amount of warranty expenses cannot be known at the time of sale.” Id. at 58. It was reasonable for Commerce to focus on the amount paid rather than the amount claimed, as the amount claimed could change as the warranty expense was negotiated. See Gov. Br. at 125; see also Section IV.C, supra. In reaching this conclusion, Commerce followed the same methodology as it did in the Issues and Decision Memorandum for the Antidumping Duty Investigation of Narrow Woven Ribbon With Woven Selvedge from Taiwan at 28-29, A-583-844, (July 19, 2010), available at http://enforcement.trade.gov/frn/ summary/taiwan/2010-17538-l.pdf (last visited Aug. 27, 2015), where Commerce stressed that it is routine practice to require not only that a warranty claim be evidenced in a respondent’s books and records at verification, but also that the respondent actually paid the claim for the expense to be counted.

Further, the Government is correct to distinguish warranty expenses from the examples U.S. Steel cites concerning interest, production, and freight costs incurred during a POI. See Gov. Br. at 124-25. Such expenses differ from warranty expenses because, unlike warranty expenses, they are capable of calculation at the time of sale. Id. at 125. This greater degree of certainty allows Commerce to include these expenses in its calculations regardless of when they are paid. See id. Conversely, there is no guarantee that a claim filed by a customer will accurately reflect the amount eventually paid. The court holds that Commerce’s decision to exclude the three warranty claims from the warranty expense calculation was reasonable, supported by substantial evidence, and in accordance with law.

X. HYSCO’s Short-Term U.S. Interest Rate

HHU reported at the outset of verification that it had mistakenly included interest expenses related to long-term loans in the numerator of its short-term interest rate calculation, and it requested that these loans be excluded from the calculation. I & D Memo at 55. Commerce subsequently verified the correction, revised the short-term interest rate, and used the resulting figure to calculate HYS-CO’s U.S. credit expenses and inventory carrying costs. Id. at 55-56. The submission of this “minor correction,” however, revealed for the first time that HHU’s short-term borrowing involved affiliated transactions. U.S. Steel Br. at 52. U.S. Steel argued that Commerce should have used the interest expense ratio HYSCO originally reported (i.e., with the long-term loans included) as partial AFA because HYSCO had failed to reveal the role of an affiliated party in HHU’s short-term borrowings prior to verification. I & D Memo at 54. Commerce rejected this contention, explaining that relevant expenses “are an inherent part of the relationship between affiliated parties,” that there was no information on the record suggesting that the verified information should be rejected, and that HHU borrowed from unaffiliated parties. Final Sales Calculation Memorandum for HYSCO at 4, CD 432 (July 10, 2014).

U.S. Steel argues that Commerce erroneously failed to apply partial AFA when calculating the interest expense ratio for the short-term borrowings of HYSCO’s U.S. affiliate HHU. U.S. Steel Br. at 50-53. U.S. Steel argues that HYSCO failed to disclose that HHU’s short-term interest rate was determined based on transactions with an affiliated party, namely HYSCO itself. 31 Id. at 51. U.S. Steel claims that this failure interfered with Commerce’s ability to investigate whether HHU’s short-term interest rate reflected arm’s-length transactions. Id. U.S. Steel further claims that HYSCO violated its obligation to disclose all relationships with affiliates that could affect the sale or distribution of the subject merchandise, including any relationships related to “borrowings.” Id. U.S. Steel’s arguments lack merit.

The use of facts otherwise available is appropriate only when there are gaps in the record evidence and Commerce must depend on other sources to complete the record. Fine Furniture (Shanghai) Ltd. v. United States, 865 F.Supp.2d 1254, 1260 (CIT 2012). “Absent a valid decision to use facts otherwise available, Commerce may not use an adverse inference.” Shandong Huarong Mach. Co., 30 CIT at 1289, 435 F.Supp.2d at 1301.

Commerce explained that the expenses at issue are an inherent part of the relationship between affiliated parties. HYS-CO had already revealed the fact that HYSCO and HHU were affiliated, just not the particular transactions. The disclosure of these transaction, however, only confirmed what Commerce already logically presumed. U.S. Steel has not pointed to any authority suggesting that Commerce’s analysis is required to take account of these specific affiliated transactions or that Commerce normally treats such transactions as significant in determining an appropriate dumping margin. And as HYSCO explains, “because HHU obtained borrowings from unaffiliated banks, [the rates reported by HHU] reflect the market rate associated with HHU’s actual interest expense. Moreover, HYSCO’s involvement makes any resulting interest rate all the more probative of the imputed credit and inventory costs associated with HYSCO’s sales to the United States through HHU.” Hyundai HYSCO’s Resp. in Opp’n to Consol. Pis. Maverick and U.S. Steel Corp.’s Rule 56.2 Mots, for J. on the Agency R. 29-30, ECF No. 155. U.S. Steel has not shown that the interest expense ratio was miscalculated, nor has it shown that the disclosure of the particular transactions at issue is something that normally would affect Commerce’s analysis. Because U.S. Steel has failed to show that there was a gap in the record, the use of AFA is not appropriate.

Furthermore, even assuming that HYS-CO should have disclosed the affiliated transactions earlier, Commerce was justified in relying upon the verified information in the record rather than using AFA. This ease is readily distinguishable from Tianjin Magnesium International Co. v. United States, 844 F.Supp.2d 1342 (CIT 2012), upon which U.S. Steel heavily relies in support of its argument that AFA should have been used. In Tianjin, the respondent attempted to submit false voucher books after their falsity previously had been determined during a failed verification. Id. at 1347. The matter was remanded because Commerce had never addressed this conduct, which appeared designed to mislead Commerce. See id. at 1347-48. Here, Commerce verified the correction made to HHU’s short-term interest rate, and there was never any reason for Commerce to think HYSCO’s data were false. The .present case is not one where Commerce ignored the challenged action, and the analogy U.S. Steel draws between the two cases is unfounded. HYSCO could have been more explicit in disclosing the affiliated transactions associated with HHU’s short-term interest rate, but regardless of this “transgression,” it was reasonable and permissible for Commerce to decline to apply AFA, especially when the adverse facts suggested by U.S. Steel were known by Commerce to be inaccurate.

CONCLUSION

For the foregoing reasons, Commerce’s Final Determination is remanded in part for Commerce to reconsider its failure to select ILJIN as a mandatory respondent and for it to reconsider its calculation of CV profit. In all other respects, Commerce’s Final Determination is sustained. Any change to NEXTEEL’s or HYSCO’s dumping margins shall be reflected in the all-others rate assigned to Husteel, AJU Besteel, SeAH, and ILJIN (if ILJIN is not individually examined on remand). Commerce shall have until November 2, 2015, to file its remand results. The parties shall have until December 2, 2015, to file objections, and the government shall have until December 17, 2015, to file its response. Should Commerce determine on remand that individual examination of ILJIN is appropriate, however, the parties shall promptly notify the court and propose an appropriate timeframe for completion of the remand proceedings.

2 . The normal value of the subject merchandise is defined as “the price at which the foreign like product is first sold :.. for consumption in the exporting country, in the usual commercial quantities and in the ordinary course of and, to the extent practicable, at the same level of trade as the export price or constructed export price.” 19 U.S.C. § 1677b(a)(l)(B)(i) (2012). If normal value cannot be determined pursuant to 19 U.S.C. § 1677b(a)(l)(B)(i), then the constructed value of the subject merchandise may be used in place of normal value. 19 U.S.C. § 1677b(a)(4). A dumping margin is "the amount by which the normal value exceeds the export price or constructed export price of the subject merchandise.” 19 U.S.C. § 1677(35)(a).

3 . CEP is "the price at which the subject merchandise is first sold (or agreed to be sold) in the United States before or after the date of importation by or for the account of the producer or exporter of such merchandise or by a seller affiliated with the producer or exporter, to a purchaser not affiliated with the producer or exporter.” 19 U.S.C. § 1677a(b).

4 . Export price is “the price at which the subject merchandise is first sold (or agreed to be sold) before the date of importation by the producer or exporter of the subject merchandise outside of the United States to an unaffiliated purchaser in the United States or to an unaffiliated purchaser for exportation to the United States.” 19 U.S.C. § 1677a(a).

5 . [[Confidential Data Deleted ]] of the companies listed in the CBP data [[Confidential Data Deleted]].

6 . Together, HYSCO and NEXTEEL accounted for approximately [[Confidential Data Deleted]] percent of the Korean OCTG imports captured in the CBP date. Respondent Selection Memo at 5.

7 . SeAH adopted the arguments set forth by Husteel regarding mandatory and voluntary respondent selection. SeAH Br. at 5-6.

8 . The government suggests that the number of exporters or producers involved in this investigation was twenty-two, which was the total number of companies listed in the CBP data. The government argues that Commerce is not required to conduct a pre-investigation into the accuracy of the CBP data for purposes of respondent selection, as such an investigation is not contemplated by the statute and would hinder the timely completion of the proceedings. Def.’s Resp. in Opp'n to Mots, for J. upon the Administrative R. 69-70, ECF. No. 146 ("Gov. Br.”). Although this argument may have some weight in certain situations, to the extent that Commerce wishes to rely on CBP data for respondent selection, it is unreasonable for Commerce to ignore evidence on the face of that data suggesting that the actual number of potential respondents is likely less than the number of companies separately listed. The court agrees with Husteel that it is unreasonable "to suggest that it requires a ‘pre-investigation’ to determine that an exporter appearing in the. CBP data with a quantity of [[Confidential Data Deleted ]] is not a potential respondent” and that "[l]ikewise, where the same exporter appears multiple times in the same dataset, it is hardly unreasonable to expect Commerce to recognize that exporter will constitute only a single respondent." Reply Br. of PL Husteel Co., Ltd. in Resp. to Def.’s and Def.-Intvnrs. Brs. 35, ECF No. 187 (confidential version).

9 . ILJIN also argues that Commerce failed to "individually address” certain comments submitted by several companies that reviewing their data, for various reasons, would not have been particularly burdensome. ILJIN Br. at 21-22. The court rejects this argument, as ILJIN never raised this issue in its case brief to Commerce. See ILJIN Case Brief, PD 446 (June 8, 2014); Pakfood Pub. Co. v. United States, 34 CIT 1122, 1143-44, 724 F.Supp.2d 1327, 1349-50 (2010) (discussing general rule that a party must present all of its arguments in its case brief in order to exhaust its administrative remedies). The court briefly notes, however, that many of the arguments and much of the data that Commerce supposedly failed to "individually address” was not submitted until after Commerce had made its decision regarding the number of mandatory respondents. These submissions would appear to be more germane to the issue of voluntary respondent selection.

10 . According to ILJIN, seamless OCTG commands a price premium of approximately [[Confidential Data Deleted ]] percent on average over welded OCTG. Br. of Pl.-Intvnr. IL-JIN Steel Crop, in Supp. of Its Mot. for J. on the Agency R. 24, ECF No. 88-1 (confidential version).

11 . The court additionally notes that selecting a larger and/or more representative sample of exporters might mitigate potential distortions caused by Commerce's obligation to include the rates calculated for voluntary respondents when calculating an "all-others” rate for non-examined producers or exporters. See Mac-Lean-Fogg Co. v. United States, 753 F.3d 1237, 1244 (Fed.Cir.2014). Only companies with relatively lower margins are likely to request voluntary respondent status, and thus the inclusion of their rates is likely to skew the all-other’s rate, which is based on an average of the rates calculated for all individually examined respondents. 19 U.S.C. § 1673d(c)(5). By broadening the potential pool of mandatory respondents, Commerce is more likely to capture a variety of Commercial experiences, rather than being limited to the experience of the largest firms and the voluntary respondents that essentially selected themselves. Whether Commerce in the foreseeable future will accept voluntary respondents, given its view of its resources and the adverse precedent, is unknown.

12 . The court acknowledges that mandatory respondents are unlikely to match non-examined producers in all respects. Some deviation is inherent when Commerce limits the number of individually examined respondents, and Commerce might not need to provide a comprehensive response to every representativeness claim in every case. Here, however, there is an entirely distinct type of OCTG and record evidence shows that its production process and price differ significantly from the OCTG produced by the other respondents, and Commerce failed to explain why representativeness of the entire subject product is not a pertinent factor in selecting mandatory respondents.

13 . ILJIN additionally argues that Commerce should not have assigned it a rate- based on the margins found for the mandatory respondents. ILJIN Br. at 36-37. The court need not discuss this argument in detail, as it is remanding Commerce's decision to exclude ILJIN from the mandatory respondents selected. The court notes, however, that if Commerce lawfully declined to individually examine ILJIN, this argument likely would lack merit. The statute provides that Commerce "shall” calculate the margins of non-examined producers or exporters by using "the weighted average of the estimated weighted average dumping margins established for exporters and producers individually investigated.” 19 U.S.C. § 1673d(c)(5). ILJIN has not pointed to any authority suggesting that Commerce can or should depart from this statutory mandate.

14 . The court notes that before this case was argued, Congress amended 19 U.S.C. § 1677m(a). See Trade Preferences Extension Act of 2015, Pub.L. No. 114-27, § 506, 129 Stat. 362, 386-87. The statute now specifies certain factors Commerce may consider in determining whether the examination of a voluntary respondent would be "unduly burdensome.” See id. Commerce has indicated that this change will apply to its determinations issued on or after August 6, 2015. Dates of Application of Amendments to the Anti-dumping and Countervailing Duty Laws Made by the Trade Preferences Extension Act of 2015, 80 Fed.Reg. 46,793, 46,795 (Dep't Commerce Aug. 6, 2015). The court therefore analyzes the statute as it existed when Commerce issued the Final Determination in this case (i.e., July 18, 2014), but notes that its ultimate conclusion would be the same even if it considered the statute as amended.

15 . The court rejects ILJIN’s arguments regarding why fairness required it to be selected as a voluntary respondent. Commerce provided an adequate explanation for why examination of any additional respondents would have been unduly burdensome, and nothing in 19 U.S.C. § 1677m(a) suggests an extraordinary need to accommodate voluntary respondents in order to ensure that margins are representative beyond that required in mandatory respondent selection.

16 . ILJIN, SeAH, and AJÜ Besteel did not independently brief this issue beyond arguing that because their rates were based on the margins assigned to NEXTEEL and HYSCO, any change to the NEXTEEL or HYSCO margins should apply to them. ILJIN Br. at 15, 37-38; .SeAH Br. at 2-3, 6; AJU Besteel Br. at 2. Accordingly, they have adopted the arguments presented by Husteel, NEXTEEL, and HYSCO pertaining to this issue. ILJIN Br. at 15, 37-38; SeAH Br. at 2-3, 6; AJU Besteel Br. at 2.

17 . No party has suggested that Commerce could have or should have calculated a profit figure pursuant to § 1677b(e)(2)(A).

18 . Commerce determined that it could not rely on alternative (ii), because there were no other respondents subject to the investigation. I & D Memo at 15-16. The parties do not contest this conclusion.

19 . Commerce also considered and rejected using the financial statements of four Indian companies. See I & D Memo at 19-20. No party has suggested that Commerce should have used the profit contained in any of these financial statements to calculate CV profit.

20 . The court is also cognizant of the fact that Commerce's decision to use this data resulted in an apparent deviation from its prior practice in calculating CV profit. Commerce on prior occasions had relied on sales of nonOCTG pipes to calculate CV profit. See Oil Country Tubular Goods, Other Than Drill Pipe, from Korea: Preliminary Results of Antidumping Duty Administrative Review, 72 Fed.Reg. 51,793, 51,796 (Dep’t Commerce Sept. 11, 2007) (using financial statement of SeAH to calculate Husteel’s CV profit), unchanged in Oil Country Tubular Goods, Other Than Drill Pipe, from Korea: Final Results of Antidumping Duty Administrative Review, 73 Fed.Reg. 14,439 (Dep’t Commerce Mar. 18, 2008); Certain Oil Country Tubular Goods from Mexico; Preliminary Results of Antidumping Duty Administrative Review and Partial Rescission, 71 Fed.Reg. 27,676, 27,679 (Dep't Commerce May 12, 2006) ("[W]e based our profit calculations and indirect selling expenses on the income statement of Hylsa’s tubular products division, a general pipe division that produces OCTG and products in the same general cate gory.”), unchanged in Notice of Final Results and Partial Rescission of Antidumping Duty Administrative Review: Certain Oil Country Tubular Goods from Mexico, 71 Fed.Reg. 54,-614 (Dep’t Commerce Sept. 18, 2006). In fact, earlier in the proceedings, Commerce’s questionnaires to the respondents specifically referred to their standard and line pipe products as goods that were in the “same general category of products.” Department's NEXTEEL Suppl. Section D Questionnaire at 11, PD 170 (Nov. 26, 2013); Department’s HYS-CO Suppl. Section D Questionnaire at 11, PD 173 (Dec. 4, 2013). Moreover, this appears to be the first time that Commerce had relied upon a CV profit source that was not based on either production or sales in the home market. See PI. NEXTEEL’s Reply Br. in Supp. of Its Rule 56.2 Mot. for J. upon the Agency R. 8-9, ECF No. 195. The court recognizes that Commerce might have legitimate justifications for this departure, but it does not change the fact that Commerce used data that was submitted late to come to a conclusion that was seemingly at odds with its prior practice, with the result being a large increase in the respondents’ dumping margins sufficient to support an order. This is a make or break issue and Commerce should do its utmost to be fair in such circumstances.

21 . Commerce looked to NEXTEEL’s commercial relationship with its customer [[Confidential Data Deleted]]. Commerce determined that [[Confidential Data Deleted ]] NEXTEEL’s U.S. sales were made through [[Confidential Data Deleted ]], and that both of these companies were [[Confidential Data Deleted]]. Commerce extended its affiliation finding and concluded that NEXTEEL was also affiliated with [[Confidential Data Deleted]] through POSCO’s [[Confidential Data Deleted]] and NEXTEEL's supposed affiliation with POSCO. NEXTEEL Br. at 49-50; NEXTEEL Affilia tion Memorandum at 4-5, CD 443 (July 10, 2014) (“NEXTEEL Affiliation Memo”).

22 . AJU Besteel adopts NEXTEEL’s argument regarding this issue and asserts that AJU Besteel’s all-others margin should incorporate any changes made to NEXTEEL’s margin if this challenge is successful. See AJU Besteel Br. at 2.

23 . POSCO provided [[Confidential Data Delet ed1]] percent of NEXTEEL’s total POI purchases of hot-rolled coil used for OCTG production, and the remaining [[Confidential Data Deleted]] percent was used [[Confidential Data Deleted]]. NEXTEEL Affiliation Memo at 3. Hot-rolled coil from POSCO accounted for [[Confidential Data Deleted ]] percent of NEXTEEL’s POI consumption of hot-rolled coil during the POI. Id. Hot-rolled coil accounted for [[Confidential Data Deleted]] percent of NEXTEEL’s OCTG cost of manufacturing. Id. Additionally, [[Confidential Data Deleted ]] although Commerce acknowledge that [[Confidential Data Deleted J]. Id. at 3-4.

24 . To the extent that U.S. Steel relies on NEC Home Electronics, Ltd. v. United States, 18 CIT 336, 1994 WL 176914 (1994), to cast doubt on Commerce’s use of NEXTEEL’s 2010 and 2011 warranty expenses, the court notes that U.S. Steel failed to raise any issues with this data before the agency, and that in any event, U.S. Steel’s cursory argument regarding this case in its briefs is unpersuasive. Even U.S. Steel appears to recognize the limited relevance of that case in its reply brief. See Reply Br. in Supp. of PI. United States Steel Corp.’s Mot. for J. on the Agency R. Under Rule 56.2 13-14, ECF No. 190.

25 . [[Confidential Data Deleted ]].

26 . "Company A” refers to [[Confidential Data Deleted ]].

27 . As an example, Maverick argues that for one grade of hot-rolled coil, the weighted average price of the coil to POSCO’s affiliates was [[Confidential Data Deleted ]] KRW, [[Confidential Data Deleted ]] KRW for its unaffiliated customers, and [[Confidential Data Deleted ]] KRW for [[Confidential Data Deleted ]]. Maverick Br. at 27.

28 . It is Maverick's view that the price charged to the [[Confidential Data Deleted]] of its affiliate, NEXTEEL, is the only reasonable benchmark price, yet it is not clear why Maverick concludes this is true. See NEXTEEL's Resp. at 48. As NEXTEEL points out, POSCO may have decided to charge a premium for its sales to [[Confidential Data Deleted ]] of one of its affiliates, which would make this price more of a marketplace outlier. See id. It is equally plausible that any differences in the prices POSCO charged its customers were due to [[Confidential Data Deleted ]] as argued by NEXTEEL. See id. Furthermore, although [[Confidential Data Deleted]] was NEXTEEL's [[Confidential Data Deleted ]], Maverick’s argument does not explain why POSCO would be willing to sell hot-rolled steel at favorable prices to [[Confidential Data Deleted]].

29 . The affidavit states that POSCO [[Confidential Data Deleted]]. There is no indication that the agreement was only with certain Korean producers or that POSCO discriminated in its prices to Korean producers. U.S. Steel’s Comments re: POSCO's Questionnaire Resp. at Ex. 3, CD 304 (Mar. 21, 2014).

30 . Maverick made additional arguments involving a comparison of the prices POSCO charged to certain unaffiliated Korean OCTG producers and the pricing data for Korean hot-rolled coil from MEPS International Steel Review that was included in the petition. See Maverick Br. at 26. As the government notes, Maverick failed to make any arguments based on the MEPS data before Commerce, and Maverick failed in its reply brief to suggest any exception to the generally applicable rule that all arguments must be presented first to the agency. Gov. Br. at 85. The court will not consider this argument. See 28 U.S.C. § 2637(d) (providing that the court “shall, where appropriate, require the exhaustion of administrative remedies”).

31 . Documents obtained at verifications revealed that [[Confidential Data Deleted]]. U.S. Steel Br. at 7-8. The documents showed that HHU [[Confidential Data Deleted]]. Id. at 51.