Full opinion text
Opinion CARMAN, Judge. This matter comes before the Court, on a motion for judgment on the agency record brought by Plaintiff, SeAH Steel Corporation (“SeAH”), pursuant to Rule 56.2 of the Rules of the United States Court of International Trade (“USCIT”). Plaintiff, challenges numerous aspects of the United States Department of Commerce’s (“Commerce” or “Department”) administrative determination with respect to Certain Welded Stainless Steel Pipes From the Republic of Korea: Final Results of Antidumping Duty Administrative Review, 74 Fed.Reg. 31,242 (June 30, 2009), Public Record Doc. No. 77 {“Final Results”). SeAH contends that certain findings made by Commerce are unsupported by substantial evidence or otherwise not in accordance with the law. {See PL’s R. 56.2 Mot. for J. Upon Agency Rec. (“PL’s Brief.”).) SeAH’s motion is opposed by Commerce, as well as Defendant-Intervenor, Bristol Metals. Whereas DefendanWntervenor urges the Court to affirm, in their entirety, Commerce’s Final Results {see generally Resp. Brief of Def.-Int. in Opp’n to PL’s Mot. for J. On the Agency R. (“Def.-Int.’s Brief’)), Defendant requests that the Court sustain its findings with regard to its calculation of normal value and costs of production, but requests voluntary remand of its major input and transactions disregarded findings, {see Def.’s Mem. in Opp’n to PL’s Mot. for J. Upon the Agency R. at 39 (“Def.’s Brief’)). For the reasons set forth below, the Court sustains Commerce’s Final Results in part, and remands them in part. Jurisdiction The Court has jurisdiction over this matter pursuant to 19 U.S.C. § 1516a(a)(2) and 28 U.S.C. § 1581(c). Standard of Review When reviewing the final results of anti-dumping administrative reviews, “[t]he court shall hold unlawful any determination, finding, or conclusion found ... to be unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i). “Substantial evidence is more than a mere scintilla.” Consol. Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938). “Substantial evidence is ‘such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.’ ” Huaiyin Foreign Trade Corp. (30) v. United States, 322 F.3d 1369, 1374 (Fed.Cir.2003) (quoting Consol. Edison Co., 305 U.S. at 229, 59 S.Ct. 206). In determining the existence of substantial evidence, a reviewing Court must consider “the record as a whole, including evidence that supports as well as evidence that ‘fairly detracts from the substantiality of the evidence.’ ” Huaiyin, 322 F.3d at 1374 (quoting Atl. Sugar, Ltd. v. United States, 744 F.2d 1556, 1562 (Fed.Cir.1984)). The possibility of drawing two inconsistent conclusions from the evidence “does not prevent an administrative agency’s finding from being supported by substantial evidence.” Consolo v. Federal Maritime Comm’n, 383 U.S. 607, 620, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1966) (citations omitted). There must be a “rational connection between the facts found and the choice made” in an agency determination if it is to be characterized as supported by substantial evidence and otherwise in accordance with law. Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962). Background Dumping takes place when goods are imported into the United States and sold at a price lower than their normal value. 19 U.S.C. § 1677(34). Under the statute, Commerce is required to impose duties on dumped merchandise to offset the effects of dumping. § 1673. The antidumping statutes provide for periodic administrative reviews of antidumping duty orders—at the request of an interested party—to update the applicable antidumping duty rate. § 1675. The case at bar challenges the results of such an administrative review. In January 2008, at the request of Bristol Metals, Commerce initiated a periodic administrative review of the anti-dumping duty order currently in place for welded stainless steel pipes (“WSSP”) from Korea for the period December 1, 2006 through November 30, 2007. Initiation of Antidumping and Countervailing Duty Administrative Reviews and Requests for Revocation in Part, 73 Fed. Reg. 4,829 (Jan. 28, 2008); see also Anti-dumping Duty Order and Clarification of Final Determination: Certain Welded Stainless Steel Pipes From Korea, 57 Fed.Reg. 62,301 (December 30, 1992). In the Preliminary Results, Commerce followed its conventional methodology of using SeAH’s period of review annual weighted-average costs of production to determine the appropriate dumping margin, and preliminarily calculated a rate of 4.10%. Preliminary Results, 73 Fed.Reg. at 79,052, 79,054. Shortly thereafter, the Department requested that SeAH provide quarterly cost information “in order to analyze the magnitude of cost changes throughout the POR” to determine whether it was appropriate to use shorter cost averaging periods for the Final Results. On April 29, 2009, Commerce issued its post-preliminary calculations of SeAH’s cost of production and constructed value information, which was based on an analysis of SeAH’s quarterly cost information. (Memorandum from Gina Lee, to Neal M. Halper, Proposed Adjustments to the Cost of Production and Constructed Value Information, PR 63, CR 31 (“Proposed Cost Adjustments Memo”).) The decision to proceed with a quarterly cost analysis had implications beyond Commerce’s normal value calculations, and affected how the Department conducted its price-to-price comparisons between home market and U.S. sales, as well as its statutorily mandated cost recovery methodology. Based on its evaluation of the quarterly cost information, Commerce made an upward adjustment of SeAH’s dumping margin prior to completion of the Final Results. (Id.) Following publication of the post-preliminary calculations, SeAH’s advocacy before Commerce focused principally on the issues in dispute in this action: (1) whether Commerce erred in using quarterly cost information rather than annual period of review average costs in the agency’s normal value calculations; (2) whether Commerce erred in its decision not to apply its normal “90/60” day window period for comparing U.S. and home market sales, and instead to make comparisons within a given quarter; (3) whether Commerce’s adjusted cost-recovery methodology was consistent with its statutory mandate; and (4) whether Commerce erred in its application of the “transactions disregarded” and “major input” rules. (See generally Brief from Law Firm of Troutman Sanders (“Case Brief’), PR 68, CR 34.) In the Final Results, Commerce rejected SeAH’s arguments on all the issues, and further adjusted its dumping margin to 9.05%. Final Results, 74 Fed.Reg. at 31,243. This action followed, contesting the Department’s determination in the Final Results. Plaintiff asserts that calculating its production costs on a quarterly basis results in an improper inflation of its dumping margin. (See Pl.’s Brief at 12.) SeAH further contends that the Department’s application of the major input and transactions disregarded provisions of the statute had a similar effect. (Id. at 10.) As a result, Plaintiff requests that this matter be remanded to Commerce with instructions to recalculate SeAH’s dumping margin using annual weighted average costs, and to conduct its major input analysis on a grade and specification basis. (Pl.’s Brief at 50.) Discussion I. Commerce’s Use of Quarterly Costs Versus Period of Review Average Costs for the Cost of Production Analysis A. Statutory Scheme In an administrative review, Commerce determines the antidumping duties to be imposed by first calculating the dumping margin for each of a foreign producer or exporter’s individual U.S. transactions, which is the amount by which the normal value of the imported subject merchandise in the exporter’s home market exceeds the export price or the constructed export price of that merchandise. See 19 U.S.C. § 1677(35)(A). Normal value is the basic conceptual focus in deriving the foreign benchmark value for the subject merchandise. The relevant portions of the statute require normal value to be calculated as follows: In determining under this subtitle whether subject merchandise is being, or is likely to be, sold at less than fair value, a fair comparison shall be made between the export price or constructed export price and normal value. In order to achieve a fair comparison with the export price or constructed export price, normal value shall be determined as follows: (1) Determination of normal value (A) In general The normal value of the subject merchandise shall be the price described in subparagraph (B), at a time reasonably corresponding to the time of the sale used to determine the export price or constructed price.... (B) Price The price referred to in subparagraph (A) is— (i) the price at which the foreign like product is first sold (or, in the absence of a sale, offered for sale) for consumption in the exporting country, in the usual commercial quantities and in the ordinary course of trade and, to the extent practicable, at the same level of trade as the export price or constructed export price.... § 1677b(a). The preferred, and by far most common, method for making the determination of normal value is through the use of sales of the subject merchandise in the home market of the exporter/producer. See 19 C.F.R. § 351.404(a). Sales made in the home country for less than the cost of production, however, may be disregarded in the calculation of normal value. 19 U.S.C. § 1677b(b). This exclusion may significantly raise the ultimate normal value established by Commerce since, when it is applied, it eliminates the lowest of the producer’s home market sales from the data used. A higher normal value, of course, results in an increased dumping margin. The “sales below cost” provision of § 1677b(b) thus assumes an important role in many dumping determinations. The statute permits Commerce to disregard home market sales below cost only under certain circumstances: Whenever the administering authority has reasonable grounds to believe or suspect that sales of the foreign like product under consideration for the determination of normal value have been made at prices which represent less than the cost of production of that product, the administering authority shall determine whether, in fact, such sales were made at less than the cost of production. If the administering authority determines that sales made at less than the cost of production— (A) have been made within an extended period of time in substantial quantities, and (B) were not at prices which permit recovery of all costs within a reasonable period of time, such sales may be disregarded in the determination of normal value. Whenever such sales are disregarded, normal value shall be based on the remaining sales of the foreign like product in the ordinary course of trade. If no sales made in the ordinary course of trade remain, the normal value shall be based on the constructed value of the merchandise. 19 U.S.C. § 1677b(b)(l). The statute further provides that the “cost of production” should be an amount equal to the cost of materials, fabrication, general and administrative expenses, and packaging during a period of time which would “ordinarily permit the production of the foreign like product in the ordinary course of business.” § 1677b(b)(3). If the Department determines that sales below the cost of production should be excluded, the remaining sales will be used to determine normal value in the foreign market. § 1677b(b)(l). In establishing whether a particular sale was made at less than the cost of production, “Commerce’s normal practice is to use annual averages when conducting its cost of production analysis.” Def.’s Brief at 8. That is to say that Commerce applies a “cost test” that involves a comparison of the home market sales price of a particular model to that model’s annual weighted average cost of production for the period of review. See Antidumping Methodologies for Proceedings that Involve Significant Cost Changes Throughout the Period of Investigation (POI)/Period of Review (POR) that May Require Using Shorter Cost Averaging Periods; Request for Comment, 73 Fed.Reg. 26,364 (May 9, 2008) (“Request for Comment”). Commerce, however, also has a long-standing practice of departing from annual averaging and employing shorter (usually quarterly) cost-averaging periods when two factors are present: (1) consistent and significant cost variation during the period of review, and (2) evidence of linkage between the cost variation and changes in sales prices within the shorter averaging period. Id. Here, Plaintiff contests the manner in which Commerce conducted the two-prong test, arguing that Commerce abruptly (1) changed the manner in which it determined that cost variation was significant, and (2) allowed the requirement for direct linkage between cost variation and price changes to be satisfied on a much looser correlation standard. B. Parties’ Arguments SeAH argues that Commerce traditionally uses shorter cost averaging periods only when record evidence clearly shows (1) a significant and consistent increase in costs during the period of review and (2) that rising costs can be “directly linked” to sales in the shorter cost averaging period. (PL’s Brief at 10-11.) Plaintiff alleges that Commerce departed from this established practice and thus its decision to apply quarterly cost averaging was “unreasonable and not in accordance with law.” (Id. at 11.) Plaintiff maintains that Commerce’s two prong test for use of shorter cost averaging periods, in effect at the time SeAH’s administrative review began, should have been used in this review. (Id. at 14.) SeAH references the Department’s Request for Comment as evidence of its long-standing practice with regard to the use of quarterly cost averaging. (Id. at 17 n. 9) (citing Request for Comment, 73 Fed.Reg. at 26,366.) The Department’s change in practice, Plaintiff asserts, results in a test that no longer requires the change in costs to be consistent, but simply requires the costs to increase significantly (i.e., greater than 25%) between any two quarters of the period of review. (Pl.’s Brief at 21.) SeAH’s objection is that “even if the significant change in COM represented just a temporary spike between two quarters ... quarterly costs would be applied to the entire POR, Le., even as to quarters not impacted by significant changes.” (Id.) The Department’s new test, argues SeAH, contradicts the rationale underlying Commerce’s previous practice, which was “based on the fact that short-term cost fluctuations are mitigated by the use of annual average costs and that resorting to quarterly or monthly costs in cases where the cost changes were not consistent across the POR could cause aberrations.” (Id. at 22.) This, says Plaintiff, constitutes an abrogation of the consistency requirement, for which Commerce failed to offer an adequate explanation. (See id.) SeAH further complains that Commerce has “abandoned its direct linkage requirement in favor of a watered-down test” that is met so long as costs and prices in a given quarter generally trend in the same direction. (Id. at 24.) SeAH claims, once again, that the explanation offered by Commerce is insufficient and runs counter to the Department’s prior concerns about direct linkage. (See id.) Finally, SeAH characterizes as flawed the Department’s analysis of quarterly average price and cost changes for the five largest U.S. and home market control numbers (“CONNUMs”). Plaintiff maintains that Commerce’s examination of the CONNUMs demonstrates that there is no correlation between the raw material costs and the directly related sales transactions occurring in the third quarter (the only quarter with significant cost variation). (See id. at 25.) Therefore, without a clear link between changes in third quarter costs and changes in sales prices within that same quarter, Commerce’s decision to use quarterly costs in place of annual average costs is unsupported by substantial evidence and otherwise not in accordance with law. (See id. at 25-26.) In response, Commerce acknowledges that its general practice is to use annual averages when conducting its cost of production analysis. The Department goes on to note, however, that it has departed from this practice in cases where the agency has concluded that, because of significant cost or home market price changes during the period of review, application of an annual average cost period would be distortive. (Def.’s Brief at 9-10.) As a result, both Defendant and Defendant-Intervenor challenge SeAH’s assertion that Commerce has deviated from its previous practice. (Id. at 10; Def.-Int.’s Brief at 8.) According to Commerce, it has consistently determined that it may depart from its normal methodology and review shorter cost periods when two factors exist: 1) the cost changes throughout the period of review are significant, and 2) sales during the shorter cost averaging period could be accurately linked with the cost of production during the same averaging period. (Def.’s Brief at 10.) In this review, as in prior administrative reviews, the agency’s determination of whether a cost change was significant was made by calculating the difference between the low quarterly average cost of manufacture and the high quarterly average cost of manufacture. (Id. at 12.) If this figure exceeds 25 percent, as is the case here, Commerce considers the significant cost change threshold to be met. (Id.) For example, the Department points to the “dramatic fluctuations” in the prices of nickel and hot-rolled coils during the period of review as evidence of the significant increase in respondent’s cost of manufacturing. (Id. at 13.) With regard to the second of the two inquiries, the Department claims that the agency’s definition of linkage does not require direct traceability between specific sales and specific production costs. (Id. at 16.) Rather, the standard is whether Commerce identifies pricing data which indicate “that both prices and costs were trending in the same direction” throughout the period of review. (Id. at 17.) The Department cites to its examination of the top five CONNUMs sold in the U.S. market and home country market and states that, because in “every instance but three, the change in the average quarterly cost trended consistently with the change in the average quarterly prices ... a reasonable correlation can be found between rising costs of manufacturing and sales prices.” (Id. at 19-20 (citation omitted).) Moreover, Commerce claims, these data demonstrate SeAH’s ability to revise its prices in response to the fluctuations in material costs, and points to SeAH’s reported inventory turnover periods for raw materials and finished goods being within the quarterly cost averaging period used by Commerce. (Adjustments to the Cost of Production and Constructed Value Information for the Final Results (“Final Cost Adjustments Mem.”) at 3-4, PR 75, CR 37.) From this, Commerce concluded that SeAH was able to respond to the volatility in material costs and adjust its sales prices accordingly within a given quarter. (Def.’s Brief at 19.) Commerce denies that the agency changed its methodology, but argues that it was, nonetheless, permitted to do so by virtue of its legislative mandate. Because § 1667b(b)(3) does not dictate the method for calculating the cost of production, nor does it provide a definition of the term “period,” Commerce was permitted to revise its methodology as long as it complied with the statute’s notice provisions. (See Def.’s Brief at 12-13 (citing SKF USA v. United States, 537 F.3d 1373 (Fed.Cir.2008); 19 U.S.C. §§ 1677b and 1677m(g)).) In so arguing, the Department relies on the deference a court must afford an agency’s reasonable interpretation of a statute, if that statute is silent on a particular methodology to be employed. (See Def.’s Brief at 12; SKF USA, 537 F.3d at 1381-82.) Moreover, the Department asserts, SeAH was given more than adequate notice that a shortened review period for costs might be used, and cites as evidence of this notice Plaintiffs ability to comment before the Final Determination was made. (Def.’s Brief at 12-13.) Defendant-Intervenor refutes Plaintiffs assertion that there is no correlation between raw material costs and the sales transactions occurring within the same quarter, and offers as support for this position SeAH’s ability to quickly pass on changes in the cost of manufacturing to its buyers through higher prices. (Def.-Int.’s Brief at 18-19.) This, according to Bristol Metals, was in large part due to the manner in which SeAH purchased its inputs, which permitted SeAH to easily identify its increases in costs and quickly pass such costs on to its customers. (Id. at 19.) The ease with which SeAH was able to respond to cost increases, says Defendant-Intervenor, is proof of a “near lockstep correspondence” between costs and prices. (Id.) C. Analysis Commerce is generally at liberty to discard one methodology in favor of another when necessary to calculate a more accurate dumping margin, subject to two important considerations. See SKF USA Inc. v. United States, 31 CIT 951, 491 F.Supp.2d 1354, 1362 (2007) (“[I]t is within Commerce’s expertise and discretion to update its methodology for both increased accuracy and ease of use”). The first restriction is that Commerce may not alter its methodology where a respondent has detrimentally relied on an old methodology used in previous reviews. See Fujian Mach. & Equip. Import & Export Corp. v. United States, 25 CIT 1150, 1169-70, 178 F.Supp.2d 1305, 1327 (2001). Second, Commerce must explain the basis for its change of methodology and demonstrate that its explanation is in accordance with law and supported by substantial evidence. See id. Plaintiffs challenge to Commerce’s use of quarterly cost averages in place of the longer annual averages fails for several reasons. In the first instance, SeAH has not presented nor attempted to present an argument based on detrimental reliance. While Commerce “may not make minor disruptive changes in methodology where a respondent demonstrates its specific reliance on the old methodology,” id., the party claiming the benefit of this rule must show detrimental reliance on the previous methodology, see NSK Ltd. v. United States, 21 CIT 617, 639, 969 F.Supp. 34, 56 (1997), aff'd in part, rev’d in part on separate grounds sub nom. NSK Ltd. v. Koyo Seiko Co., Ltd., 190 F.3d 1321 (Fed.Cir.1999). Instead, Plaintiffs argument focuses on the second of the two requirements, specifically that Commerce failed to adequately explain its change in practice. This, however, ignores the Court’s previous rulings, and vitiates a wealth of controlling authority. Even assuming that Plaintiffs argument was properly framed, SeAH’s contention that the application of Commerce’s quarterly cost methodology was unlawfully retroactive, is similarly flawed. There is an inherent retroactivity to antidumping administrative review determinations, and “[c]hanges in methodology, like all other antidumping review determinations, permissibly involve retroactive effect.” SKF USA, 537 F.3d at 1381 (internal citations and quotation marks omitted); see also American Permac, Inc. v. United States, 10 CIT 535, 539, 642 F.Supp. 1187, 1191 (CIT 1986) (stating that “19 U.S.C. § 1675(a)(2) expressly calls for the retrospective application of antidumping review determinations”). Having failed to establish detrimental reliance on Commerce’s previous practice, and given the inherently retroactive nature of the antidumping statutory scheme, Plaintiffs arguments fail to establish that Commerce improperly changed its cost production methodology. As a threshold matter, it is hardly clear that Commerce has in fact changed its methodology at all. As the DefendantIntervenor notes, Commerce has applied quarterly costs in the same manner as used in the present case in several other administrative reviews. (Def.-Int.’s Brief at 6-7.) The Department’s approach in these past proceedings are representative of the agency’s long-standing and well-recognized test for use of alternative cost averaging periods. What is significant, however, is the Department’s interpretation of the central terms of the test’s two requirements. Commerce explained in the Issues & Decision Memo that: The Department has articulated in several past proceedings that the use of an alternative cost averaging period may be appropriate in situations where a reliance on our normal annual weighted average cost method would be distortive due to significant cost changes.... [W]e recognize the importance of having a consistent and predictable approach to analyzing the issue and determining when to deviate from our normal annual average cost methodology.... The Department conducted a careful review of the comments received in response to the [Request for Comment ]. We also considered interested party comments on the same issue in [Turkish Rebar 2008 ], [Sheet from Mexico ] and [Plate from Belgium ], and reaffirmed in the final results of these cases that the two most important factors in considering whether to deviate from our normal average cost methodology are 1) whether the cost changes throughout the POI or POR were significant, and 2) whether sales during the shorter cost averaging period could be accurately linked with the COP during the same averaging period. (Issues & Decision Memo at 5-6.) In these prior determinations, Commerce established a presumption that costs are deemed to vary significantly when the range between the quarterly costs of manufacture of the subject merchandise exceeds 25 percent. (Id. at 9.) More importantly, however, Commerce computed (on a CONNUM specific basis) the percentage difference between the low quarterly average cost of manufacture and the high quarterly average cost of manufacture. (Id.) If the percentage difference exceeded 25 percent, Commerce deemed “the significant cost change threshold to be met.” (Id.) In two of the cases cited by Commerce, the agency also introduced, under the linkage requirement, its “reasonable correlation” analysis. (Id. at 12 (citing Sheet from Mexico, and Plate from Belgium).) Thus, Commerce’s methodology had now begun to rely on cost changes that were measured through differences between quarters and were linked to prices by way of a reasonable correlation. Plaintiff characterizes Commerce’s approach as one that has “abandoned the requirement that changes in costs be consistent over the POR and ... eliminated the requirement of a direct link between costs and prices.” (Pl.’s Brief at 18.) The Court disagrees. While it is true that the Department has shifted position as to the requisite relationship between production costs and sales prices, it has consistently rejected a direct traceability requirement since early 2008. (See Def.’s Confidential App. Tab F, Nucor Corp. v. United States, Final Results of Redetermination Pursuant to Court Remand, at 14 (“There is no requirement of direct traceability between specific sales and their specific production costs to prove linkage in the Department’s practice.”), available at http://ia.ita.doc. gov/remands/09-20.pdf (last visited May 3, 2010); Def.’s Confidential App. Tab G, Habas Sinai v. United States, Final Results of Redetermination Pursuant to Court Remand, at 13 (same), available at http://ia. ita.doc.gov/remands/09-55.pdf (last visited May 3, 2010); Plate from Belgium Issues & Decision Memo at 18, available at http:// ia.ita.doc.gov/frn/summary/belgium/E829410-l.pdf (last visited May 3, 2010) (“Our definition of linkage in the instant case does not require direct traceability between a specific sale and its specific production cost, but rather relies on whether there are elements which would indicate a reasonably positive correlation between the underlying costs and the final sales prices”); Sheet from Mexico Issues & Decision Memo at 21 (same), available at http://ia.ita.doc.gov/frn/summary/ mexico/E9-2667-l.pdf (last visited May 3, 2010); Issues & Decision Memo at 13 (“As noted, our definition of linkage does not require direct traceability between specific sales and their specific production costs.”).) As Commerce noted in the Issues & Decision Memo, “the Department has approached its consideration of linkage between sales and costs in various ways and to varying levels of precision.” (Issues & Decision Memo at 13.) The statute “does not dictate the method by which Commerce may calculate costs of production, nor ... define the term ‘period,’ ” and Commerce is afforded considerable discretion in formulating its practices in this regard. (Def.’s Brief at 9.) Commerce has not deviated from the application of its two-prong test for deciding whether to resort to shorter cost averaging periods. All it has done, in this case, is exercise its discretionary authority to more clearly define the significance and linkage thresholds—factors for which the Department elicited suggestions in the Request for Comment. (See 73 Fed.Reg. at 26,367.) Even had Commerce changed its methodology, the statute only requires the agency to provide the affected parties with notice and an opportunity to comment before the final determination is made. See Shikoku Chem. Corp. v. United States, 16 CIT 382, 388-89, 795 F.Supp. 417, 421-22 (1992) (finding that principles of fairness can prevent Commerce from changing its methodology without adequate notice). The Department was only obligated to notify Plaintiff prior to its final determination in this matter, and it did. Koyo Seiko Co. v. United States, 31 CIT 1512, 1520, 516 F.Supp.2d 1323, 1334 (2007). In sum, Plaintiff has failed to establish that Commerce’s use of an alternative cost averaging period was unreasonable, or constituted an abrupt change in methodology. For the reasons stated above, the Court finds that Commerce’s decision to depart from its general practice of using an annual cost averaging period, and to instead rely on quarterly costs, was supported by substantial evidence and otherwise in accordance with law. 2. Commerce’s Cost Recovery Methodology A. Statutory Scheme As previously noted, below cost sales may be excluded from the calculation of normal value only if they “have been made within an extended period of time in substantial quantities,” and “were not at prices which permit recovery of all costs within a reasonable period of time.” 19 U.S.C. § 1677b(b)(l)(A), (B). In determining whether such sales are at prices which permit the recovery of costs, the statute further provides: If prices which are below the per unit cost of production at the time of sale are above the weighted average per unit cost of production for the period of investigation or review, such prices shall be considered to provide for the recovery of costs within a reasonable period of time. 19 U.S.C. § 1677b(b)(2)(D). It is normally the case that Commerce calculates a weighted average per unit cost on an annual basis in order to make this comparison. As long as the respondent’s sales price is above that annual weighted average per unit cost, the costs are considered to be recovered, and thus, included in the calculation of normal value. In this way, the cost recovery test accounts for fluctuations in costs throughout the period of review (which covers one year). In the underlying administrative review, however, Commerce determined that the calculation of an unadjusted annual weighted average per unit cost would not smooth out the fluctuations in costs, but would result in significant distortions in the cost recovery test. (Issues & Decision Memo at 19.) At issue is the indexing methodology Commerce employed in calculating the weighted average per unit cost of production for the period of review. B. Parties’ Arguments SeAH alleges that Commerce ignored the requirements of section 1677b(b)(2)(D) when it “calculated a distinct CONNUMspecific COP for each quarter of the POR and then compared home market sales prices to the COP for the quarter in which the sale was made,” after which Commerce excluded from the calculation of normal value those sales “whose prices were below that quarterly weighted average per-unit COP,” regardless of whether such prices were above the weighted average per-unit cost for the period of review. (Pi’s Brief at 30 n. 12.) The failure to apply a weighted average per unit cost for the entire period of review resulted in the exclusion of “all sales that were found to be below cost based on a comparison of the selling price to the restated quarterly weighted-average per unit costs calculated by Commerce.” {Id,, at 30 (emphasis in original).) This, according to Plaintiff, is contrary to the clear and unambiguous statement by Congress that the cost recovery test must compare prices to a weighted average cost for the entire period of review. {See id. at 31.) As support for this position, Plaintiff cites to the relevant portions of the legislative history which Plaintiff claims confirm “Congress’s intent that the cost recovery test is to be based exclusively on POR weighted average costs.” {Id. at 30.) Plaintiff further argues that, as correctly interpreted, the statute provides for only one cost recovery test “in all circumstances.” (Pl.’s Brief at 31) (citing Acciai Speciali Terni S.P.A v. United States, 25 CIT 245, 274, 142 F.Supp.2d 969, 997 (2001).) In addition, SeAH points to earlier administrative proceedings in which Commerce “expressly and routinely recognized that the statute required Commerce to conduct the cost recovery test using POR (or POI) weighted average costs even when it had determined to otherwise calculate COP using quarterly (or even monthly) weighted-average costs.” (PL’s Brief at 32 (citing Dynamic Random Access Memory Semiconductors of One Megabit or Above from Taiwan, 64 Fed.Reg. 28,983, 28,988 (May 28, 1999).) Therefore, because Commerce has failed to apply the cost recovery test in a manner consistent with the plain language of the statute, its quarterly cost determination is unsupported by substantial evidence and otherwise not in accordance with law. (PL’s Brief at 33.) Essentially, Defendant does not disagree with Plaintiffs interpretation of the statute’s requirements, but rather contests Plaintiffs characterization of its methodology as one that deviates from the statutory mandate. (Def.’s Brief at 35.) According to Commerce, because of the significant changes in SeAH’s costs during the period of review, and its determination to use a quarterly cost averaging period, the Department concluded that “it must adjust its normal cost-recovery methodology to account [for] the distortive effect of significant cost changes.” {Id. at 33.) Commerce insists, however, that this change in methodology did not alter the “weighted average per unit cost required by the statute.” {Id.) The Department explains that if it were to use “an unadjusted weighted average per unit cost for the POR for purposes of the cost recovery test, sales prices which were determined to be below cost may be erroneously considered to have recovered costs based simply on the timing of the sale.” (Issues & Decision Memo at 19.) While Commerce concedes that, in most instances, application of an unadjusted weighted average is proper, Commerce determined that the volatility in SeAH’s cost of manufacturing required a quarterly indexing of SeAH’s costs in order to “neutralize” the distortive effects of these fluctuations. (Def.’s Brief at 33-34.) This, Commerce claims, was entirely consistent with the terms of the statute in that the Department continued to apply a period-wide weighted average per unit cost, although within that framework it incorporated a quarterly indexing methodology. (Id. at 35.) As argued by Commerce, the statute does not limit the agency’s calculations “to a simple weighted annual average that fails to take into consideration significant changes in the cost of production.” (Id. at 37.) Rather, it provides for the “rehabilitation” of sales below cost if those sales prices are above a weighted average per unit cost of production for the period of review. (Id.) Commerce interprets section 1677b(b)(2)(D) as providing the Department with the authority to “consider relevant factors during the period of review that would result in the use of costs that ‘reasonably reflect’ SeAH’s costs of production.” (Id.) C. Analysis The nature of the parties’ disagreement focuses not on divergent interpretations of the antidumping statute, but rather on whether or not Commerce’s actions comport with the statute’s substantive requirements. Both parties agree that section 1677b(b)(2)(D) requires that below cost prices, found to be above the weighted average costs for the period of review, are considered to provide for the recovery of costs within a reasonable period time. (See Pl.’s Brief at 29; Def.’s Brief at 35.) Thus, those sales are to be used in the calculation of normal value. However, whereas Commerce describes its practice of applying an indexed weighted average per unit cost of production for the period of review as being consonant with this provision, Plaintiff characterizes the quarterly indexing methodology as unlawful. In order for the Court to make a determination on this issue, it must do so solely on the grounds invoked by the agency. “If those grounds are inadequate or improper, the court is powerless to affirm the administrative action by substituting what it considers to be a more adequate or proper basis.” SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947). Further, if “the administrative action is to be tested by the basis upon which it purports to rest, that basis must be set forth with such clarity as to be understandable.” Id. Thus, the Court cannot “be expected to chisel that which must be precise from what the agency has left vague and indecisive.” Id. at 197. Applying this rule and its corollary, the Court cannot sustain the Department’s use of a quarterly-based indexing adjustment of SeAH’s weighted average costs for the cost recovery test. The Court is unable to discern from the record here the precise manner in which Commerce “adjusted” its cost recovery methodology to comply with both the statute’s requirement of weighted average costs for the period of review and Commerce’s belief that such an analysis, left unadjusted, would distort the cost recovery test results. To be sure, both parties place great emphasis on the effect of Commerce’s quarterly-based adjustment process, but neither provides an adequate explanation of its underpinnings. It is, however, up to Commerce to justify its determination with a reasoned explanation that is supported by substantial evidence on the record. While Commerce has asserted that its adjusted cost recovery methodology fully complies with the requirements of § 1677b(b)(2)(D), the Court must examine the basis upon which this conclusion was drawn. In other words, the Court must determine whether Commerce is correct in arguing that its quarterly indexing methodology conforms to the statute’s requirement of a “weighted average per unit cost of production for the period of investigation or review.” 19 U.S.C. § 1677b(b)(2)(D). As Commerce explains the methodology, the agency: first computed indices for each quarter of the period of review that reflected the relative cost of the hot-rolled coils for each grade that SeAH used to produce stainless steel pipe. These indices were ratios comparing the weighted-average cost of hot-rolled coils used in each quarter to a base quarter during the period of review. Using the calculated indices, Commerce then restated each reported quarter’s CONNUM-specific average hot-rolled coil cost to a single quarter’s “constant cost level.” After this restatement of costs, Commerce extended the four quarters’ restated costs by SeAH’s respective quarterly production quantities and calculated a “restated annual average direct material cost,” by weight-averaging the “constant cost levels” on an annual basis. Once again using the computed quarterly indices, Commerce then restated the calculated “constant cost level” annual weighted-average cost of hot-rolled coils back to each quarter’s calculated “cost levels.” (Def.’s Brief at 34 (citations omitted).) 1. Cost Recovery Analysis The core question as to cost recovery appears to be this: did Commerce’s quarterly indexing adjustments produce a “weighted average per unit cost of production for the period of ... review” as required by 19 U.S.C. § 1677b(b)(2)(D)? The Court finds itself unable to answer this core question on the record here. The problem is two-fold. a. Inadequate Explanation of Cost Recovery Test First, the Court finds Commerce’s explanation of its methodology lacking. Although Commerce explained broadly that the methodology “addressed, and attempted to neutralize, the distortive affect [sic] of significantly changing hot-rolled coil costs,” Commerce did not adequately explain why it implemented this methodology via the particular “multi-part analysis” employed here. (See Def.’s Brief at 33-34.) For example, using quarterly indices, Commerce restated each quarter’s CON-NUM-specific average materials costs into a “constant cost level,” then weighted those restated figures by quarterly production quantities to produce a “restated annual average direct material cost,” then restated those figures “back to each quarter’s calculated ‘cost levels.’ ” (Id. at 34.) Commerce has failed to describe how this specific, rather complex, mechanism of calculating the benchmark weighted average per unit cost of production “addressed” and served to “neutralize” the distortion caused by significant changes in hot-rolled coil costs. (Id. at 33-34.) Not only that, but Commerce has simply asserted, without adequate explanation, that the chosen methodology did not effectively substitute quarterly weighted averages for the period of review-wide weighted average required by the statute. Without an explanation as to why and how this particular methodology reconciles Commerce’s preferred quarterly cost of production examination with the period of review-wide cost of production examination called for in 19 U.S.C. § 1677b(b)(2)(D), the Court cannot ensure that Commerce conducted the cost recovery test in accordance with law. b. Inadequate Record of Calculations Used Commerce’s failure to adequately explain its methodology was compounded by the inclusion in the record of only a limited amount of the underlying data resulting from its calculations. For example, Commerce cites the Proposed Cost Adjustments Memo for support of its methodology. (See Def.’s Brief at 33-34.) While that document describes the functions used in Commerce’s quarterly indexing methodology, it does not provide a complete record of the representative calculations. (See Proposed Cost Adjustments Mem. at 3-5 (describing the proposed methodology later employed by Commerce), Att. 3 (containing only samples of the results produced by the methodology); compare Final Cost Adjustments Mem. at 5-6 (methodology unchanged) and Atts. 1 & 2 (spreadsheets of results of methodology not provided).) Insofar as the Department has acknowledged a deviation from its “standard cost-recovery test” (Def.’s Brief at 37), the sparse record here does not allow the Court to determine whether Commerce’s application of the new cost recovery test in fact complied with the statute. While the data in the record illustrates Commerce’s methodology, the record does not contain the data for each CONNUM that resulted from Commerce’s calculations of a “quarterly indexed” “adjusted” weighted average cost of production for the period of review. (See Proposed Cost Adjustments Mem., Att. 3 (containing “sample” results of various steps of the methodology, but not complete data for the five CONNUMS used in the calculation of normal value).) Without disclosing the totality of the evidence upon which Commerce relied, no adequate explanation is presented. The Court declines to read into the record a justification which Commerce itself did not provide. 2. Proper Explanation of the Cost Recovery Test It seems to the Court that Commerce, to adequately explain its cost recovery methodology, should provide clear descriptions and data that compare the results obtained using its standard cost-recovery test to the results obtained via the adjusted quarterly indexed methodology used in the Final Results. The Court’s need to compare the results of the two methods comes from the cost recovery statute. The statute limits the Department’s discretion in disregarding below-cost-of-production sales by mandating that, for sales made at a price “above the weighted average per unit cost of production for the period of ... review, such prices shall be considered to provide for recovery of costs within a reasonable period of time.” 19 U.S.C. § 1677b(b)(2)(D) (emphasis added). Such sales, consequently, cannot meet the statutory criteria for exclusion from NV calculations. 19 U.S.C. § 1677b(b)(l) (stating that Commerce may disregard sales made at less than cost of production if the sales “were not at prices which permit recovery of all costs within a reasonable period of time”). The disregarded sales statute, on the other hand, is worded permissively: it states that, for sales meeting its criteria, “such sales may be disregarded in the determination of normal value.” Id. (emphasis added). This is consistent with the cost recovery statute, which only limits Commerce’s discretion to disregard sales priced above the statutory price floor, but leaves Commerce with the discretion to determine whether a sale priced below the floor might provide for recovery of costs within a reasonable time. § 1677b(b)(2)(D). The cost recovery statute, then, creates a price floor, above which home market sales “shall be” recovered and considered in establishing normal value. § 1677b(b)(2)(D). Because § 1677b(b)(2)(D) is non-discretionary, that price floor must be calculated in a manner that is consistent with the statutory language. The specified manner is by calculating the “weighted average per unit cost of production for the period of ... review!)]” Id. The Court must invalidate as contrary to law any cost recovery test that excludes home market sales prices that are above that price floor. Id. In essence, Plaintiff argues that the Department’s quarterly indexing methodology has erected an artificial floor—a raised stage—above the statutory price floor, and excluded sales that are above the floor but underneath the stage established by the methodology. The argument is compelling, because it is difficult to see how Commerce’s quarterly indexing would be useful except as a means to exclude certain sales despite the fact that they are priced above the statutory floor of the ordinary weighted average per unit cost of production. Indeed, Defendant comes close to saying as much, stating that “comparing prices that failed the below-cost test on a quarterly basis, with an unadjusted weighted average per-unit cost for the period of review, might result in below-cost prices erroneously considered to have recovered costs based simply on the timing of the sale. This is because comparing costs on a quarterly basis, and then comparing them a second time on an annual basis would produce anomalous results.” (Def.’s Brief at 33 (internal quotes and citations omitted).) The cost recovery statute explicitly constrains where the price floor for cost recovery may be set. Commerce apparently faced the dilemma of reconciling what the cost recovery statute actually requires with what Commerce wishes the statute required. To resolve this dilemma, it appears that Commerce’s quarterly indexing methodology built a stage above the statutory cost recovery price floor. Commerce then apparently excluded any home market sales that fell below the artificial floor set by Commerce’s methodology even if those sales were priced above the statutory price floor (and thus recoverable pursuant to the statute). The Court cannot, however, conduct its review solely upon appearances. In order to evaluate whether Commerce’s quarterly indexing methodology was in accordance with law, the Court requires Commerce to identify all of those sales that would be recoverable using the ordinary weighted average per unit cost of production for the period of review, but were excluded under the quarterly indexed version of the cost recovery test. Only then can the Court determine whether Commerce’s methodology is consistent with 19 U.S.C. § 1677b(b)(2)(D). 3. Remand Based on the concerns described above, the Court remands the cost-recovery component of the administrative review to Commerce for the following action. First, on remand Commerce shall calculate the normal value of Plaintiffs home market sales using both the quarterly-indexed cost recovery test employed in the Final Results and using the ordinary weighted av'erage per unit cost of production for the period of review. Second, Commerce shall include in the record the specific figures used in and resulting from these calculations. Third, in its remand redetermination, Commerce shall identify all those sales that are recoverable using the ordinary weighted average per unit cost of production for the period of review, but subject to exclusion under the quarterly indexed version of the cost recovery test. Fourth, Commerce shall explain which of the two methodologies it adopts to conduct the cost recovery test, stating in clear terms why the particular steps of that methodology are appropriate in the context of the requirements of 19 U.S.C. § 1677b (b)(2)(D). 3. Commerce’s Decision to Compare U.S. and Home Market Prices on a Quarterly Basis and Eliminate the “90/60” Day Window Period A. Statutory Scheme As discussed above, Commerce is obligated under the antidumping statute to determine the amount by which the normal value of the subject merchandise exceeds the export price or constructed export price. 19 U.S.C. § 1677(35). In most instances, the actual determination of an antidumping duty involves a comparison of the prices of the subject merchandise in the relevant home market of the foreign producer to those in the United States., While the basic price comparison may seem to be a simple mathematical exercise, the substantial body of law and practice which has developed in this regard demonstrates otherwise. In fact, the comparison of home market prices to U.S. prices involves a detailed and sometimes complex methodology designed to ensure that certain economic and business realities are considered. See Smith-Corona v. United States, 713 F.2d 1568, 1571 (Fed.Cir.1983) (“[Normal value] and United States price represent prices in different markets affected by a variety of differences in the chain of commerce”). To ensure that these economic realities are properly considered, Commerce has promulgated regulations to guide its analysis. Specifically, 19 C.F.R. § 351.414(c)(2) identifies the Department’s preference for use of the average-to-transaction method in making comparisons of export price with normal value in an administrative review. The application of this method is described under 19 C.F.R. § 351.414(e), and is more commonly referred to as the “90/60” day rule. Under this rule, Commerce will first attempt to compare U.S. sales of subject merchandise with weighted average home market sales within the same thirty day period. Where no sales of the like product are made in the exporting country in the month of the U.S. sale, Commerce will attempt to find a weighted average monthly price one month prior, then two months prior, and then three months prior to the month of the U.S. sale. (See Issues & Decision Memo at 16.) If unsuccessful, the Department looks one month after, and, finally, two months after the month of the U.S. sale. (See id.) At issue in the present case is the Department’s decision to depart from the contemporaneity guideline in its margin analysis, and instead to match sales only within the same quarterly cost averaging period used in the cost test. Under Commerce’s new approach, U.S. sales could only be compared to home market sales in the same quarter—as identified by Commerce—effectively limiting the potential matching period from six months to three. B. Parties’ Arguments Plaintiff challenges the Department’s elimination of the 90/60 day window period, asserting that Commerce’s alternate approach created clear distortions in respondent’s dumping margin. (Pl.’s Brief at 34-35.) As Plaintiff explains it, “[m]any U.S. sales were matched to less similar home market products—despite the fact that there were available above-cost sales of the identical or more similar products ... in the normal 90/60-day window period.” (Id. at 35.) SeAH attributes this to the fact that U.S. sales made in the first month of the four quarters defined by Commerce could only be compared to home market sales in the same month and the two following months even though the regulation assigns a higher preference to the two months prior to the month of the U.S. sale. (Id. at 34.) Thus, according to Plaintiff, this new methodology artificially inflated the company’s dumping margin “for reasons having nothing to do with a more accurate matching of costs and prices.” (Id. at 35 (internal quotation omitted).) SeAH further alleges that Commerce’s use of a quarterly time frame for the price-to-price comparisons is an unlawful deviation from the “contemporaneous month” requirement of 19 C.F.R. § 351.414(e)(2). (Id. at 38-39.) Recognizing that the regulation is qualified by the word “normally,” Plaintiff argues that any notional authority this language may confer upon Commerce to depart from the definition of contemporaneous month is rendered nugatory by the methodology actually employed. (Id. at 40.) SeAH maintains that the record fails to support Commerce’s decision to depart from the 90/60 day rule, specifically the agency’s determination that a significant increase in costs made use of the 90/60 day matching period unsustainable. Plaintiff challenges the evidence on which Commerce relied in making this inference. (Id. at 41-42.) At the core of Plaintiffs argument is the criteria Commerce used in its determination that costs had increased considerably. Because a cost increase of 25% between any two quarters is the threshold required for a departure from the use of annual averages in cost averaging, SeAH argues that price increases in the range reflected by the record “are per se insignificant and thus fail to support a departure from the 90/60-day rule.” (Id. at 42.) SeAH avers that the Department failed to adequately explain, or support with record evidence, what numerical threshold it used in reaching the conclusion that costs had increased significantly. (Id.) Without such a baseline, Commerce’s decision to deviate from the 90/60 day contemporaneity period is unsupported by substantial evidence. In addition, Plaintiff attacks the cost data on which Commerce based its analysis. (Id. at 43.) Because SeAH was never asked to report its costs for the pre-period of review quarter, Commerce relied on surrogate production costs taken from an affiliate of SeAH. However, other evidence on the record, asserts Plaintiff, indicated that SeAH’s cost of manufacturing was not increasing. For instance, the surrogate’s data included not just production costs but the prices at which the surrogate sold hot coil steel to SeAH. Inasmuch as this data demonstrates that the prices SeAH paid to its affiliate for hot coil steel remained the same from the pre-period of review quarter to the first quarter period of review, Plaintiff claims there was no need to consider the surrogate’s costs because the surrogate’s “selling prices to SeAH are SeAH’s costs.” (Id. at 44 (emphasis in original).) Thus, Commerce was in error to conclude that Plaintiffs costs were increasing at the start of the period of review. As a result, the premise on which Commerce based its decision to depart from the 90/60 day period was flawed. (Id. at 45.) Finally, Plaintiff challenges the Department’s product matching methodology itself, averring that the process created massive distortions in the dumping margin calculation. (See id.) SeAH points to the fact that Commerce’s use of quarterly costs resulted in lower costs in the first quarter of the period of review as compared to the annual average costs calculated by Commerce in the Preliminary Results. (See Case Brief at 26.) Although the lower costs resulted in a higher percentage of home market sales passing the cost test, SeAH’s dumping margin increased by almost 5 percent. (Id.) This, says Plaintiff, is a consequence of the Department’s revised product matching methodology, which prevents sales made in the first quarter from being matched to home market sales in the pre-period of review window period. (Pl.’s Brief at 46.) SeAH claims that this forced “certain high-volume U.S. products to be matched to less similar home market products,” even though more similar sales matches existed within the normal 90/60 day window period. (Id. at 6, 46.) Defendant’s counter-argument is, for the most part, based upon its interpretation of 19 C.F.R. § 351.414. Specifically, Commerce points to the regulation’s inclusion of the terms “in general” and “normally.” (Def.’s Brief at 21-22.) According to Commerce, the regulation only provides what the agency must do under “normal” circumstances, yet affords Commerce the discretion to decide “when the ‘normal’ situation ‘is inapplicable.’ ” (Id. at 24 (quoting KYD, Inc. v. United States, 33 CIT-, 613 F.Supp.2d 1371, 1382 (2009)).) In other words, 19 C.F.R. § 351.414(e) qualifies its instructions for application of the average-to-transaction methodology. Therefore, Commerce argues, the Department is free to depart from the 90/60 day contemporaneity guideline in certain anomalous situations. (Id. at 23-24.) Recently, Commerce has eliminated the 90/60 day window period “where costs and prices [have] changed significantly due to high inflation.” (Id. at 22-23 (citing Notice of Final Results of Antidumping Duty Administrative Review: Certain Welded Carbon Steel Pipe and Tube From Turkey, 61 Fed.Reg. 69,067 (Dec. 31, 1996) and Certain Porcelain-onr-Steel Cookware From Mexico: Final Results of Antidumping Duty Administrative Review, 62 Fed.Reg. 42,946 (Aug. 7, 1997)).) Here, Commerce concluded that use of the six month contemporaneity window would render distorted margin calculations because of the significant change in SeAH’s costs and home market prices over the period of review. (Id. at 23.) The Department reasoned that comparing lower priced home market sales from the pre-period of review window with first quarter U.S. sales, when the unadjusted home market price does not reflect the contemporaneous price increases that have occurred through the date of the U.S. sale, would result in a distorted analysis. (Issues & Decision Memo at 16.) In this way, Commerce analogized its prior rationale for departing from the normal 90/60 day window period (i.e., cases involving hyper-inflationary economies) with those conditions present here. Thus, Commerce found that “prieeto-price comparisons should be made over a shorter period of time to lessen the distortive effects of changes in sales price which result from significantly increasing costs.” (Id.) Consistent with its shortened cost averaging period, Commerce limited the extension of price comparisons to only the three months of a given quarter, effectively reducing the normal contemporaneity window by half. (Def.’s Brief at 24.) The Department maintains that “if six months is considered contem