AL Tech Specialty Steel Corp. v. United States

Opinion Tsoucalas, Senior Judge: Plaintiffs, AL Tech Specialty Steel Corporation, Carpenter Technology Corporation, Crucible Specialty Metals Division, Crucible Materials Corporation, Electralloy, Division of G.O. Carlson, Inc., Republic Engineered Steels, Slater Steels Corporation, Talley Metals Technology, Inc. and The United Steelworkers of America, AFL-CIO/CLC (collectively “AL Tech”), move this Court for judgment upon the agency record pursuant to Rule 56.2 of the Rules of this Court challenging certain aspects of the International Trade Administration, Department of Commerce’s (“Commerce” or “ITA”) final determination, entitled Notice of Final Determination of Sales at Not Less — 1345 — Than Fair Value: Stainless Steel Bar from Italy (“Final Results”), 59 Fed. Reg. 66,921 (1994). Background On January 27,1994, Commerce published a notice of its initiation of an investigation concerning stainless steel bar imported from Italy during the period of July 1,1993 through December 31,1993. See Initiation of Antidumping Duty Investigations: Stainless Steel Bar From Brazil, India, Italy, Japan and Spain, 59 Fed. Reg. 3844 (1994). Commerce published the preliminary results of its determination of the less than fair value investigation (“LTFV”) on August 4,1994. See Preliminary Determination of Sales at Less Than Fair Value and Postponement of Final Determination: Stainless Steel Bar From Italy, 59 Fed. Reg. 39,736 (1994). On December 28,1994, Commerce issued its final determination finding de minimus margins for both Valbruna and Foroni. See Final Results, 59 Fed. Reg. at 66,921. AL Tech challenges the following actions by Commerce as being inconsistent with law and unsupported by substantial evidence on the agency record: (1) applying a tax rate to U.S. price in excess of the average tax rate assessed on home market sales; (2) excluding B-2 and B-3 home market sales of Acciaierie Valbruna S.r.L. (“Valbruna”) from its model match calculation after permitting only limited reporting of such sales; (3) using Valbruna’s cost of production (“COP”) data from the period of investigation to calculate Valbru-na’s dumping margin; (4) treating Valbruna’s home market pre-sale costs associated with maintaining inventory as direct selling expenses; (5) applying weighted-average calculated margin of Foroni, S.p.A. and Foroni Metals of Texas, Inc. (collectively “Foroni”) to comparisons involving Foroni’s unreported U.S. sales; and (6) permitting Foroni to assign unique grade codes to non-specified grades. Discussion The Court’s jurisdiction in this action is derived from 19 U.S.C. § 1516a(a)(2) (1994) and 28 U.S.C. § 1581(c) (1994). The Court must uphold Commerce’s final determination unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B) (1994). Substantial evidence is “more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Universal Camera Corp. v. NLRB, 340 U.S. 474, 477 (1951) (quoting Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229 (1938)). “It is not within the Court’s domain either to weigh the adequate quality or quantity of the evidence for sufficiency or to reject a finding on grounds of a differing interpretation of the record.” Timken Co. v. United States, 12 CIT 955, 962, 699 F. Supp. 300, 306 (1988), aff'd, 894 F.2d 385 (Fed. Cir. 1990). — 1346 — 1. Tax Rate: In the Final Results, Commerce compared a weighted-average price of a group of home market sales that incurred a value-added tax (“VAT”) of less than 19 percent to U.S. prices that had been adjusted by a VAT rate of 19 percent. See Computer Program for Final Determination, C.R. Doc. No. 80, Pls.’ App., Ex. 11. AL Tech argues that pursuant to 19 U.S.C. § 1677a(d)(1)(C) (1988), an adjustment for tax forgiven on export sales is appropriate only where the comparison home market sales have actually been taxed. Accordingly, AL Tech requests a remand with instructions to Commerce to recalculate the adjustment to U.S. price for those sales compared to a foreign market value (“FMV”) that is based on a mixture of sales subject to the VAT and sales not subject to the VAT. On remand, AL Tech states that Commerce should compute a weighted average tax rate for the home market groups and apply the determined rate to the U.S. comparison prices. Pis.’ Mem. Supp. Mot J. Agency R. at 7-13. Commerce concedes that it erred by improperly comparing home market sales that incurred a VAT of less than 19 percent to U.S. prices that had been adjusted by a VAT rate of 19 percent. Commerce states that to comply with 19 U.S.C. § 1677a(d)(1)(C), it should have applied a weighted-average tax rate to both the home market and U.S. sales. However, Commerce insists that even if it had applied the appropriate tax rate, the margins for both Valbruna and Foroni would have been de minimus (0.17 percent and 0.05 percent, respectively). Therefore, Commerce argues that the error was harmless and does not warrant a remand. Def.’s Opp’n to Mot. J. Agency R. at 59-60. Defendant-intervenors Foroni and Valbruna both agree that if Commerce improperly applied the VAT, the error was harmless. Foroni’s Opp’n to Mot. J. Agency R. at 10-12; Valbruna’s Opp’n to Mot. J. Agency R. at 41-42. In rebuttal, AL Tech argues that the Court does not have the authority to accept Commerce’s new calculations without providing AL Tech with an opportunity to comment. According to AL Tech, the proper remedy for the error is to remand the issue so that Commerce may perform the calculations to determine whether the margin rises above de mini-mus. Pis.’ Reply to Opp’n to Mot. J. Agency R. at 2-4. Section 1677a(d)(1)(C), Title 19, United States Code, states that United States price shall be adjusted by the amount of any taxes imposed in the country of exportation directly upon the exported merchandise or components thereof, which have been rebated, or which have not been collected, by reason of the exportation of the merchandise to the United States, but only to the extent that such taxes are added to or included in the price of such or similar merchandise when sold in the country of exportation. (Emphasis added.) As Commerce concedes, by comparing a weighted average price of a group of home market sales that incurred a VAT of less — 1347 — than 19 percent to U.S. prices that had been adjusted by 19 percent, Commerce failed to comply with the statute. In reviewing Commerce’s actions, the Court is restricted to the administrative record and the evidence contained therein. Atcor, Inc. v. United States, 11 CIT 148, 154, 658 F. Supp. 295, 300 (1987). The Court will not accept post hoc rationalizations in place of evidence on the agency record. Id.; see also Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168-69 (1962). In the present case, Commerce admits that its calculations were flawed hut insists the error was harmless. Commerce has not attempted to provide the Court with a reason for its methodology but, rather, has offered the Court new calculations. The evidence presented by Commerce amounts to post hoc rationalization which cannot support a finding for Commerce. The Court cannot simply accept Commerce’s characterization of the error as harmless. It is for the Court, not Commerce, to determine the prejudicial effect of the error, and for the Court to do so, it must examine record evidence. The new calculations presented by Commerce cannot be considered record evidence. The Court also notes that Commerce does not include in its brief before the Court the actual calculations relied upon to support its position. Instead, Commerce merely states a conclusion without providing the Court with a detailed description of its calculations. As such, Commerce does not provide the Court with sufficient information concerning its recalculations. Finally, AL Tech should be provided with an opportunity to respond to the new calculations. Accordingly, the Court finds that a remand is appropriate for Commerce to recalculate the VAT adjustment by calculating a weighted-average tax rate for each of the subject home market comparison groups. 2. Exclusion of Certain Home Market Sales: On April 1, 1994, Valbruna wrote a letter to Commerce requesting permission to report limited data concerning certain home market sales because the difference in merchandise (“difmer”) adjustment would be greater than 20 percent for a majority of the comparisons. See Letter to Sec’y of Commerce from William Silverman, Rogers and Wells (Apr. 1, 1994), ER. Doc. No. 91, Pis.’ App., Ex. 3. Commerce granted Valbruna’s request to exclude certain sales from its home market sales listing, adding that if Commerce determined after verification that Valbruna should have reported such data, Commerce would resort to use of best information available (“BIA”). Letter to William Silverman from David L. Binder, Director, Antidumping Division II (Apr. 1,1994), ER. Doc. No. 89, Pis.’ App., Ex. 3. Commerce also required Valbruna to provide worksheets demonstrating the size of the difmer for any sales not reported solely on that basis. See id. Finally, Commerce granted Valbruna’s request to report only the first four criteria out of the six model matching methodology criteria for those home market products which are not comparable to U.S. products. See Mem. of V Irene Darzenta, Import — 1348 — Compliance Specialist, Antidumping Division II (April 6, 1994), ER. Doc. No. 93, Pis.’ App., Ex. 3. In the Final Results, Commerce decided to exclude B-2 and B-3 home market sales from the final margin calculations stating the following: We verified the fact that these sales would not be used for matching purposes. Therefore, consistent with our preliminary determination, we have continued to disregard sales in File 2 for purposes of our margin calculations. 59 Fed. Reg. at 66,925. AL Tech argues that the exclusion of these sales was improper because Commerce failed to verify that the sales would not be used for comparison purposes. AL Tech maintains that Commerce violated its statutory duty to verify the accuracy of data submitted by respondents. Pis.’ Mem. Supp. Mot. J. Agency R. at 14 (relying on 19 U.S.C. § 1677e(b) (1988)). In an effort to demonstrate the unreliability of Valbruna’s arguments concerning these sales, AL Tech emphasizes inconsistencies in Valbruna’s arguments to Commerce concerning the exclusion of the sales. AL Tech points out that first Valbruna asserted that the Files B-2 and B-3 merchandise would not be used because the difmer adjustment would be greater than 20 percent, and then Valbruna claimed that the merchandise would not be used for comparison purposes because none of those sales were within the first five identical or most similar product matches. Pis.’ Mem. Supp. Mot. J. Agency R. at 14. Commerce defends its decision to permit limited reporting of B-2 sales by Valbruna as consistent with its six criteria model match methodology. Commerce explains that it verified Valbruna’s computer reporting codes within each of the six criteria demonstrating that the B-2 sales were neither identical nor the most similar matches to U.S. sales. Commerce contends that AL Tech incorrectly relied on only four of the six criteria in alleging that Valbruna’s B-2 home market sales were similar to its U.S. sales. Def.’s Opp’n to Mot. J. Agency R. at 20-24. Futher, once it determined that the sales at issue were not similar or identical to Valbruna’s U.S. sales, Commerce asserts that it appropriately permitted Valbruna to report limited information about the B-2 sales. Commerce emphasizes that it warned Valbruna of Commerce’s power to resort to BIA if it later determined that B-2 sales were indeed identical or similar to U.S. sales, and insists that it properly verified the information Valbruna reported concerning its B-2 sales. Id. at 24-29. Regarding the B-3 sales, Commerce maintains that AL Tech failed to exhaust its administrative remedies by not raising the issue at the administrative level. However, if the Court concludes that AL Tech did not need to exhaust administrative remedies, Commerce advances the same arguments in support of its treatment of the B-3 sales as those advanced for the B-2 sales. Id. at 29-33. Valbruna, defendant-intervenor, supports Commerce’s actions by emphasizing the broad discretion the statute grants to Commerce in determining what constitutes similar merchandise. Valbruna states — 1349 — that it applied Commerce’s product comparison criteria before requesting the exclusion of certain home market sales from the reporting requirements. According to Valbruna, Commerce, not Valbruna, concluded that the B-2 and B-3 sales would not be used as comparison models based on Commerce’s own model matching methodology. Valbruna’s Opp’n to Mot. J. Agency R. at 16-23. In the Final Results, Commerce explained its decision to exclude the B-2 sales as follows: With regard to petitioners’ argument that Valbruna failed to provide worksheets showing difmers in excess of 20 percent for sales in File 2, our letter of April 1,1994, to Valbruna stated that we would require worksheets for any sales not reported solely because of the size of the difmer (as opposed to those that did not match to a U.S. sale based on product characteristics). As respondent states, and as we verified, because the sales in File 2 would never match to U.S. sales based on the six product characteristics specified in Appendix V of the questionnaire issued in this case, there was no need for respondent to provide worksheets. Finally, concerning petitioners’ argument that a comparison of File 2 sales to U.S. sales shows several products with identical matches, we agree with respondent that this argument is incorrect because petitioners based their analysis on only the first four product characteristics as opposed to the six point characteristics that the Department required for matching purposes in Appendix V of the questionnaire. As explained above, when all of the matching characteristics are considered, the sales in question would not be used for matching purposes. 59 Fed. Reg. 66,925-26. This Court has frequently acknowledged Commerce’s broad discretion in devising a methodology for determining what constitutes similar merchandise pursuant to 19 U.S.C. § 1677 (16) (1988). See Koyo Seiko Co. v. United States, 19 CIT 1085, 1092, 898 F. Supp. 915, 921-22 (1995), rev’d on other grounds, 92 F.3d 1162 (Fed. Cir. 1996); Torrington Co. v. United States, 19 CIT 403, 412, 881 F. Supp. 622, 634 (1995). The issue in this case, however, is not the validity of Commerce’s methodology but, rather, whether Commerce actually applied its own methodology or relied completely on the information provided by Valbruna without adhering to proper verification procedures. Section 1677e(b), Title 19, United States Code, instructs Commerce to verify information in the following manner: The administering authority shall verify all information relied upon in making— (1) a final determination in an investigation, * * * In publishing [a final determination], the administering authority shall report the methods and procedures used to verify such information. If the administering authority is unable to verify the accuracy of the information submitted, it shall use the best — 1350 — information available to it as the basis for its action, which may include * * * the information submitted in support of the petition. While the statute clearly requires Commerce to verify information, it also affords Commerce latitude in implementing verification procedures. See American Alloys, Inc. v. United States, 30 F.3d 1469, 1475 (Fed. Cir. 1994); PPG Industries, Inc. v. United States, 15 CIT 615, 620, 781 F. Supp. 781, 787 (1991); Kerr-McGee Chem. Corp. v. United States, 14 CIT 344, 362, 739 F. Supp. 613, 628 (1990). In determining the propriety of Commerce’s verification procedures, the issue is whether there is substantial evidence on the agency record to support Commerce’s chosen methodology. Hercules, Inc. v. United States, 11 CIT 710, 726, 673 F. Supp. 454, 469 (1987). Commerce may not need further information if submissions by a respondent are complete and the explanations are sound. Kerr-McGee, 14 CIT at 344, 739 F. Supp. at 628. In fact, the Court has approved Commerce’s practice of accepting a proposed methodology from respondents as long as the methodology is reasonable and “based on information in the administrative record, is likely to be representative of the underlying information.” Torrington Co. v. United States, 17 CIT 967, 972, 832 F. Supp. 405, 410 (1993). Commerce provided the following description of its verification procedures for product comparisons: For verification, Valbruna officials provided a worksheet listing the various grades of merchandise that Valbruna produces and how the stainless steel grades fit into this listing * * *. Valbruna provided a “Type and Finishing Operation Conversion Table” which converts Valbruna’s grades and product descriptions to the Appendix V hierarchy. Valbruna officials explained the manner in which they compiled the three computer files used to report sales to the Department. File 3 contains grades other than 304, 304L, 316, and 316L (these 300 grades are all contained in Files 1 or 2). From their computer system they compiled a total of all stainless steel bar sold during the [period of investigation] (using gross unit price, quantity, date of sale, customer, and product description). Then they used the conversion sheets (Valbruna product codes cross-referenced to DOC Appendix V codes) to sort the data by the information requested by the Department. Officials explained that certain grades were sold in both the U.S. and home markets but had different finishes, i.e., centerless grinding and hot finished peeled are sold in the home market but not in the United States. Since these sales would not be used for comparison purposes, the Department allowed limited reportingrequirements for these sales. Valbruna reported the product code and the quantity sold in the home market for these sales in File 2. Verification of the Responses of Acciaierie Valbruna, S.r.L. in the Antidumping Duty Investigation of Stainless Steel Bar from Italy (“Verification of Valbruna”), C.R. Doc. No. 65, Def.’s App., Ex. 11, at 6. — 1351 — The Court finds that Commerce’s verification procedure was supported by substantial evidence on the agency record. In this case, Val-bruna relied on Commerce’s own methodology rather than suggesting a different method. AL Tech’s central concern that Commerce did not verify the information used in applying the model matching methodology is without merit. Commerce is not required to examine each piece of information reported by a respondent. See PMC Specialties Group, Inc. v. United States, 20 CIT 1130, 1133, Slip Op. 96-153, at 7 (Aug. 30, 1996); See also Monsanto Co. v. United States, 12 CIT 937, 947, 698 F. Supp. 275, 283-84 (1988). The verification report demonstrates that Commerce adequately verified Valbruna’s assertion that the sales at issue would not be used for comparison purposes. Valbruna provided Commerce with all information requested by Commerce and its explanation for limiting its reporting of B-2 and B-3 sales was sound. Commerce verified Valbruna’s grades and product descriptions accord-ingto Commerce’s own criteria set forth in Appendix V of the questionnaire.1 There was no reason for Commerce to doubt the accuracy of any of Valbruna’s reported information. Further, Commerce provided a reasonable explanation of its actions in the Final Results. As such, the Court finds Commerce’s exclusion for Valbruna’s B-2 and B-3 home market sales from its margin calculations to be supported by substantial evidence on the agency record.2 3. Valbruna’s Cost of Production Data: Pursuant to 19 U.S.C. § 1677b(b) (1988), whenever Commerce has “reasonable grounds to believe or suspect that sales in the home market * * * have been made at prices which represent less than the cost of producing the merchandise in question, it shall determine whether, in fact, such sales were made at less than [cost of production].” Commerce determined that a cost of production investigation was necessary with respect to Valbruna and, therefore, requested reporting of production costs and constructed value data. In accordance with Commerce’s request, Valbruna reported weighted-average production data based on costs incurred during the period of investigation (“POI”). See Final Results, 59 Fed. Reg. at 66,929. AL Tech objects to the reporting methodology permitted by Commerce, claiming that the costs reported were not representative of the costs actually incurred because the merchandise sold in the home market during the POI was produced prior to the POI and held in inventory until it was sold. AL Tech insists that Valbruna’s calculations resulted in an understatement of the actual costs incurred by Valbruna to manufac — 1352 — ture stainless steel bar sold in Italy during the POI. Pis.’ Mem. Supp. Mot. J. Agency R. at 16-18. While AL Tech acknowledges Commerce’s presumption that merchandise sold during the POI was also produced during the POI, AL Tech argues that the presumption may be rebutted by evidence demonstrating that the merchandise sold during the POI was produced at a time when costs of production differed. AL Tech argues that Commerce’s practice has been to deviate from the presumption in order to use costs representative of costs actually incurred to produce the merchandise sold during the POI. Pls.’ Mem. Supp. Mot. J. Agency R. at 18-21 (citing Final Determination of Sales at Less Than Fair Value: Dynamic Random Access Memory Semiconductors of One Megabit and Above From the Republic of Korea (“DRAMS”), 58 Fed. Reg. 15,467, 15,473 (1993); Final Determination of Sales at Less Than Fair Value: Canned Pineapple Fruit From Thailand (“Canned Pineapple”), 60 Fed. Reg. 29,553, 29,564 (1995); Notice of Final Determination of Sales at Less Than Fair Value: Stainless Steel Bar From India (“Stainless Steel Bar From India”), 59 Fed. Reg. 66,915, 66,920 (1994)). AL Tech directs the Court’s attention to the affirmance of Commerce’s decision in DRAMS to use costs from a period commencing prior to the POI in its COP calculation. Pis.’ Mem. Supp. Mot. J. Agency R. at 19-20 (citing Micron Technology, Inc. v. United States, 19 CIT 829, 837-38, 893 F. Supp. 21, 31-32 (1995)). AL Tech insists that Commerce should apply a similar practice in this case, because the cost of stainless steel scrap (one of the main raw materials used to produce stainless steel bar) fluctuated such that it was higher during the period in which Valbruna actually produced the stainless steel bar sold in Italy during the POI, than it was during the POI. Pls.’ Mem. Supp. Mot. J. Agency R. at 17-19. In response, Commerce asserts that AL Tech raised the issue too late in the investigation for Commerce to thoroughly analyze the allegation. Commerce maintains that although AL Tech alerted Commerce to the issue seven days before the scheduled date of the commencement of verification in compliance with 19 C.F.R. § 353.31(a)(1)(i) (1994), there was insufficient time for Commerce to determine whether a change in the cost data reporting period was warranted. Def.’s Opp’n to Mot. J. Agency R. at 35-37. In addition, Commerce claims that AL Tech failed to present any evidence to justify a departure from Commerce’s general practice of relying on weighted-average cost of production data incurred during the POI. Commerce insists that AL Tech did not demonstrate that unique circumstances existed to support a change in methodology. Commerce further disputes the assertion that Valbruna’s use of the POI period to calculate its cost of production resulted in a substantial understatement of actual costs. Commerce points out that AL Tech relies on average international prices and dollar denominated costs, rather than Valbru-na’s actual costs and experience, to support its contention that raw materials were more expensive for Valbruna during the time period — 1353 — before the POI (the time during which Valbruna produced the merchandise sold during the POI) than during the POI. Id. at 37-40. Valbruna supports Commerce’s position, arguing that this case is distinguishable from past administrative reviews in which Commerce found unique circumstances to require reliance on periods other than the POI to compute cost of production data. Valbruna distinguishes the present review from Canned Pineapple, arguing that in that case, the seasonality of the product required the use of a different time period. Valbruna also highlights differences between this case and DRAMS by arguing that in DRAMS, Commerce lagged the reportingperiod because of rapidly changing costs of semiconductors. In contrast, Valbruna asserts that the cost of manufacturing stainless steel bar was stable. Val-bruna negates AL Tech’s claim that dollar-denominated scrap prices during the POI were lower than prices before the POI. Valbruna explains that Valbruna purchased stainless steel scrap in Italian lire and that the lire-denominated costs did not change in the same manner as the dollar-denominated costs because the lire weakened against the dollar during the period at issue. Valbruna’s Opp’n to Mot. J. Agency R. at 24-29. In the Final Results, Commerce addressed this issue as follows: Section D of the questionnaire clearly requests weighted-average production data based on costs incurred during the POI. The Department has departed from this general policy only when unique circumstances arise, such as when there was no production during the POI. Furthermore, companies, frequently hold inventory for a period of time between production and shipment and raw materials are held for a period of time between purchase and production. An average inventory holding period or length of time between order and production are only estimates. Sales are sometimes made from existing stock or may be produced to order, or even a combination of both. Petitioners raised the issue for the first time in the pre-verification comments — too late in the investigation for the Department to perform the appropriate analysis to determine whether a change in the cost data reporting period is warranted. Furthermore, if the Department was to accept petitioners’ argument, the CV data would be based on a different accounting period than the COP data, effectively doubling the burden on all parties. Accordingly, absent strong evidence to the contrary, the Department assumes that the cost structure prevailing during the POI is representative and can be used to calculate COE 59 Fed. Reg. at 66,929. Commerce’s regulations require parties to submit factual information to be considered in the final determination no later than “seven days before the scheduled date on which the verification is to commence.” 19 C.F.R. § 353.31(a)(1)(i). Commerce does not dispute the fact that AL Tech complied with the deadline set forth in 19 C.F.R. § 353.31(a)(1)(i) by submitting its comments regarding Valbruna’s — 1354 — reporting of cost of production seven days before the verification. However, Commerce maintains that it did not have sufficient time to review AL Tech’s allegations. The Court finds this argument to be without merit. While AL Tech may not have submitted the information at the earliest possible moment, AL Tech complied with the regulatory time limits. Commerce cannot apply these time limits arbitrarily or capriciously by refusing to accept information submitted before the applicable deadline. See Böwe-Passat v. United States, 17 CIT 335, 338 (1993) (holding that Commerce acted arbitrarily and capriciously by refusing to accept supplemental explanatory information submitted in an untimely manner when Commerce failed to indicate deficiencies in respondent’s prior responses to requests for information). Having decided that AL Tech’s submission was timely, the Court now turns to the merits of AL Tech’s arguments. One of the cases AL Tech cites in support of its position is Micron. While it is true that in Micron, the court upheld Commerce’s use of a time lag to compare constructed value and COP to sales, AL Tech fails to recognize the distinctions between that case and the one presently before the Court. Micron involved the production and sales of dynamic random access memory semiconductors (“DRAMS”). The court stated that “the semiconductor industry is characterized by relatively rapidly declining per-unit costs due to increased in yield, design changes, and other efficiency improvements over time.” Micron, 19 CIT at 838, 893 F. Supp. at 32. Thus, the court concluded that in order to properly match cost of production to sales over a given period of time, a lag was necessary to account for the fact that production costs for DRAMS in any given month were not the same as the cost of the individual DRAMS sold at the end of the month. Id. The facts of Micron are not analogous to those of the case at bar. The steel industry is not subject to rapid changes due to sudden advances in technology. While normal market fluctuations may affect the price of inputs, such as scrap metal, the changes are not nearly as dramatic as AL Tech would have the Court believe. AL Tech also relies on two other administrative reviews to support its position—Canned Pineapple and Stainless Steel Bar From India. Neither of these prior reviews require the result AL Tech seeks in this case. Canned Pineapple involved unique circumstances due to the growth period of respondent’s self-grown pineapples. The seasonality of the pineapples required Commerce to adopt a different approach in determining cost of production. 60 Fed. Reg. at 29,564. In Stainless Steel Bar From India, 59 Fed. Reg. at 66,920, Commerce determined that respondent’s weighted-average nickel cost was based on quantities in excess of quantities used. In that review, during the last month of the POI, the respondent purchased nickel at a price significantly less than the nickel price it had paid during the other months of the POI. Id. Commerce concluded that the price of nickel during the last month of the POI could not represent the cost of production during the POI since the nickel purchased at that time would not have been converted to stainless steel bar — 1355 — before the end of the POI. Id. In contrast, there is no evidence in this case that the weighted-average cost was based on quantities in excess of quantity used. Further, the prices during the POI did not significantly decrease during the last month. In fact, the prices of both scrap metal and nickel increased towards the end of the POI. See Petitioners’ Case Brief Regarding Valbruna, S.p.A. and Avesta Sheffield, Inc., C.R. Doc. No. 71, at Ex. 5, Pls.’ App., Ex. 12. After examining the record, the Court finds that although prices of scrap metal (according to the London Metal Exchange) were lower during the POI than during the period before the POI, there were no erratic fluctuations to indicate an unstable market. On the contrary, the decrease in price was gradual and in relatively small increments. See id. As Valbruna correctly notes, some of the apparent price changes were actually due to the weakening value of the Italian lire. See Exhibits for Valbruna’s Cost Verification Report, C.R. Doc. No. 68, Valbruna’s App., Ex. 14. While the currency conversion rate does not account for the full amount of the decline in price, it provides further support for Commerce’s decision to adhere to-its general practice of comparing sales made during the POI to manufacturing costs from the POI.3 Based on the evidence contained in the record, the Court concludes that this was not a unique situation warranting a departure from Commerce’s usual methodology. Holding merchandise in inventory for a period of time is a frequent practice. The decline of prices of scrap metal and nickel was not significant enough to render the data supplied by Val-bruna unrepresentative of the actual costs incurred during the POI. Thus, the Court finds Commerce’s actions to be supported by substantial evidence on the agency record and in accordance with law. 4. Home Market Pre-Sale Inventory Costs: Valbruna maintained five regional warehouses in Italy to service customers located in those regions. In the contested review, Commerce treated Valbruna’s cost of maintaining inventory at those warehouses (pre-sale movement expenses, pre-sale warehousing expenses and pre-sale inventory carrying expenses) as direct selling expenses. See Final Results, 59 Fed. Reg. at 66,928. AL Tech argues that Commerce erred by treating these expenses as direct expenses because Valbruna failed to tie the pre-sale expenses to particular sales. AL Tech alleges that Valbruna used an overall allocation methodology to calculate the average pre-sale warehousing and pre-sale inventory-carrying expenses for each of its regional service centers. As such, AL Tech insists that Commerce’s actions are inconsistent with the review, entitled Final Determination of Sales at Less Than Fair Value: Polyethylene Terephthalate Film, Sheet, and Strip From Japan (“PET Film”), 56 Fed. Reg. 16,300, 16,303 (1991), in which Commerce — 1356 — treated the pre-sale expenses of respondent Toray Industries, Inc. (“Toray”) as indirect because Toray failed to relate the expenses to specific sales. According to AL Tech, Valbruna does not dedicate certain merchandise in inventory to particular customer accounts. Due to the fungible nature of stainless steel bar, AL Tech insists that it would not make economic sense for Valbruna to set aside particular merchandise for customers. Pis.’ Mem. Supp. Mot. J. Agency R. at 21-25. Commerce counters that it treated Valbruna’s pre-sale expenses associated with maintaining inventory as direct selling expenses because Valbruna provided sufficient evidence to demonstrate that the expenses were indeed related to particular sales. Commerce cites the following four facts observed at verification that resulted in Commerce’s determination that the pre-sale expenses were directly related to particular sales: 1) “open order” tags reflecting specific inventory levels at its service centers against a customer’s supply forecast; 2) Valbruna’s accounting system which tracks additional stock by quantity going to a warehouse and states that the merchandise is destined for a specific customer; 3) Valbruna’s “just-in-time” commercial strategy which keeps stock in inventory at all times to meet a customer’s needs at the customer’s request; 4) home market purchase orders from resellers and end users stipulating that Valbruna keep “minimum buffer stock” or “minimum inventory levels” at particular service centers depending on the customer’s location. Def.’s Opp’n to Mot. J. Agency R. at 44. Commerce further claims that the facts in this case resemble those of a prior administrative review, entitled Final Determination of Sales at Less Than Fair Value: Certain Hot Rolled Lead and Bismuth Carbon Steel Products From the United Kingdom ("Lead and Bismuth”), 58 Fed. Reg. 6207, 6210 (1993), in which Commerce treated pre-sale expenses as direct expenses after the respondent demonstrated that it maintained customer-specific inventory pursuant to requests by certain customers. Commerce also refers to its treatment of respondent Teijin, Ltd. (“Teijin”) in PET Film, 56 Fed. Reg. at 16,303, claiming that it accepted Teijin’s contention that pre-sale warehousing expenses were directly related to home market sales based on evidence that the stock in question was only available for sale to specific customers. Def.’s Opp’n to Mot. J. Agency R. at 44-47. Valbruna supports Commerce’s treatment of its pre-sale inventory costs arguing that there was adequate record evidence demonstrating that the subject expenses were tied to specific sales. Valbruna explains that pursuant to its “just-in-time” supply service, Valbruna maintains stock in its regional warehouses to meet each customer’s specific needs. Valbruna points out that Commerce verified Valbruna’s practice of keeping a minimum level of inventory of specifically identified products for particular customers. Valbruna’s Opp’n to Mot. J. Agency R. at 33-37. — 1357 — Valbruna further argues that Commerce was not required to examine each sale made from the service centers to determine whether such sale was dedicated to a specific customer. Valbruna emphasizes that Commerce has broad discretion in determining how to conduct a verification proceeding. Id. at 37-38. In the Final Results, Commerce supported its decision by describing the following verification procedures: At verification we reviewed customer purchase orders and Val-bruna order confirmations which stipulated that Valbruna was required to keep on hand a specific amount of subject merchandise with certain specifications for particular customers at particular service centers. The record contains no indication that Valbruna sold this merchandise to customers other than the ones for which the particular merchandise was held in inventory. In fact, company officials stated that the merchandise is usually so specialized that Valbruna would be unable to sell it to other customers. We also observed duringthe plant tour merchandise with “open order” tags reflecting open orders against a customer’s supply forecast for which Valbruna was required to maintain specific inventory levels at its service centers. Furthermore, we observed that Valbruna’s accounting system tracks additional stock going to a warehouse; it lists the quantity but not the price, and states the merchandise is destined for a specific customer. 59 Fed. Reg. at 66,928. In Torrington Co. v. United States, 82 F.3d 1039, 1050-51 (Fed. Cir. 1996), the CAFC shifted the analysis of direct versus indirect expenses from the recordation and allocation to the calculation of such expenses. Specifically, the court clarified that the “allocation of expenses in a manner different from the calculation of the expenses * * * does not alter the relationship between the expenses and the sales under consideration.” Id. at 1051. While the decision of the CAFC addressed post-sale price adjustments, the analysis concerning whether of not an expense is direct remains the same for pre-sale expenses. Thus, AL Tech incorrectly focuses on Valbruna’s allocation methodology to support its argument. The appropriate determination for the Court is whether Valbruna’s pre-sale movement, warehousing and inventory carrying expenses were indeed related to particular sales. However, because pre-sale expenses, by their very nature, are incurred before any actual sale has occurred, it is more difficult to demonstrate a direct relationship between the expense and a particular sale. Torrington Co. v. United States, 20 CIT 632, 635, 926 F. Supp. 1151, 1156 (1996); Ad Hoc Comm. AZ-NM-TX-FL Producers of Gray Portland Cement v. United States, 19 CIT 692, 694, Slip Op. 95-91, at 5, 1995 WL 351035, *2 (May 15, 1995), aff'd 95 F.3d 1164 (Fed. Cir. 1996). The limited circumstances in which Commerce recognizes freight and warehouse expenses as direct are those involving “products channeled or customized for certain buyers.” Ad Hoc, 19 CIT at 694-95,, Slip Op. 95-91, at 5 (citing Final Results, 59 Fed. Reg. at 66,928; PET Film, 56 — 1358 — Fed. Reg. at 16,303). In Torrington, 20 CIT at 637, 926 F. Supp. at 1156, the Court upheld Commerce’s treatment of pre-sale expenses as indirect expenses where the respondent failed to present any evidence supporting a finding that expenses were incurred “in the course of ‘channeling’ or ‘positioning’ sales for particular customers.” (Quoting Ad Hoc, 19 CIT at 694-95, Slip Op. 95-91, at 5). In contrast, Valbruna presented, and Commerce verified, facts supporting its claim that the expenses at issue were incurred as a result of particular merchandise being set aside for certain customers. See Final Results, 59 Fed. Reg. at 66,928; see also Verification of Valbruna, C.R. Doc. No. 65, Def.’s App., Ex. 11, at 19-21. The inventory-carrying costs were incurred at specific service centers rather than at Valbruna’s headquarters where Valbruna maintains general inventory. Valbruna Sales Verification Report, C.R. Doc. No. 65, Valbruna’s App., Ex.8, at 19. The service centers exist to satisfy customers in specific geographic locations. Id. Valbruna’s explanations and Commerce’s observations of “just-in-time” inventory, “open order” tags and Valbruna’s accounting system provided adequate evidence of a direct link between the inventory and warehousing expenses incurred at these service centers and the merchandise held for specific customers.4 As such, the Court sustains Commerce on this issue. 5. Use ofForoni’s Weighted-Average Margin: In the review at issue, Foroni failed to report a portion of U.S. sales. Rather than applying adverse best information available (“BIA”), Commerce agreed to use Foroni’s overall weighted-average calculated margin to account for the unreported sales. Final Results, 59 Fed. Reg. at 66,924-25. AL Tech argues that Commerce’s failure to resort to adverse BIA was inconsistent with 19 U.S.C. § 1677e(c). According to AL Tech, Commerce’s reliance on the weighted-average margin as a neutral surrogate for a percentage of U.S. sales caused an artificial reduction ofFor-oni’s dumping margin. Pls.’ Mem. Supp. Mot. J. Agency R. at 25-26. Commerce responds that in light ofForoni’s overall cooperation with the investigation, the application of neutral BIA was appropriate. Commerce points to the unique circumstances of this case resulting in For-oni’s inadvertent reporting error. Commerce explains that to comply with Commerce’s requirement of reporting sales according to purchase order dates, Foroni manually searched its files for all invoices generated during and after the period of investigation pursuant to purchase orders issued within the period of investigation. At verification, Commerce also manually searched Foroni’s files and verified that a percentage ofFor-oni’s sales invoices that had been created after the period of investigation had been inadvertently overlooked. Commerce states that an — 1359 — examination of these sales established that the gross prices indicated on the invoices were comparable to those for sales of the same products that were reported and verified. Commerce emphasizes that aside from this error, which was not purposeful, Foroni had complied with Commerce’s requests for information. Finally, Commerce maintains that the omission did not significantly impede the investigation. Def.’s Opp’n to Mot. J. Agency R. at 48-53. In support of Commerce, defendant-intervenor Foroni adds that the margins were not underestimated because the allocation of all indirect expenses over only a portion of sales had the effect of overstating expenses on a per sale basis, thereby lowering the U.S. sales price and exaggerating any dumping margins. Foroni’s Opp’n to Mot. J. Agency R. at 13-14. Section 1677e(c) of Title 19, United States Code, states that Commerce “shall, whenever a party or any other person refuses or is unable to produce information requested in a timely manner and in the form required, or otherwise significantly impedes an investigation, use the best information otherwise available.” (Emphasis added). Regarding Commerce’s selection of BIA, in Allied-Signal Aerospace Co. v. United States, 996 F.2d 1185, 1191 (Fed. Cir. 1993), the CAFC noted that “because Congress has ‘explicitly left a gap for the agency to fill’ in determining what constitutes the best information available, the ITA’s construction of the statute must be accorded considerable deference” (quoting Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843-44 (1984)). Thus, the Court must grant Commerce considerable deference in its choice of BIA. The Court has upheld Commerce’s use of adverse BIA even where a respondent claimed inability to provide information requested. Koyo Seiko Co. v. United States, 20 CIT 920, 924-26, 936 F. Supp. 1040, 1043-45 (1996), appeal docketed, No. 97-1031 (Fed. Cir. Oct. 17, 1996) (upholding Commerce’s use of adverse BIA where respondent failed to report any home market sales because to do so would have required a manual review of all sales); Koyo Seiko Co. v. United States, 19 CIT 1272, 1278, 905 F. Supp. 1112, 1116-17 (1995) (holding that Commerce appropriately resorted to adverse BIA where respondent did not report cost of production data in possession of an uncooperative related company). On the other hand, the court has recognized that the application of neutral BIA is appropriate when a cooperative respondent substantially complies with information requests and where any deficiencies arise from factors beyond the respondent’s control. Unisor Sacilor v. United States, 19 CIT 1314, 1316, 907 F. Supp. 426, 429 (1995). Neutral BIA may also be applied where a minor or insignificant adjustment is involved. AD Hoc Comm. of AZ-NM-TX-FL Prod. v. United States, 18 CIT 906, 915, n.22, 865 F. Supp. 857, 865, n.22 (1994), aff'd 68 F.3d 487 (Fed. Cir. 1995). The facts of this case support the application of neutral BIA. In the cases cited above involving Commerce’s application of adverse BIA, the — 1360 — respondents failed to report a substantial amount of information. In the present case, the missing information constituted a small portion of U.S. sales reported. This is not a situation in which Foroni did not report any information concerning its U.S. sales. On the contrary, unlike the respondent in Koyo Seiko, 20 CIT at 924-26, 936 F. Supp. at 1043-45, Foroni manually reviewed its U.S. sales invoices in order to comply with Commerce’s information request. The omission of a small percentage of the U.S. sales was due to an inadvertent error rather than a refusal to report the information. See Verification of the Responses of Foroni Metals of Texas, Inc. in the Antidumping Duty Investigation of Stainless Steel Bar from Italy, C.R. Doc. No. 62, Def.’s App., Ex. 9, at 8-11. Thus, the size of the omission combined with Foroni’s substantial compliance justified Commerce’s application of a neutral BIA. The Court concludes that Commerce acted within its discretion and in accordance with law. 6. Foroni’s Unique Grade Codes: In the Final Results, Commerce accepted Foroni’s assignment of unique grade codes to sales of 316LUG and 316LN merchandise. 59 Fed. Reg. at 66,925. By doing so, AL Tech claims that Commerce permitted Foroni to contravene Commerce’s instructions in the original questionnaire. AL Tech argues that Commerce’s own inspection of Foroni’s books contradicted Foroni’s assertion that 316LN was so different from 316L as to warrant a separate grade code. According to AL Tech, the effect of assigning unique control numbers to 316LN and 316LUG was to decrease the dumping margin to a de minimus level. Pis. ’ Mem. Supp. Mot. J. Agency R. at 26-28. Commerce emphasizes that it possesses broad discretion to determine what constitutes similar merchandise under 19 U.S.C. § 1677(16). Commerce supports its actions by insisting that it verified that the products sold in the home market (316LUG and 316LN) have different chemical compositions than the 316L product sold in the United States. Based on this determination, Commerce maintains that it properly determined that the merchandise at issue were not the most similar to grade 316L merchandise sold in the United States. Def.’s Opp’n to Mot. J. Agency R. at 54-58. Foroni agrees with Commerce’s position and argues that its actions were in fact consistent with the questionnaire instruction to assign additional grade codes to each different grade of stainless steel bar sold during the period of investigation. Foroni’s Opp’n to Mot. J. Agency R. at 18-21. The Court has already stated that Commerce has broad discretion to determine what constitutes similar merchandise. See supra at 11. In devising the appropriate methodology, Commerce seeks accurate investigations based on comparisons that are as similar as possible. Torring1361 ton, 19 CIT at 414, 881 F. Supp. at 634. In granting Foroni’s request to assign unique grade codes, Commerce stated the following: We reviewed the information provided by Foroni regarding the different chemical compositions and material costs of each product prior to, as well as during, verification and determined that grades 316LUG and 316LN are in fact chemically different from grade 316L. Based on our review of the chemical compositions and material costs stated above, we determined that these products are not the most similar to grade 316L sold in the United States. Furthermore, we disagree with petitioners’ contention that For-oni is attempting to alter the matching hierarchy. Grade, which takes into account chemical composition, is in fact one of the matching criteria in Appendix V of the questionnaire. Final Results, 59 Fed. Reg. at 66,925. Commerce’s decision and explanation were reasonable. AL Tech correctly asserts that Commerce may not abdicate its responsibility to determine what constitutes similar merchandise to an interested party. See Timken Co. v. United States, 10 CIT 86, 98, 630 F. Supp. 1327, 1338 (1986). However, Commerce did not assign the task of defining similar merchandise to Foroni. Rather, the record indicates that Commerce verified Foroni’s assertion that the merchandise at issue did indeed deserve unique grade codes. AL Tech quotes only a portion of the verification report, leaving out the fact that Commerce verified that some differences did exist among the three grades of merchandise. Verification of the Responses of Foroni S.p.A. in the Antidumping Duty Investigation of Stainless Steel Bar from Italy, C.R. Doc. No. 63, Pls.’ App., Ex. 7, at 7-8. This is not a situation like Timken Co. where Commerce failed to collect any data on home market sales characterized by the respondent as not being similar to United States sales. 10 CIT at 98, 630 F. Supp at 1338. Here, Commerce examined all of the information and explanations submitted by Foroni, and after verification, agreed with Foroni’s assessment. Agreement is not the same as abdication. Thus, the Court concludes that Commerce’s actions were supported by substantial evidence on the agency record and in accordance with law. Conclusion For the foregoing reasons, the Court remands this case to Commerce to recalcaulate the VAT adjustment and apply a weighted-average tax rate to certain home market comparison groups and U.S. prices. Commerce is sustained as to all other issues. Remand results are due within sixty (60) days of the date this opinion is entered. Any comments or responses are due within thirty (30) days thereafter. Any rebuttal comments are due within fifteen (15) days after the date responses or comments are due. Footnotes [1] Appendix V lists the six criteria relied upon by Commerce on the following hierarchial basis: (1) general type of finish-, (2) grade; (3) remelting; (4) type of finishing operation; (5) size and (6) shape. See Letter from David L. Binder, Antidumping Division H (Mar. 7, 1994), P.R. Doc. No. 57, Def.’s App., Ex 1. [2] Because the Court finds Commerce’s decision to exclude Valbruna’s B-2 and B-3 sales from the final margin calculations to be supported by substantial evidence on the agency record, the Court will not address the exhaustion issue raised by Commerce. [3] The Court also recognizes the possibility that the London Metal Exchange prices may not accurately represent the actual costs to Valbruna for purchasing scrap metal and nickel (another component of stainless steel bar). See Rebuttal Brief of Acciaierie Valbruna S.r.L., C.R. Doc. No. 73, Valbruna’s App., Ex. 15, at 16. [4] The Court also rejects AL Tech’s assertion that Valbruna would not want to set aside merchandise for customers due to economic reasons. AL Tech’s argument amounts to business advice rather than evidence of Valbruna’s actions. While, generally, fungible merchandise would not be held for particular customers, a company may decide to grant certain customers special treatment in order to maintain good customer relations. VaJbruna provided sufficient evidence and explanations supporting its claim that it does indeed hold specific merchandise for particular customers.