Timken Co. v. United States

240 F. Supp. 2d 1228, 26 Ct. Int'l Trade 1072, 26 C.I.T. 1072, 24 I.T.R.D. (BNA) 1964, 2002 Ct. Intl. Trade LEXIS 106

Opinion

OPINION

POGUE, Judge.

This consolidated action is before the Court on cross-motions for judgment on the agency record, pursuant to USCIT Rule 56.2. The parties challenge aspects of the Department of Commerce’s (“Commerce” or “the Department”) final results regarding sales at less than fair value (“LTFV”) of Tapered Roller Bearings (“TRBs”) from Japan covering the period of October 1, 1998 through September 30, 1999. See Twpered Roller Bearings and Parts Thereof, Finished and Unfinished, From Japan, and Tapered Roller Bearings, Four Inches or Less in Outside Diameter, and Components Thereof, from Japan, 66 Fed.Reg. 15,078 (Dep’t Commerce Mar. 15, 2001) {“Final Results”) and the accompanying Issues and Decision Memorandum, P.R. Doc. No. 141 (Mar. 7, 2001) {“Decision Mem.”). The parties include several foreign and domestic producers of TRBs. The Court has jurisdiction over this matter pursuant to 19 U.S.C. § 1516a(a)(2)(B) and 28 U.S.C. § 1581(c).

Foreign TRB producers Koyo Seiko Ltd. and Koyo Corp. of America (collectively “Koyo”) claim (1) Commerce violated its international obligations by applying the “arm’s-length” test to exclude certain home market sales to affiliated customers; (2) Commerce violated its international obligations by “zeroing” the margins on negative-margin transactions when calculating Koyo’s weighted average dumping margins; and (3) Commerce erred in its treatment of imputed expenses in the calculation of profit for Koyo’s CEP sales. 1

Domestic producer The Timken Company (“Timken”) argues that (1) Commerce improperly calculated Koyo’s constructed export price (“CEP”) by applying adverse facts to Koyo’s entered value, rather than Koyo’s sales value; and (2) for purposes of a level of trade (“LOT”) adjustment to NTN’s normal values, Commerce erred in its decision to weight percentage differences in sales prices observed at different levels of trade by the sum of the quantities of sales at both levels of trade, rather than the lesser of the sales quantities of the two LOTs being compared. 2

In response to Timken’s second claim, NTN argues that Timken’s LOT adjustment claim presents no case or controversy, and therefore cannot be considered by this Court. See U.S. Const. Art. 3, § 2 (prohibiting issuance of advisory opinions).

Standard of review

The Court will uphold a final determination by Commerce in an antidumping investigation unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B).

Discussion

I. Commerce’s Application of Adverse Facts Available to Determine Koyo’s Dumping Margin

A. Background

An antidumping duty is imposed upon imported merchandise if that merchandise is sold or likely to be sold in the United States at less than fair value, and an industry in the United States is materially injured or is threatened with material injury. See 19 U.S.C. § 1673. To determine whether merchandise is being sold at less than fair value, Commerce compares the price of the imported merchandise in the United States to the normal value (“NV”) 3 for the same or similar merchandise in the home market. See 19 U.S.C. § 1677b. The United States price is calculated using either the export price (“EP”) or constructed export price (“CEP”). See 19 U.S.C. § 1677a(a), (b). Commerce uses a CEP if, “before or after the time of importation, the first sale to an unaffiliated person is made by (or for the account of) the producer or exporter or by a seller in the United States who is affiliated with the producer or exporter.” Uruguay Round Agreements Act, Statement of Administrative Action, H.R. Doc. No. 103-826 (1994), reprinted in 1994 U.S.C.C.A.A.N. 4040, at 822 (“SAA”). 4 Various adjustments may be made to CEP, including reduction by “the cost of any further manufacture or assembly” in the U.S. See 19 U.S.C. § 1677a(d)(2).

Here, Commerce chose to use CEP. 5 As there was value added to the subject merchandise in the United States after importation, Commerce required a Section E response from Koyo. 6 Koyo, however, chose not to file Section E of the questionnaire. See Letter from Koyo Seiko Co. to the Department of Commerce, P.R. Doc. No. 59 at 6 (May 2, 2000) (“Koyo’s Refusal Letter”). As a result of Koyo’s deliberate noncompliance, Commerce calculated Koyo’s CEP using adverse facts available. Commerce chose as adverse facts available the rate of 41.04 percent. Decision Mem. at 8. This was the cash deposit rate established in the 1993-94 administrative review, see Tapered Roller Bearings and Parts Thereof, Finished and Unfinished From Japan, and Tapered Roller Bear ings, Four Inches or Less in Outside Diameter, and Components Thereof, from Japan, 63 Fed.Reg. 20,585, 20,611 (Dep’t Commerce 1998), and the highest rate ever calculated for Koyo in any segment of the A-588-604 case. Decision Mem. at 8. Commerce applied this rate to the entered value of Koyo’s further-manufactured merchandise in order to calculate Koyo’s CEP.

While Commerce’s decision to use adverse facts is undisputed, Timken believes that Commerce’s application of adverse facts to Koyo’s entered value did not create a fully adverse inference. Timken’s Mem. Supp. Mot. J. Agency R. at 12 (“Timken’s Mem.”). Timken points out that Commerce used the same methodology here as in previous administrative reviews in which Koyo also refused to supply further-manufactured information. See id. at 8-12; see also Tapered Roller Bearings and Parts Thereof, Finished and Unfinished, from Japan, and Tapered Roller Bearings, Four Inches or Less in Outside Diameter, and Components Thereof, from Japan, 65 Fed.Reg. 11,767 (Dep’t Commerce March 6, 2000) (1997-98 review period); 63 Fed.Reg. 2,558 (Dep’t Commerce Jan. 15, 1998) (1995-96 review period). Timken argues that Koyo’s earlier noncompliance with this methodology suggests that Commerce should alter the methodology in order to obtain Koyo’s compliance. Timken’s Mem. at 12. Timken suggests that Commerce should apply the percentage rate to Koyo’s U.S. sales values, which would result in a higher dumping margin. Id. at 10. In essence, Timken argues that Commerce should have applied “a more adverse ‘facts available’ ” to calculate Koyo’s dumping margin. Id. at 12-13.

Commerce rejected Timken’s approach, explaining that the application of the 41.04 rate to Koyo’s sales value would be unduly punitive. Decision Mem. at 8. Commerce also points out that “Timken has failed to offer arguments or provide record evidence demonstrating that the rate selected is not reasonably adverse.” Id.

B. Analysis

Commerce’s application of the adverse facts available rate to the entered value rather than the sales value is consistent with Commerce’s regulation for determining assessment rates, which states that Commerce “normally will calculate the assessment rate by dividing the dumping margin found on the subject merchandise examined by the entered value of such merchandise for normal customs duty purposes.” 19 CFR § 351.212(b)(1) (emphasis added). Additionally, this Court has previously decided that CEP can be calculated by applying adverse facts available to a party’s entered value when there is further manufacturing. 7 See NTN Bearing Corp. of Am. v. United States, 26 CIT -, -, 186 F.Supp.2d 1257, 1315 (2002) (sustaining Commerce’s application of adverse facts available rate to Koyo’s entered value to determine the CEP of Koyo’s further manufactured merchandise). 8 We find no reason to change our position on this matter.

Even though the issue and parties are identical, Timken argues that this Court’s ruling in NTN Bearing II does not preclude application of adverse facts available to Koyo’s sales value. Timken’s Reply Br. Supp. Mot. J. Agency R. at 15. Timken suggests that the difference is in Commerce’s knowledge: before NTN Bearing II, Commerce could not have known Koyo would repeatedly decline to comply with Commerce’s requests for Section E data, whereas after NTN Bearing II, Commerce should have known that Koyo’s noncompliance would continue. Id. at 16. In other words, Timken argues, Commerce should alter its methodology (i.e. apply the percentage rate to sales values instead of entered values) in order to effectively induce Koyo’s compliance. Id.

This argument is flawed for several reasons. First, Commerce has increased Koyo’s rate since NTN Bearing II, from 36.21 percent, see Tapered Roller Bearings, 63 Fed.Reg. at 2,562, to the rate of 41.04 percent used here. Timken’s argument that Commerce’s methodology does not attempt to induce Koyo’s compliance fails to recognize the higher dumping margin imposed by this higher rate.

Second, although it is true that Commerce applied the same rate of 41.04 percent in this review as it did in a previous administrative review for Koyo that took place after NTN Bearing II, see Issue and Decision Mem., Tapered Roller Bearings, 65 Fed.Reg. 11,767 at Comment 1, Commerce is “not required by the statute to select a method that is ‘the most’ or ‘more’ reasonably adverse.” See NTN Bearing II, 26 CIT at -, 186 F.Supp.2d at 1315 (agreeing with Commerce that it need not apply the most or more adverse facts) (internal citations and quotations omitted). Rather, Commerce should adhere to the overriding goal of the antidumping law, which is not to create a punitive result, i.e., “unreasonably high rates with no relationship to the respondent’s actual dumping margin,” F.LLI De Ceceo Di Filippo Fara S. Martino S.p.A. v. United States, 216 F.3d 1027, 1032 (Fed.Cir.2000) (“De Cecco”), but rather to create a result that determines “current margins as accurately as possible.” Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed.Cir.1990); NTN Bearing Corp. v. United States, 74 F.3d 1204, 1208 (Fed.Cir.1995).

In using the 41.04 percent rate and applying that rate to Koyo’s entered value, Commerce is appropriately balancing this goal of accuracy against the risk of creating a punitive margin. Commerce specifically declined to apply adverse facts to Koyo’s sales value because it believed that such an application “would be unduly punitive, given that a substantial amount of value was added to the imported components in the United States.” Decision Mem. at 8. Commerce reasonably denied Timken’s suggestion to apply adverse facts to a value that has been substantially increased after importation because such an application could result in an unreasonably high dumping margin. Commerce’s decision therefore adheres to the purpose of and restrictions on adverse facts available. See De Cecco, 216 F.3d at 1032; Reiner Brack GmbH & Co. KG v. United States, 26 CIT -, 206 F.Supp.2d 1323 (2002).

Timken questions Commerce’s ability to determine that Timken’s suggested approach' would be punitive, because Koyo failed to supply any information upon which Commerce could make a fact-based estimate. Timken’s Mem. at 14. However, Commerce has “extensive experience with and knowledge of Koyo’s further-manufactured sales and the calculation of the value added in the United States with respect to these sales.” Decision Mem. at 6. Additionally, Koyo submitted data to Commerce, which Commerce verified, demonstrating that the value added in the United States substantially exceeds the value of the imported merchandise. See Letter from Koyo Seiko Co. to the Department of Commerce, P.R. Doc. No. 37 (Feb. 11, 2000), amended by Letter from Koyo Seiko Co. to the Department of Commerce, P.R. Doc. No. 59 at 2 (Oct. 2, 2000); Letter from Koyo Seiko Co. to the Department of Commerce, P.R. Doc. No. 55 at 2 (October 24, 2000). Therefore, even without Koyo’s Section E response, Commerce still had enough information to make a reasonable assessment of the impact of applying adverse facts. Commerce is “in the best position, based on its expert knowledge of the market and the individual respondent, to select adverse facts that will create the proper deterrent to noncooperation with its investigations and assure a reasonable margin.” De Cecco, 216 F.3d at 1032.

Finally, because Commerce determined that “a substantial amount of value was added to the imported components in the United States,” Timken’s suggested methodology — that Commerce apply its adverse facts to Koyo’s sales value — would effectively apply a dumping margin to value added after the merchandise was imported into the United States. Decision Mem. at 8. Such a result is contrary to the purpose of the anti-dumping investigation, which is to “determine whether dumping duties should be imposed on subject merchandise when it is imported into the United States.” Pesquera Mares Australes Ltda. v. United States, 266 F.3d 1372, 1374 (Fed.Cir.2001) (emphasis added).

In sum, Timken offers no evidence demonstrating that the rate selected is not reasonably adverse, nor any well-founded claim that Commerce’s chosen methodology is not in accordance with law. For these reasons, this Court upholds Commerce’s application of adverse facts available to Koyo’s entered value as supported by substantial evidence and in accordance with law.

II. Level of Trade (“LOT”) Adjustment for NTN

Because normal value is based on exporting country (“EC”) sales at the same LOT as the EP or CEP, 9 19 U.S.C. § 1677b(a)(7) directs Commerce to adjust normal value to account for any price differential between sales at different LOTs. In order to determine the amount of the adjustment, “Commerce for each NTN model sold at both LOTs in the [home market] calculated the difference between the weighted-average prices at the two LOTs as a percentage of the weighted-average price at the comparison LOT.” Def.’s Mem. Opp’n Pis’ Mot. J. Agency R. at 26 (“Def.’s Mem.”); 19 C.F.R. § 351.412(e). 10 The LOT adjustment was then calculated by applying the weighted-average percentage price difference to the normal value determined at the comparison LOT. Def.’s Mem. at 26; see generally Mem. from Deborah Scott, Case Analyst, Analysis Memorandum For Preliminary Results of the 1998-99 Review — NTN Corporation, P.R. 117 (Oct. 31, 2000). 11

Here, Commerce adjusted NTN’s EP sales for price differentials accounted for by different levels of trade. Timken argues that Commerce’s computer program used the sum of the sales of models at both levels of trade to weight prices, and that this practice produces erroneous results and provides respondents with an opportunity to “game the system with isolated single sales.” Timken’s Mem. at 22. 12 Rather, according to Timken, Commerce should weight the price based on the actual number of instances where there are actual price differences. Id. at 21-23. Timken poses several hypothetical sets of facts that it claims demonstrate that Commerce’s methodology produces distorted results. See, e.g., id. at 22-23.

Timken’s hypothetical examples, however, do not prove that Commerce’s methodology for calculating the LOT adjustment produces distortive and therefore unreasonable results in this instance. Aside from providing the hypothetical examples, Timken does not offer any evidence that Commerce’s weighted averages for NTN in this review were distorted. Nor does Timken accuse NTN of attempting to “game the system;” rather, Timken argues that it is possible that some unspecified party could take advantage of the system.

Moreover, in promulgating its implementing regulation, 19 C.F.R. § 351.412(e), Commerce considered proposals similar to Timken’s, that it “should base the amount of any adjustment on the pattern of consistent price differences, rather than on a weighted average.” Antidumping Duties; Countervailing Duties; Final Rule, 62 Fed.Reg. 27,296, 27,372 (May 19, 1997) (discussion of section 351.412(e)). Commerce, however, made a policy decision based on the SAA guidelines and rejected this approach. The SAA provides that “[a]ny adjustment under Section 773(a)(7)(A) [19 U.S.C. 1677b(a)(7)(A) ] will be calculated as the percentage by which the weighted-average prices at each of the two levels of trade differ in the market used to establish normal value.” SAA at 830. Because Commerce’s policy choice is reasonable, until a party presents actual evidence that the application of Commerce’s methodology is distortive and unreasonable, this Court will respect the agency’s legitimate policy decision. Suramerica de Aleaciones Laminadas, C.A. v. United States, 966 F.2d 660, 665 (Fed.Cir.1992).

III. Commerce’s Application of the 99.5 Percent Arm’s Length Test

As noted above, “normal value” is defined as “the price at which the foreign like product is first sold ... for consumption in the exporting country, in the usual commercial quantities and in the ordinary course of trade.” 19 U.S.C. § 1677b(a)(l)(B)(i). The Department’s regulations direct it to use sales data to calculate normal value if the Department is “satisfied that the price is comparable to the price at which the exporter or producer sold the foreign like product to a person who is not affiliated with the seller,” 19 C.F.R. § 351.408(c), thereby only including data from sales made at arm’s length, i.e., in the ordinary course of trade.

Commerce has consistently applied 19 C.F.R. § 351.403(c) through a “99.5 percent arm’s-length test.” Under this test, the Department compares, on a model-specific basis, the weighted average prices of home market sales of subject merchandise to affiliated customers with the weighted average prices of home market sales of the same model to unaffiliated customers. All home market sales to affiliated customers the weighted average prices of which are less than 99.5 percent of the weighted average prices of sales to unaffiliated customers are excluded from the calculation of normal value. Sales to affiliated customers at prices that are higher than 99.5 percent of the weighted average price of sales to unaffiliated customers are automatically included, unless the party can prove that those sales are aberrationally high.

In the Final Determination, the Department excluded sales by Koyo to affiliates that failed the Department’s “arm’s-length” test. All sales at prices above 99.5 percent of the weighted average price to unaffiliated customers, however, were included in the calculation of Koyo’s dumping margin. Koyo claims that Commerce’s application of the arm’s-length test violates Article 2.1 of the Agreement on Implementation of Art. VI of the General Agreement on Tariffs and Trade (“Anti-Dumping Agreement”), as interpreted in recent WTO dispute resolution proceedings, United States — Anti-Dumping Measures on Certain Hot-Rolled Steel Products from Japan, WT/DS184/R (Feb. 28, 2001) (“Hot Rolled Steel Panel Report”) and WT/ DS184/AB/R (July 24, 2001) (“Hot Rolled Steel Appellate Body Report”). See Koyo’s Mot. at 5-6.

Commerce claims that its arm’s length test should be upheld because this Court has previously sustained the test and because Koyo is precluded from seeking a remedy in this Court based on the Anti-Dumping Agreement, pursuant to 19 U.S.C. § 3512(c) and § 3538. Timken also argues that Koyo failed to exhaust its administrative remedies by never raising this issue during the proceeding before the Department. 13

A. Exhaustion

As a preliminary matter, the Court addresses Timken’s claim that Koyo failed to exhaust its administrative remedies. “The exhaustion doctrine requires a party to present its claims to the relevant administrative agency for the agency’s consideration before raising these claims to the Court.” Timken Co. v. United States, 26 CIT --,-, 201 F.Supp.2d 1316, 1340 (2002). There is, however, “no absolute requirement of exhaustion in the Court of International Trade in non-classification cases.” Consol. Bearings Co. v. United States, 25 CIT -,-, 166 F.Supp.2d 580, 586 (2001). Rather, Congress vested the Court with discretion, pursuant to 28 U.S.C. § 2637(d), to determine the circumstances under which it is appropriate to require the exhaustion of administrative remedies.

A party “may be excused from its failure to raise the issue before Commerce [where] Commerce in fact considered the issue.” FAG Kugelfischer Georg Schafer AG v. United States, 25 CIT -, -, 131 F.Supp.2d 104, 114 (2001). The same aspects of the arm’s-length test at issue here were raised by NTN in the administrative proceeding below. The danger of the Court deciding the issue before the Department has the opportunity to examine it at the administrative level is not present because the Department did indeed consider and reject an identical claim. Id.; see also Decision Mem. at 26-27. Therefore, even though Koyo may have failed to raise the issue in the administrative proceeding, it does not appear reasonable to require further exhaustion in this case.

Moreover, “the Court has exercised its discretion to obviate exhaustion where: (1) requiring it would be futile;” (2) “a subsequent court decision has interpreted existing law after the administrative determination at issue was published, and the new decision might materially affect the agency’s actions;” (3) “the question is one of law and does not require further factual development and, therefore, the court does not invade the province of the agency” by considering the question; or (4) plaintiffs “had no reason to suspect that the agency would refuse to adhere to ‘clearly applicable precedent.’ ” FAG Kugelfischer Georg Schafer AG v. United States, 25 CIT at -, 131 F.Supp.2d at 114.

Here, neither the WTO Panel Report nor the Appellate Body Report were issued until after Koyo filed its brief with the Department. 14 To require a party to anticipate the outcome of WTO decisions would be an unreasonable application of the exhaustion requirement. Although WTO Panel Reports and Appellate Body Reports are not binding on this Court, they may help inform the Court’s decisions, and therefore it is appropriate to review Koyo’s challenge to the Department’s application of its arm’s length policy in this matter.

B. 19 U.S.C. § 3512(c)

The Department does not address Koyo’s argument that the “arm’s-length” test violates 19 U.S.C. § 1677b’s “fair comparison” requirement and therefore contradicts obligations pursuant to the Anti-dumping Agreement. Rather, Commerce focuses on 19 U.S.C. § 3512(c), arguing that the statute prohibits private parties from challenging government action on the basis that it violates a WTO agreement. Koyo, however, is not bringing this action under any WTO agreement; rather, Koyo is arguing that the Department’s application and interpretation of U.S. law violates its international obligations pursuant to a WTO agreement.

Koyo is certainly “free to argue that Congress would never have intended to violate an agreement it generally intended to implement, without expressly saying so.” Gov’t of Uzbekistan v. United States, 2001 WL 1012780, *3 (CIT Aug. 30, 2001). As in Uzbekistan, the Department’s reb-anee on section 3512(c) is an “erroneous technical bar.” Id. Therefore, Koyo’s claim is appropriately before us.

C. WTO Panel Reports

The interaction between international obligations and domestic law is interesting and complex. While an unambiguous statute wib prevail over a conflicting international obligation, Federal-Mogul Corp. v. United States, 63 F.3d 1572, 1581 (1995), an ambiguous statute should be interpreted so as to avoid conflict with international obligations. See Murray v. Schooner Charming Betsy, 2 Cranch 64, 6 U.S. 64, 2 L.Ed. 208 (1804) (“[A]n act of Congress ought never to be construed to violate the law of nations, if any other possible construction remains.... ”). In the case of statutory interpretations by agencies, however, judicial review must take place within the confines of either Chevron or Skidmore deference. See United States v. Mead Corp., 533 U.S. 218, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001) (discussing Chevron U.S.A., Inc. v. Natural Res. Def. Council, 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984) and Skidmore v. Swift & Co., 323 U.S. 134, 65 S.Ct. 161, 89 L.Ed. 124 (1944)); but cf. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Const. Trades Council, 485 U.S. 568, 574-75, 108 S.Ct. 1392, 99 L.Ed.2d 645 (1988) (holding that Chevron may yield to the Charming Betsy doctrine).

This Court does not automatically assume that the WTO Panel and Appellate Body decisions are correct interpretations of United States obligations pursuant to the GATT. Rather, they are non-binding decisions, Hyundai Elec. Co., Ltd. v. United States, 23 CIT 302, 311, 53 F.Supp.2d 1334, 1343 (1999), the reasoning of which may help inform this Court’s decision. 15 23 CIT at 312, 53 F.Supp.2d at 1343.

Here, the WTO Appellate Body found that Commerce’s 99.5 percent arm’s-length test “does not rest on a permissible interpretation of the term ‘sales in the ordinary course of trade,’” found in Article 2.1 of the Anti-Dumping Agreement. Hot Rolled Steel Appellate Body Report ¶ 158 (quoting Hot Rolled Steel Panel Report ¶ 7.112). Article 2.1 of the Anti-Dumping Agreement provides:

For the purposes of this Agreement, a product is to be considered as being dumped, i.e. introduced into the commerce of another country at less than its normal value, if the export price of the product exported from one country to another is less than the comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country.

However, neither the WTO Panel nor the Appellate Body found that 19 U.S.C. § 1677b or 19 C.F.R. § 351.403 16 violated the Anti-Dumping Agreement. Rather, the two panels determined that the Department’s policy of applying the 99.5 percent arm’s-length test resulted in an inflated normal value and lacked “even-handedness.” Hot Rolled Steel Appellate Body Report ¶ 154. According to the panels, because all high-priced sales are included, unless the exporter demonstrates through a difficult process that a given sales price is aberrationally high, sales that are not in the ordinary course of trade are often included in the antidump-ing calculation. Furthermore, the Appellate Body noted that the Department “does not have any standard, nor even guidelines, for determining the threshold of aberrationally high” sales. Id. at ¶ 151. Moreover, it was “not clear to [the Appellate Body] that exporters would have known of the rule applied to high-priced sales.” Id. at ¶ 155. The result, according to the WTO decisions, disadvantages exporters. 17

The WTO decisions found, and this Court agrees, that the statute and the Department’s regulation are consistent with the Anti-Dumping Agreement. As a result, no direct conflict exists between provisions of U.S. law and international obligations. Therefore, we focus solely on the Department’s policy interpreting its statute and regulations. However, the ambiguity of the statutes and regulations as to the definition of “ordinary course of trade,” precludes a Chevron step-one analysis. 18 Accordingly, the court must determine if the Department’s interpretation is reasonable, as informed by Chevron step-two and Charming Betsy.

This Court has previously upheld Commerce’s arm’s-length test as a reasonable method for establishing a fair basis of comparison between affiliated and unaffiliated party sales. See, e.g., Usinor Sacilor v. United States, 18 CIT 1155, 1158-59, 872 F.Supp. 1000, 1004 (1994); Micron Tech. Inc. v. United States, 19 CIT 829, 846-47, 893 F.Supp. 21, 37-38 (1995). Although several parties have argued that the test fails to discover whether the investigated party actually manipulated prices charged to a related party, this Court held that it would continue to “uphold Commerce’s arm’s length test unless the test was shown to be unreasonable because it distorted price comparability.” SSAB Svenskt Stal AB v. United States, 21 CIT 1007, 1010, 976 F.Supp. 1027, 1030 (1997). The Court has also explicitly rejected the notion that the arm’s length test is flawed because it does not take into account certain qualitative factors other than price. NSK Ltd. v. United States, 21 CIT 617, 628-29, 969 F.Supp. 34, 48 (1997). These cases, however, do not appear to consider whether it is reasonable to apply a test that automatically excludes prices below 99.5 percent while automatically including prices above 99.5 percent.

Investigating authorities, pursuant to their obligations under the Anti-Dumping Agreement, need to verify whether sales are made in the ordinary course of trade. The authorities cannot presume that all affiliate sales are outside the ordinary course of trade; in some circumstances, this may not be the case. As the WTO panels note, however, “in the ordinary course of trade” is not a phrase defined by the Antidumping Agreement. Therefore, investigating authorities have the discretion to choose different methods of testing whether a sale is made within the ordinary course of trade.

Here, Commerce determined that the 99.5 percent arm’s length test appropriately excludes sales made outside the ordinary course of trade. The Department excludes sales for which the price ratio is less than 99.5 percent because “on average, that customer was paying less than unrelated customers for the same merchandise.” Usinor Sacilor, 18 CIT at 1157, 872 F.Supp. at 1003 (1994) (internal citations and quotations omitted). The Appellate Body agreed that “a pattern of prices to affiliated customers, different from the pattern of prices to unaffiliated customers, could indicate that sales were not in the ordinary course of trade.” Hot Rolled Steel Appellate Body Report ¶ 135. We agree with the Department that it is reasonable to presume that affiliate sales with a pattern of below average prices are not in the ordinary course of trade. The Appellate Body, however, expressed concern over the fact that the 99.5 percent arm’s length policy only determines whether sales to affiliates are, on average, at lower prices than sales to unaffiliated parties, not whether prices might be on average higher to affiliates. Accordingly, what this Court must next consider is whether higher priced sales should be automatically included.

Although higher priced sales are presumed to be included in the calculation of normal value, they may be excluded upon a showing that they are aberrationally high. The Appellate Body was concerned that “[t]he rule applied to high-priced sales ... was not contained in any guidelines, or other document conveyed to the interested parties. It is, therefore, not clear to us [the Appellate Body] that exporters would have known of the rule applied to high-priced sales.” Id. at ¶ 155. Here, however, Koyo concedes that it had notice that Commerce excluded sales demonstrated to be “aberrational.” Oral Arg. Trans, at 9-11 (Aug. 2, 2002). In fact, in this same administrative review, NTN tried to demonstrate that some of its high profit sales were outside the ordinary course of trade. See Timken’s Oral Arg. Ex. 3.

Commerce argues that the investigated parties should have the burden of establishing that high priced sales between affiliated parties are not in the ordinary course of trade. According to Commerce, once a party establishes that the high priced sales are aberrational, the sales are excluded from the calculation of normal value. Commerce applies this asymmetric test because it assumes that investigated parties will supply advantageous information, such as why a high priced sale between affiliated parties is not in the ordinary course of trade, but will be reluctant to supply information that is disadvantageous, such as why low priced sales between affiliated parties are not in the ordinary course of trade. Furthermore, according to Commerce, “[t]he purpose of an arm’s length test is to eliminate prices that are distorted. We test sales between two affiliated parties to determine if prices may have been manipulated to lower normal value.” Antidumping Duties; Countervailing Duties; Final Rule, 62 Fed.Reg. 27,296, 27,356 (May 19,1997).

It may be that Commerce’s application of the 99.5 percent arm’s length test could, in another case, lack even-handedness and disadvantage exporters so as to be inconsistent with international obligations under the Anti-Dumping Agreement. In this case, however, we do not find, nor does Koyo argue, that the application of the 99.5 percent arm’s length test results in the inclusion of sales outside the ordinary course of trade in the calculation of Koyo’s normal value. Accordingly, because in this case investigated parties control the data at issue, we uphold Commerce’s application of its statutes and regulations as a reasonable interpretation of “ordinary course of trade.”

IV. Commerce’s Practice of Zeroing Negative Margins

Koyo also argues that the Department erred by refusing “to give full mathematical effect to the negative margins ... in calculating Koyo’s (and other respondents’) weighted average margins, by setting the negative margins on those transactions at zero.” Koyo’s Mot. at 34. Koyo points to a recent Appellate Body decision, European Communities — Antidumping Duties on Imports of Cotton-Type Bed Linen from India, WT/DS141/AB/R (Mar. 1, 2001) (“EC-Bed Linen Appellate Body Report”), in arguing that this practice is inconsistent with the Anti-Dumping Agreement. Id.

As with the “arm’s-length” test, the Department argues that Koyo is barred from seeking a remedy in this Court pursuant to 19 U.S.C. § 3512(c). The Department also claims that WTO cases are not binding on this Court and, more importantly, that to date, the WTO cases have not decided the issue with respect to the zeroing practice of the U.S. Finally, Timken claims that the zeroing practice actually ensures a more accurate antidumping margin.

As discussed above, 19 U.S.C. § 3512(e) does not bar Koyo’s claim. This action is commenced under U.S. law, and a party may reasonably assume that the agency will interpret U.S. law so as to avoid a conflict with international obligations.

EC Bed-Linen involved the European Community’s practice of zeroing “when establishing ‘the existence of margins of dumping.’ ” EC Bed-Linen Appellate Body Report ¶¶ 45(a), 47. The Appellate Body found EC’s “zeroing” practice to be inconsistent with Article 2.4.2 of the EC Anti-Dumping Agreement. 19 Koyo argues that the EC’s practice is similar to the one used by the Department.

As Koyo concedes, the Department’s zeroing practice has previously been affirmed by this Court, which found it to be a reasonable interpretation of 19 U.S.C. § 1673. Seram/pore Indus. Pvt. Ltd. v. Dep’t of Commerce, 11 CIT 866, 874, 675 F.Supp. 1354, 1360-61 (1987); Bowe Passat Reinigungs-Und Waschereitechnik GmbH v. United States, 20 CIT 558, 572, 926 F.Supp. 1138, 1150 (1996). The Court, however, has also stated that it would only continue to uphold the Department’s practice of zeroing “until it becomes clear that such a practice is impermissible.” Bowe Passat, 20 CIT at 572, 926 F.Supp. at 1150.

The EC Bed Linen report does not invalidate Commerce’s zeroing practice. The Appellate Body decision involved a dispute between India and the European Communities, and did not comment on U.S. practices. To date, no comparable WTO case has been decided concerning U.S. zeroing practices. Moreover, although the EC’s zeroing practice appears similar to the United States’ practice, this Court cannot determine from the Appellate Body report whether they are the same. As noted above, according to the SAA, only the ministerial body of the WTO can interpret an Appellate Body report. See SAA at 662 (discussing procedures for making decisions). It is therefore not the province of this Court to determine the extent of the similarities between EC and U.S. zeroing practices based on the Appellate Body decision.

Furthermore, the EC-Bed Linen decision involved a comparison, made during an antidumping investigation, of weighted averages for export prices and normal value, while the instant case involves a comparison, made during an administrative review, of weighted-average normal values to transaction-specific export prices. Decision Mem. at 33. The Appellate Body was limited to interpreting Article 2.4.2. An administrative review, such as the one at issue here, however, is governed by Article 9.3.1 of the Anti-Dumping Agreement. 20 Although the two proceedings are related, involving the calculation of anti-dumping margins, differences exist between them and each investigation is informed by a different article of the Anti-Dumping Agreement. Here, the statutes require Commerce to calculate a “dumping margin for each such entry.” 19 U.S.C. § 1675(a) (2) (A) (ii). “Dumping margin” is defined by 19 U.S.C. § 1677(35)(A) as “the amount by which the normal value exceeds the export price or constructed export price of the subject merchandise.” As the statute requires the calculation to be on an entry-by entry approach, and as previous cases determined, Commerce’s practice is a reasonable interpretation of 19 U.S.C. § 1675(a)(2)(A), and continues to be a reasonable interpretation of the statute despite the WTO Panel report. Therefore, EC-Bed Linen does not inform the Court on the issue of an administrative review of an existing order.

Therefore, the Appellate Body’s decision in EC-Bed Linen does not compel a change to this Court’s holding in Bowe Passat, 20 CIT at 572, 926 F.Supp. at 1150, that the Department’s zeroing practice is upheld “until it becomes clear that such practice is impermissible.”

V. Commerce’s Treatment of Imputed Expenses in the Calculation of Koyo’s Constructed Export Price

Koyo argues that the Department’s treatment of imputed expenses in the calculation of constructed export price (“CEP”) sales is not in accordance with law. The Department contends that the Court should not consider Koyo’s challenge because Koyo failed to exhaust its administrative remedies and, even if this Court does consider Koyo’s challenge, the Department properly excluded imputed credit and inventory carrying costs in its calculation of CEP.

As previously discussed, supra page 1238, there is “no absolute requirement of exhaustion,” except in classification cases. Consol. Bearings Co. v. United States, 25 CIT at -, 166 F.Supp.2d at 586. The court has the discretion to excuse the failure to exhaust administrative remedies “where appropriate,” including when it would be futile to follow the administrative remedy. 28 U.S.C. § 2637(d). Exhaustion is futile when the agency (1) consistently applies the challenged policy or methodology; (2) issues rules, regulations or bulletins promulgating such policy or methodology; and (3) rejects similar challenges. See Koyo Seiko Co. v. United States, 26 CIT -,-, 186 F.Supp.2d 1332, 1339 (quoting Von Hoffburg v. Alexander, 615 F.2d 633, 638 (5th Cir.1980)). According to Koyo, “the well-established ‘futility’ exception to the exhaustion requirement” applies here. Koyo’s Reply Br. Supp. Mot. J. Agency R. at 27; see also Asociacion Colombiana de Exportadores v. United States, 916 F.2d 1571, 1575 (Fed.Cir.1990) (explaining that “following the administrative remedy would be futile because of certainty of an adverse decision”) (internal citations and quotations omitted).

The Department’s practice for calculating CEP profit is well-established. In 1997, the Department issued a policy bulletin explaining its methodology for the calculation of profit for CEP transactions. See Import Administration Policy Bulletin 97-1: Calculation of Profit for Constructed Export Price Transactions at Step 2. The bulletin made clear that the Department’s policy is to include imputed costs such as inventory carrying costs and credit costs in “total U.S. selling expenses” but not in “total expenses.” Id. Furthermore, the Department has consistently excluded imputed expenses from “total expenses” while including them in “total U.S. selling expenses” in other antidumping proceedings. See, e.g., Antifriction Bearings (Other than Tapered Roller Bearings) and Parts Thereof from France, et al., 64 Fed.Reg. 35,590, 35,623 (July 1, 1999) (final determ.); Antifriction Bearings (Other than Tapered Roller Bearings) and Parts Thereof from France, et al., 62 Fed.Reg. 54,043, 54,072 (Oct. 17, 1997) (final admin, rev.). Accordingly, as Commerce’s position has been both issued formally, as a policy bulletin, and consistently applied, we conclude that the policy is so well-established that it would have been futile for Koyo to raise the issue in the administrative proceeding below. See NTN Bearing Corp. I, 155 F.Supp.2d at 743. As a result, the Court finds that Koyo was excused from exhausting its administrative remedies in this case.

B. Accordance with Law

Title 19 U.S.C. § 1677a(f) defines “total United States expenses” and “total expenses.” “Total United States expenses” refers to “the total expenses described in subsection (d)(1) [commissions for selling the subject merchandise in the U.S., expenses resulting from and bearing a direct relationship to the sale, any selling expenses that the seller pays on behalf of the purchaser and any other selling expenses] and (2) [cost of further manufacture or assembly] of this Section.” “Total expenses,” on the other hand, includes, in relevant part:

all expenses in the first of the following categories which applies and which are incurred by or on behalf of the foreign producer and foreign exporter of the subject merchandise and by or on behalf of the United States seller affiliated with the producer or exporter with respect to the production and sale of such merchandise:

(i) The expenses incurred with respect to the subject merchandise sold in the United States and the foreign like product sold in the exporting country if such expenses were requested by the administering authority for the purpose of establishing normal value and constructed export price.

(ii) The expenses incurred with respect to the narrowest category of merchandise sold in the United States and the exporting country which includes the subject merchandise.

(iii) The expenses incurred with respect to the narrowest category of merchandise sold in all countries which includes the subject merchandise.

19 U.S.C. § 1677a(f)(C). In interpreting these two provisions, Commerce includes imputed credit and inventory costs in U.S. total expenses as an expense having a direct relation to the sale. Commerce, however, does not impute credit and inventory expenses in total expenses where total expenses include actual credit and inventory costs.

In NTN Bearing I this Court addressed the Department’s practice of excluding imputed expenses from “total expenses” but including them in “total U.S. expenses.” The NTN Bearing I court found that the Department’s practice ignored the plain language of the statute and that the Department must include imputed credit and inventory carrying costs in total expenses when they are included in total U.S. expenses. NTN Bearing I, 155 F.Supp.2d at 748. The Department, however, contends that NTN Bearing I is not dispositive of the issue because it conflicts "with the appellate decision in U.S. Steel Group v. United States, 225 F.3d 1284 (Fed.Cir.2000).

In U.S. Steel Group, the Court of Appeals for the Federal Circuit found that symmetry between “total expenses” and “total U.S. expenses” was not necessary, in which case movement expenses could be included in one but not the other. Furthermore, the court found that total U.S. expenses were not a subset of total expenses. The Department now argues that U.S. Steel Group stands for the proposition that symmetry need not exist in the ratio for CEP transactions used here.

Unlike the court in NTN Bearing II, we do not believe that the statute clearly addresses the use of imputed expenses in the calculation of total expenses or total actual profit. Furthermore, we believe, as held in U.S. Steel Group, that Congress defined total United States expenses and total expenses differently. We agree with the Department that although the court in U.S. Steel Group focused on “movement expenses,” the reasoning of that case is applicable here.

The Federal Circuit in U.S. Steel Group looked at the relevant statutory provisions, 19 U.S.C. § 1677a, as a whole. The court held that “[t]he statute itself defines ‘total U.S. expenses’ distinctly, both structurally and substantively, from ‘total expenses.’ ” 225 F.3d at 1289. Although the court focused on movement expenses, it still found that total expenses and total U.S. expenses were defined “very differently.” Here, although the definitions of both total U.S. expenses and total expenses direct Commerce to include a figure for selling expenses, it is not clear from the statute that these figures need to be precisely the same.

Furthermore, even if U.S. Steel does not apply to selling expenses, Commerce’s methodology is a reasonable interpretation of the statute. In this situation, Commerce included a category of expenses, inventory and credit costs, when calculating both total U.S. expenses in the numerator and total expenses in the denominator of the ratio. As this Court explained in Thai Pineapple, imputed selling expenses when included in calculating total U.S. expenses also need to be included in the calculation for total expenses “unless they are already represented in total expenses in some other fashion.” Thai Pineapple Canning Indus. Corp., Ltd. v. United States, 23 CIT 286, 296, 2000 WL 174986 (2000), aff'd in part, rev’d in part, 273 F.3d 1077 (Fed.Cir.2001) (“Thai Pineapple I”) 21 (emphasis added) (citing U.S. Steel, 15 F.Supp.2d at 898). 22 Here, Koyo provided Commerce with both “imputed” numbers representing inventory and credit costs on a per-model basis for U.S. sales and “actual” numbers for total credit and inventory costs. 23 Commerce excluded imputed credit and inventory carrying costs from total expenses and total actual profit because these expenses were already accounted for; total expenses merely uses actual figures while U.S. expenses uses imputed. Therefore, the category of expenses at issue — inventory and carrying costs — are included in both total expenses and total U.S. expenses. This is consistent with U.S. Steel and the Thai Pineapple cases. See, e.g., Thai Pineapple I, 23 CIT at 289, 2000 WL 174986 (holding that “imputed expenses should be omitted from actual profit if they duplicate expenses already accounted for”).

This practice is further supported by Commerce’s, preference for the use of actual cost information rather than imputed cost information when possible. See, e.g., Antidumping Manual, Chap. 8 at 23-25 (“Our preference is to use actual credit cost information if it is available. If actual expenses are not available, we impute the cost of credit.... ”). 24 Rather than using a proxy, actual figures for the interest expenses of inventory and credit costs were included in the calculation of total expenses and total actual profit. While the imputed numbers used in total U.S. expenses may not be exactly the same as those used in total expenses, one is a reasonable surrogate for the another. See, e.g., Thai Pineapple II, 2000 WL 174986, *6 (“Theoretically, the total expenses denominator would reflect the interest expenses captured in the U.S. sales expenses numerator ... as well as ‘home’ market interest expenses, because the total expenses denominator is derived from a net unit figure based on all company interest expenses without regard to sales destination.”). Moreover, “[c]ompanies may not keep track of the costs of maintaining inventory or extending credit to their customers on a per-model basis. Nonetheless, they are real costs that a company incurs. The Department asks respondents to provide measures of these costs, to ‘impute’ them for purpose of determining normal value and U.S. price.” Timken’s Resp. to Koyo’s Mot. J. Agency R. at 39. Therefore, even if total U.S. expenses are a subset of total expenses for selling cost purposes, inventory and credit costs are accounted for in both parts of the ratio.

Accordingly, this Court, consistent with the federal circuit’s analysis in U.S. Steel Group, upholds the Department’s practice of excluding imputed expense in “total expenses” when actual expenses are used as that practice was applied here.

Conclusion

The Department’s final results are, therefore, affirmed as being supported by substantial evidence and otherwise in accordance with law.

1 . Koyo initially also argued that Commerce erred by using "adverse facts available” for calculating margins on subject merchandise further processed in the United States. See Koyo's Am. Compl. at 4-5. Koyo reasoned that because it met the criteria described in 19 U.S.C. 1677a(e), Commerce was no longer authorized to request Section E data. See Koyo's Mem. Supp. Mot. J. Agency R. at 16-17 ("Koyo's Motion”); see also note 6, infra. As a result, Koyo believed its noncompliance was justified, and thus, application of adverse facts available would be improper. Koyo's Mot. at 13-14. Prior to oral argument, Koyo abandoned this claim. See Letter from Sid-ley, Austin, Brown & Wood to United States Court of International Trade at 1 (July 31, 2002).

2 . Timken also alleged in its complaint that "[t]he ITA made other clerical errors in calculating the final results that implicate business proprietary information or the calculation methodology used to reach the final results of the administrative review.” Timken's Compl. ¶ 6(d). Timken’s subsequent Rule 56.2 Motion, however, is limited to its disagreement with treatment of Koyo's further manufactured merchandise and NTN's LOT adjustment. In it's brief, Timken abandoned its other claims. NSK Ltd. and NSK Corp. asks that any action by Timken effecting NSK's rights in this matter be dismissed. Because "any claim which is not pressed is deemed abandoned,” De Laval Separator Co. v. United States, 1 CIT 144, 146, 511 F.Supp. 810, 812 (1981), we dismiss Timken’s action as to NSK.

3 . NV is "the price at which the foreign like product is first sold ... for consumption in the exporting country, in the usual commercial quantities and in the ordinary course of trade.” 19 U.S.C. § 1677b(a)(1)(B)(i).

4 . The SAA is "an authoritative expression by the United States concerning the interpretation and application of the Uruguay Round Agreements and this Act in any judicial proceeding in which a question arises concerning such interpretation or application.” 19 U.S.C. § 3512(d).

5 . Commerce's decision to use CEP is unchallenged by any of the parties.

6 . Section E contains a request for sales and cost information for Koyo's further-manufactured sales. See Commerce's Request for Information at Section E, P.R. Doc. No. 14 at E-2.

7 . The Court’s previous decision to uphold the application of adverse facts available to the entered value of subject merchandise was in response to motions from the same parties as in this case.

8 . We refer frequently to several cases entitled NTN Bearing Corp. v. United States. Only two of these cases are discussed extensively. References to these two cases will be abbreviated in chronological order. See NTN Bearing Corp. of Am. v. United States, 25 CIT -, -, 155 F.Supp.2d 715 (2001), appeal docketed, Nos. 02-1180, 02-1181 (Fed.Cir. Feb. 5, 2002) ("NTN Bearing I"); NTN Bearing Corp. of Am. v. United States, 26 CIT -,-, 186 F.Supp.2d 1257 (2002) ("NTN Bearing II”).

9 . Sales are made at different levels of trade "if they are made at different marketing stages (or their equivalent).” Antidumping Manual, Chap. 8 at 53; 19 C.F.R. § 351.412(c)(2).

10 . Commerce's implementing regulation for LOT adjustments provides:

(e) Amount of adjustment. The Secretary normally will calculate the amount of a level of trade adjustment by:

(1)Calculating the weighted-averages of the prices of sales at the two levels of trade identified in paragraph (d), after making any other adjustments to those prices appropriate under section 773(a)(6) of the Act and this subpart;

(2) Calculating the average of the percentage differences between those weighted-average prices; and

(3) Applying the percentage difference to normal value, where it is at a different level of trade from the export price or constructed export price (whichever is applicable), after making any other adjustments to normal value appropriate under Section 773(a)(6) of the Act and this subpart.

19 C.F.R. § 351.412(e).

11 . This practice has been applied consistently by Commerce.

12 . NTN Bearing claims that Timken presents "no case or controversy” and is asking for an "advisory opinion.” NTN’s Resp. to Timken's Mem. Supp. Mot. L Agency R. at 4. "In order to satisfy the case or controversy requirement 'a litigant must have suffered some actual injury that can be redressed by a favorable judicial decision.’ ” Verson v. United States, 22 CIT 151, 153, 5 F.Supp.2d 963, 966 (1998) (quoting Iron Arrow Honor Soc. v. Heckler, 464 U.S. 67, 70, 104 S.Ct. 373, 78 L.Ed.2d 58 (1983)). While Timken does not cite to specific record evidence regarding NTN, Timken does argue that Commerce’s LOT adjustment produces erroneous results whenever it is used, including Commerce's application of the test to NTN's LOT adjustment. Therefore, a case or controversy does exist.

13 . Although the Department argues that Koyo failed to exhaust its administrative remedies for several other claims in Koyo's complaint, there is no mention of this argument by the government with regards to the "arm’s-length” test.

14 . Koyo filed its case brief with the Department on December 7, 2000. The WTO Panel report, however, was not issued until February 28, 2001 and the Appellate Body report was issued on July 24, 2001.

15 . The ministerial body of the WTO is the only body that can interpret an Appellate Body report. See SAA at 662. Furthermore, the response to "an adverse WTO panel report is the province of the executive branch and, more particularly, the Office of the U.S. Trade Representative.” Hyundai Elecs. Co., Ltd., 23 CIT at 312, 53 F.Supp.2d at 1343; 19 U.S.C. § 3538.

16 . As noted above, the Department's regulations provide that it "may calculate normal value based on that sale [to an affiliated party] only if satisfied that the price is comparable to the price at which the exporter or producer sold the foreign like product to a person who is not affiliated with the seller.” 19 C.F.R. § 351.403(c).

17 . The Appellate Body recommended that "the United States bring its measures found in this Report, and in the Panel Report as modified by this Report, to be inconsistent with the Anti-Dumping Agreement and the WTO Agreement, into conformity with its obligations under those Agreements.” Hot Rolled Steel Appellate Body Report ¶241. The United States stated its intention to implement the recommendations and rulings of the Dispute Settlement Body in the Hot-Rolled Steel rulings. "After the DSB adopted its recommendations and rulings on August 23, the United States stated its intention to implement them in a manner consistent with its WTO obligations and engaged in discussions with Japan pursuant to Article 21.3(b) in an effort to reach agreement on the reasonable period of time for U.S. implementation.” Submission of the United States, Arbitration on the "Reasonable Period of Time,” United States — Anti-Dumping Measures on Certain Hot-rolled Steel Products from Japan, ¶ 1 (Jan. 4, 2002). During oral argument for the arbitration proceeding determining the proper amount of time for implementation of the Appellate Body decision, the United States represented "that modification of the '99.5 percent’ or 'arm’s length' test applied in practice by its administrative officials has already been commenced.” Arbitration, United States — Anti-Dumping Measures on Certain Hot-rolled Steel Products from Japan, WT/ DS184/13 (Feb. 19, 2002) ¶33. (holding that the period of time for implementation "which covers both legislative and administrative phases ...” will accordingly expire on 23 November 2002). Furthermore, on August 15, 2002, the Commerce Department, in the Federal Register, published a request for comments on a proposed change in its arm’s length policy. As a result, this Court is in the unfortunate position of reviewing a policy that Commerce has already decided to modify. Nothing in this opinion should be construed as limiting the Department's obligations in this regard.

18 . As discussed supra page 1239, determination of whether the agency's statutory interpretation is in accordance with law follows the two-step analysis formulated in Chevron, 467 U.S. at 837, 104 S.Ct. 2778. If the statute is clear, "that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.” Id. at 842-43, 104 S.Ct. 2778. If the statute is ambiguous and is expressed in a format that carries the "force of law,” Christensen v. Harris County, 529 U.S. 576, 120 S.Ct. 1655, 146 L.Ed.2d 621 (2000), the agency’s interpretation of the statute is entitled to deference as long as it is reasonable. Chevron, 467 U.S. at 843-44, 104 S.Ct. 2778.

19 . Article 2.4.2 of the Anti-Dumping Agreement provides:

Subject to the provisions governing fair comparison in paragraph 4, the existence of margins of dumping during the investigation phase shall normally be established on the basis of a comparison of a weighted average normal value with a weighted average of prices of all comparable export transactions or by a comparison of normal value and export prices on a transaction-to-transaction basis. A normal value established on a weighted average basis may be compared to prices of individual export transactions if the authorities find a pattern of export prices which differ significantly among different purchasers, regions or time periods, and if an explanation is provided as to why such differences cannot be taken into account appropriately by the use of a weighted average-to-weighted average or transaction-to-transaction comparison.

20 . Article 9.3.1, provides:

When the amount of the anti-dumping duty is assessed on a retrospective basis, the determination of the final liability for payment of the anti-dumping duties shall take place as soon as possible, normally within 12 months, and in no case more than 18 months, after the date on which a request for a final assessment of the amount of the anti-dumping duty has been made. Any refund shall be made promptly and normally in not more than 90 days following the determination of final liability made pursuant to this sub-paragraph. In any case, where a refund is not made within 90 days, the authorities shall provide an explanation if so requested.

Anti-Dumping Agreement, art. 9.3.1.

21 . Although Commerce cites to Thai Pineapple I as adverse to its position we are of a contrary mind.

22 . Thai Pineapple I was remanded to Commerce in order for the agency to "demonstrate ... that the total expense denominator of the ratio to be applied to total actual profit to obtain the CEP profit adjustment contains all interest expenses (including those relating to U.S. sales) as required by 19 U.S.C. § 1677a(f)(2)(C).” Thai Pineapple Canning Indus. Corp. Ltd. v. United States, 2000 WL 174986, **5-6 (CIT Feb. 10, 2000) (‘‘Thai Pineapple II”).

23 . Commerce requests respondents to report total interest expenses covering inventory car-lying costs and credit extension expenses for cost of production purposes. "For price adjustment purposes, however, Commerce requires respondents to impute interest expenses separately for U.S. sales, even though companies may not account for such expenses separately.” Thai Pineapple II, 2000 WL 174986, *6.

24 .It should be noted that while the Anti-dumping Manual "is not a binding legal document, it does give insight into the internal operating procedures of Commerce.” Koenig & Bauer-Albert AG v. United States, 24 CIT -, -, 90 F.Supp.2d 1284, 1292 n. 13 (2000), aff'd in part, vacated in part, remanded by 259 F.3d 1341 (Fed.Cir.2001).