A-100-001 AR:98/99 Public Document G1O3: AFB Team MEMORANDUM TO: Troy H. Cribb Acting Assistant Secretary for Import Administration FROM: Richard W. Moreland Deputy Assistant Secretary for Import Administration SUBJECT: Issues and Decision Memorandum for the Administrative Reviews of Antifriction Bearings (other than tapered roller bearings) and parts thereof from France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom - May 1, 1998, through April 30, 1999 Summary We have analyzed the case and rebuttal briefs of interested parties in the May 1, 1998, through April 30, 1999, administrative reviews of the antidumping duty orders covering antifriction bearings (other than tapered roller bearings) and parts thereof from France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom. As a result of our analysis, we have made changes, including corrections of certain inadvertent programming and clerical errors, in the margin calculations. We recommend that you approve the positions we have developed in the Discussion of the Issues section of this memorandum. Below is the complete list of the issues in these administrative reviews regarding which we received comments and rebuttal comments by parties: 1. Facts Available 2. Revocation 3. Export Price/CEP A. CEP Offset B. CEP Profit C. Other Expenses 4. Discounts and Rebates 5. Price Adjustments A. Indirect Selling Expenses B. Inventory Carrying Costs C. Credit Expenses D. Commissions E. Advertising Expenses F. Technical-Service Expenses G. Bank Charges H. Repacking Expenses I. Other Direct Selling Expenses 6. Level of Trade 7. Samples and Sales Outside the Ordinary Course of Trade 8. Cost of Production and Constructed Value A. Profit for Constructed Value B. Affiliated-Party Inputs C. General, Selling, and Administrative Expenses D. When to Use CV E. Inventory Write-offs F. Allowance for Doubtful Accounts G. Marketable Securities 9. Packing and Movement Expenses 10. Romania-Specific Issues 11. Miscellaneous A. Programming and Clerical Errors B. Date of Sale C. Sample Weeks D. Clerical Error in Respondent's Data Background On April 6, 2000, the Department of Commerce (the Department) published the preliminary results of the administrative reviews of the antidumping duty orders on antifriction bearings (other than tapered roller bearings) and parts thereof from France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom (65 FR 18033). The reviews cover 36 manufacturers/exporters. The period of review is May 1, 1998, through April 30, 1999. We invited interested parties to comment on our preliminary results. At the request of certain parties, we held hearings for Germany-specific issues on May 17, 2000, and for Japan-specific issues on May 22, 2000. Company Abbreviations Asahi - Asahi Seiko Co., Ltd. Barden - Barden Corp. (U.K.) Ltd. FAG Italy - FAG Italia S.p.A. (including all relevant affiliates) FAG Germany - FAG Kugelfischer Georg Shaefer AG FAG U.K. - FAG (U.K.) Ltd. IJK - Inoue Jukuuke Kogyo IKS - Izumoto Seiko Co., Ltd. INA - INA Walzlager Schaeffler oHG Koyo Japan - Koyo Seiko Co. Ltd. Koyo Romania - S.C. Romania S.A. KYK - Tottori Yamakei Bearing Seisakusho Ltd. MPT - Prazisionsteile GmbH Mittweida Nachi - Nachi-Fujikoshi Corp. NMB/Pelmec - NMB Singapore Ltd./Pelmec Industries (Pte.) Ltd. NPBS - Nippon Pillow Block Sales Co. NSK - NSK Ltd. NTN Germany - NTN Kugellager Fabrik (Deutschland) GmbH NTN Japan - NTN Corp. Paul Mueller - Paul Mueller GmbH and Co. KG SKF France - SKF Compagnie d'Applications Mecaniques, S.A. (Clamart) (including all relevant affiliates) SKF Germany - SKF GmbH (including all relevant affiliates) SKF Italy - SKF Industrie, S.p.A. (including all relevant affiliates) SKF Sweden - SKF Sverige AB (including all relevant affiliates) SNFA France - SNFA S.A. SNFA U.K.- SNFA (U.K.) Bearings, Ltd. (including all relevant affiliates) SNR France - SNR Roulements SNR Germany - SNR Roulements (exports of German-origin bearings) SNR U.K. - SNR Roulements (exports of U.K.-origin bearings) Somecat - Somecat, S.p.A. Takeshita - Takeshita Seiko Co., Ltd. TIE - Tehnoimportexport, S.A. Torrington - The Torrington Company Torrington Nadellager - Torrington Nadellager GmbH Other Abbreviations AFB - antifriction bearing BB - ball bearing CAFC - Court of Appeals for the Federal Circuit CEP - Constructed Export Price CIT - Court of International Trade COP - Cost of Production CRB - cylindrical roller bearing CV - Constructed Value G&A - General and Administrative Expenses LTFV - Less Than Fair Value NME - Non-Market Economy OEM - Original Equipment Manufacturer POR - Period of Review SAA - Statement of Administrative Action accompanying the URAA, H.R. Doc. 316, Vol. 1, 103D Cong. (1994) SG&A - Selling, General, and Administrative SPB - spherical plain bearing URAA - Uruguay Round Agreements Act AFB Administrative Determinations LTFV Investigation - Final Determinations of Sales at Less than Fair Value; Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof from the Federal Republic of Germany, 54 FR 19006 (May 3, 1989). AFBs 1 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof from the Federal Republic of Germany; Final Results of Antidumping Duty Administrative Review, 56 FR 31692 (July 11, 1991). AFBs 2 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al.; Final Results of Antidumping Duty Administrative Reviews, 57 FR 28360 (June 24, 1992). AFBs 3 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al.; Final Results of Antidumping Duty Administrative Reviews and Revocation in Part of an Antidumping Duty Order, 58 FR 39729 (July 26, 1993). AFBs 4 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews, Partial Termination of Administrative Reviews, and Revocation in Part of Antidumping Duty Orders, 60 FR 10900 (February 28, 1995). AFBs 5 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews and Partial Termination of Administrative Reviews, 61 FR 66472 (December 17, 1996). AFBs 6 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews and Partial Termination of Administrative Reviews, 62 FR 2081 (January 15, 1997). AFBs 7 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews and Partial Termination of Administrative Reviews, 62 FR 54043 (October 17, 1997). AFBs 8 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews and Partial Termination of Administrative Reviews, 63 FR 33320 (June 18, 1998). AFBs 9 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews, 64 FR 35590 (July 1, 1999). Discussion of the Issues 1. Facts Available Comment 1: NTN Germany argues that the Department's regulations require that the Department must publish preliminary results of a review including margins and an invitation for argument. NTN Germany comments that the preliminary notice for this review did not contain a margin for NTN Germany and that the review for NTN Germany had been rescinded. Thus, NTN Germany argues, the Department's notification after the preliminary results of its intention to apply a facts-available rate to NTN Germany was invalid because the Department did not follow its own regulations. Torrington observes that the Department published its mistaken belief that the review request on NTN Germany had been withdrawn but that, once it recognized the error, it sent the letter to NTN Germany describing the action it was taking. According to Torrington, the letter amends the erroneous notice of preliminary results. Torrington contends that the purpose of publishing preliminary results of review is to provide interested parties an opportunity to submit comments so that the Department may implement appropriate corrections or revisions of calculations and that the Department's May 2, 2000, letter to NTN Germany served this purpose. Department's Position: We published the preliminary results of review and invited argument. See Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al.; Preliminary Results of Antidumping Duty Administrative Reviews, Partial Rescission of Administrative Reviews, and Notice of Intent to Revoke Orders in Part, 65 FR 18033 (April 6, 2000). In the preliminary results we rescinded the review with regard to NTN Germany, based on the erroneous impression that the request for review of NTN Germany had been withdrawn. Promptly after issuance of the preliminary results, we realized that the request to review NTN Germany had, in fact, not been withdrawn. Thus, we had to re- instate the review with regard to NTN Germany and notify the affected parties of our intent to apply facts available to NTN Germany because NTN Germany had not replied to our questionnaire. To give NTN Germany a reasonable opportunity to comment on our intended application of facts available, we sent a letter to the firm providing notice of the re- instatement and the preliminary facts-available rate. We also, at NTN Germany's request, granted an extension of time for NTN Germany to submit its case brief. As a result, NTN Germany was given due process and an opportunity to participate fully in this administrative proceeding. Comment 2: NTN Germany argues that the Department should not have applied an adverse facts-available rate to NTN Germany for not responding to the Department's questionnaire. NTN Germany claims that it could not justify the economic expense of filing a complete response and that, given its past cooperation and low margins, the Department's use of an adverse inference was unwarranted. NTN Germany argues that the Department should have assigned it the rate it received in the most recent review. NTN Germany argues further that, even if an adverse inference is warranted, the use of a 12-year-old rate from another company violates the statutory requirement for corroboration of secondary evidence used to determine facts available. NTN Germany contends that Congress intended, in amending the statute, to moderate the Department's automatic use of the highest rate ever assigned in a particular case. Citing Ferro Union Inc. v. United States, 44 F. Supp.2d 1310, 1334 (1999) (Ferro), NTN Germany also claims that the CIT held that the Department "cannot select a rate which focuses only on inducing the exporter to cooperate and ignores the interest in selecting a margin which relates to the past practices of the industry." Furthermore, the respondent contends, the Department did not discuss other rates or state why such rates would be less probative of NTN's rate. Thus, NTN Germany concludes, if an adverse inference is warranted, the Department should use a rate previously calculated for NTN Germany and that the highest rate it had received previously would serve as an inducement to future cooperation. Torrington contends that, because NTN Germany did not act to the best of its ability to comply with the Department's request for information the application of adverse facts available is appropriate. Torrington also alleges that NTN Germany's reliance on Ferro is misplaced because that case concerned a respondent that furnished information. Torrington alleges that NTN Germany could have responded had it chosen to do so, citing its cooperation with the Department's requests for information in prior reviews. Torrington contends further that what NTN Germany is seeking is inappropriate because it would allow respondents to choose whether to answer questionnaires based on self-serving assessments of economic advantage. Finally, Torrington argues that, by choosing not to respond to the questionnaire, NTN Germany invites speculation that its dumping margin would have been large had it answered the questionnaire. Department's Position: NTN Germany did not respond to our request for information. Rather, NTN Germany submitted a letter in which it stated that it was not responding to our questionnaire because it was not economically feasible to do so given the volume and value of its sales during the POR. This is not cooperation or an attempt to cooperate with our request for information. Furthermore, the firm's cooperation in past reviews is unrelated to the question of its cooperation in this review. NTN Germany could have cooperated had it chosen to do so. Because it chose not to respond to our questionnaire, the use of total adverse facts available is appropriate. However, we find that the rate we preliminarily chose, the highest rate ever found for any company in this proceeding, is inappropriate, because we have been unable to corroborate it in the context of this administrative review and an alternative adverse rate which can be corroborated is available. Therefore, for these final results, as adverse facts available for NTN Germany, we have selected the highest calculated rate found for any company in any segment of this proceeding (i.e., BBs from Germany). This rate is 70.41 percent, which we found for FAG in the LTFV investigation. We find that it is sufficiently high as to reasonably ensure that NTN Germany does not obtain a more favorable result by failing to cooperate. In addition, we find that this rate has probative value. Although the rate was calculated for a period approximately 11 years earlier than the instant POR, as discussed below, it is within the range of dumping margins we have actually observed in this review with respect to this merchandise and thus can be corroborated by evidence on the record of this review. Because there are no independent sources for calculated dumping margins from a prior segment, it is not necessary to question the reliability of the margin for that time period. See, e.g., Extruded Rubber Thread From Malaysia; Final Results of Antidumping Duty Administrative Review, 65 FR 6140, 6141 (February 8, 2000). To further examine the probative value of the selected margin, we compared the selected margin to margins calculated on individual sales of German-origin BBs made by four companies in this review to determine whether such companies continue to dump subject merchandise in the United States at a 70.41 percent level. For these companies, we have found a substantial number of sales, made in the ordinary course of trade, in commercial quantities, with dumping margins near or exceeding 70.41 percent. (The details of this analysis are contained in the proprietary version of the NTN Germany analysis memorandum, dated July 28, 2000.) This evidence supports an inference that the selected rate might reflect NTN Germany's actual dumping margin. With respect to the relevance aspect of corroboration, the Department will consider information reasonably at its disposal as to whether there are circumstances that would render a margin not relevant. Where circumstances indicate that the selected margin is not appropriate, the Department will attempt to find a more appropriate basis for facts available. See, e.g., Fresh Cut Flowers from Mexico; Final Results of Antidumping Duty Administrative Review, 61 FR 6812, 6814 (Feb. 22, 1996) (Fresh Cut Flowers) (where the Department disregarded the highest margin as adverse best information available because the margin was based on another company's uncharacteristic business expense resulting in an unusually high margin). Further, in accordance with F.LII De Cecco Di Filippo Fara S. Martino S.p.A. v. United States, No. 99-1318 (CAFC June 16, 2000), we also examined whether any information on the record would discredit the selected rate as reasonable facts available for NTN Germany. There is no information on the record that demonstrates that the rate selected is not an appropriate total adverse facts-available rate for NTN Germany. In particular, there is no information, such as reliable evidence of NTN Germany's export prices, that might lead to a conclusion that a different rate would be more appropriate. Therefore, we consider the selected rate to have probative value with respect to NTN Germany in this review and to reflect an adverse inference. With regard to NTN Germany's contention that the rates it received in prior reviews would be more probative of its rate, the Department is not required to select the most probative rate for NTN Germany. In fact, such an effort would have required NTN Germany's cooperation in this review. Because we are applying an adverse inference with regard to NTN Germany, there is no reason for us to choose a lower rate than 70.41 percent, which is probative with respect to NTN Germany in this review and therefore corroborated. Comment 3: Torrington argues that the Department should apply adverse facts-available margins to all of NTN Japan's U.S. sales of bearing models which were also sold to home-market affiliates not reporting resale information. Torrington contends that the record supports a determination that: 1) NTN Japan did not act to the best of its ability in obtaining resale information from its home-market affiliates, and 2) NTN Japan itself was the cause of the affiliates' failure to report the sales data because NTN Japan declined to share its computer program for identifying the sales. Furthermore, Torrington notes that NTN had stated initially that it was unable to obtain this information from its affiliated resellers for proprietary reasons. Torrington concludes that the fact that at verification NTN Japan changed its explanation for not reporting the downstream sales calls into question NTN Japan's previous explanations. NTN Japan contends that the facts of this case do not meet the statutory conditions necessary for adverse facts available and that Torrington provided no evidence that NTN Japan's actions warrant such treatment. NTN Japan claims that it was unable to obtain the resale data from its affiliated resellers because the affiliates' records are not computerized or compatible with NTN's records. NTN Japan contends that, by reporting the sales to the resellers, it acted to the best of its ability to comply with the Department's requests for information. NTN Japan also asserts that its computer program is proprietary and that it should not be required to divulge it to other companies. Also, NTN Japan contends that, even if it did share the program with its affiliates, the resellers do not have the computer capability to run NTN Japan's program. Department's Position: NTN Japan did not act to the best of its ability in reporting sales by home-market affiliates. NTN Japan's statements of the reasons for why it could not obtain this information from those affiliates in which it holds a majority interest are not convincing. We are unable to go into further detail because of the proprietary nature of our analysis. See the NTN Japan final results analysis memorandum dated July 28, 2000, for our analysis. Because NTN Japan did not act to the best of its ability in responding to our requests for information on sales by affiliated resellers, we find appropriate the use of adverse facts available with regard to sales by resellers in which NTN Japan owns a majority interest. As adverse facts available, for each model sold to these affiliated resellers, we have replaced the price to the affiliated party with the highest home-market price at which NTN Japan sold the same model at the same level of trade to other customers during the POR. For models sold by these affiliated resellers that NTN Japan did not sell to other customers, we have increased the net home-market price to implement our adverse determination. To do so, we applied the weighted average of the ratio of the highest price for each model to the respective prices of those models sold to the affiliated resellers. In addition, because these facts available are our estimation of arms-length prices, we did not apply the arm's-length test to these sales. Comment 4: Gulf Bearing, an importer of BBs exported by KYK during the POR, states that, based on its timely and certified submission, the Department was able to establish that KYK went bankrupt during the POR and that KYK is not able to resume business operations in the future. Gulf Bearings argues that, because this information was submitted in the course of the review, the Department need not corroborate the information from independent sources in order to rely upon it in issuing its final results of reviews. Moreover, Gulf Bearings argues that, since no other party has questioned, challenged, or otherwise commented on Gulf Bearings' factual submissions, the Department should continue to rely on this information as the basis of applying the average margin for all of the Japanese firms involved in this administrative review as a non-adverse facts-available dumping margin for KYK. Department's Position: While we do not disagree with Gulf Bearing's argument that we may consider information submitted by it as to the financial viability of a reviewed producer, we consider it appropriate to examine information from more than one source in this case. Since the preliminary results, we have reviewed other sources seeking evidence of KYK's bankruptcy, including U.S. Customs entries. We requested information from the U.S. Embassy in Tokyo and a Business Information Report from Dunn & Bradstreet. The Dunn & Bradstreet report states that banks suspended transactions with KYK in September 1999 and that the firm was bankrupt. Based upon the information submitted by Gulf Bearing and the Dunn and Bradstreet report, we have used the same methodology we employed for the preliminary results to calculate a nonadverse dumping margin for KYK for these final results, i.e., we have applied the average margin for all firms involved in the review of BBs from Japan. Comment 5: Torrington argues that INA's response is deficient because INA did not report sales to its U.S. affiliate of balls that were further manufactured into linear motion devices. Citing United States v. Citroen, 223 U.S. 407, 56 L.Ed. 486 (1911) (Citroen), and The Carrington Co. v. United States, 61 CCPA 77, 81, C.A.D. 1126 (1974) (Carrington), Torrington argues that the subject balls imported by INA were still describable as BB parts at the time of importation and that the dutiable status of imported goods must be ascertained by an examination of the imported article in the condition in which it is imported. Torrington asserts that, because INA did not report the sales related to the balls at issue, the Department should apply facts available to determine the values and dumping margins for these sales. INA claims that it did not report the ball sales at issue because they were used to produce linear motion devices. Citing the Final Scope Ruling on the Request by Allergen Medical Optics dated June 18, 1992, INA asserts that balls are only considered to be parts of BBs subject to the order when they are employed as the rolling element of AFBs. Thus, according to INA, because linear motion devices are not subject to the antidumping duty orders under review, the balls imported for use in the linear motion devices are also not subject to the orders. INA asserts that the cases cited by Torrington are inapposite because they involved an unrelated issue, which was whether merchandise imported in one condition should be classified as if imported in another condition. INA also argues that the Department has consistently held that scope is governed by the narrative description of the merchandise contained in an order, not by the tariff classification. Finally, INA contends that, even if the imported balls used in the production of linear motion devices were subject merchandise, the Department had already excused INA from reporting sales of imports that were further manufactured in the United States. Department Position: INA's assertion that balls not ultimately used in BBs are outside the scope of the order is incorrect. The order applies to balls suitable for use in BBs unless the Department concludes in a scope determination that specific balls are used to manufacture some product other than the BBs subject to this order. Furthermore, the scope ruling to which INA cites is inapposite. In that ruling, we stated that "the answer to the question of whether the [merchandise in question] constitute[s] a 'part' of an AFB is dependent upon whether [they] are 'rolling elements' suitable for use in an AFB." See the Final Scope Ruling on the Request by Allergen Medical Optics dated June 18, 1992, at page 4. In addition, the ruling stated that the merchandise in question could not function as a ball in an AFB. Finally, it does not appear, nor did INA claim, that the balls INA used in linear motion devices are either identical or similar to the stainless steel balls at issue in the scope ruling. Thus, the scope ruling to which INA cites is not dispositive as to whether the balls which INA claims to further process into linear motion devices are within the scope of the order. If INA believes that its imports of such balls should be excluded from the order, it must request a scope ruling with regard to that merchandise. While the imports of the balls in question are subject to the orders of these reviews, we find that the use of facts available is not warranted in INA's case. In our letter to INA dated August 18, 1999, we advised INA that it did not need to respond to the further-manufacturing elements of our questionnaire. Furthermore, after the Germany-specific hearing for this review, we sent INA a supplemental questionnaire, requesting that INA revise its section A response regarding further-processed merchandise so that we could determine whether INA still qualified for this treatment. After reviewing INA's response of June 30, 2000, we have determined that it is appropriate to apply the special rule for merchandise with value added after importation (section 772(e) of the Act) with regard to INA because 1) the value added in the United States exceeds substantially the value of the subject merchandise and 2) there is a sufficient quantity of U.S. sales of subject merchandise that was not further processed in the United States upon which we can calculate a margin. In its supplemental questionnaire response, INA indicated that it had declared balls ultimately used in linear motion devices as within the scope of the order on BBs from Germany at the time of importation. We are satisfied that liquidation of entries of such balls was suspended and deposits were paid. Therefore, when we send out the liquidation instructions for such entries, entries of these balls will be liquidated at the applicable INA-specific rate. Comment 6: Torrington argues that the Department should apply adverse facts available to INA's CEP sales involving models purchased from third- country affiliates. Torrington contends that INA's response is deficient with respect to the activities of the third-country affiliates supplying the models, such as movement expenses, indirect selling expenses, pre-sale warehousing expenses, and inventory carrying costs. Torrington also asserts that the "proxy" data and "estimates" INA submitted in its supplemental questionnaire response are inadequate for antidumping reporting. Torrington argues that INA possessed accurate information and should have reported it. Torrington also argues that INA's claim that inventory carrying costs and indirect selling expenses which its third- country affiliates incurred have no impact on the margin analysis is incorrect. Torrington contends that such expenses are highly relevant to analyzing selling activities at the CEP level of trade and in determining whether INA qualifies for a CEP-offset adjustment in the calculation of normal value. Therefore, Torrington argues, the Department should apply adverse facts-available margins to INA's CEP sales of models purchased from third-country affiliates. INA contends that its sales records provide no means for identifying purchases from such affiliates and that it acted reasonably and to the best of its ability in providing the information the Department requested. INA argues that, because of this, the use of adverse facts available is not warranted. INA also asserts that the sales in question comprise a small portion of INA's total sales of subject merchandise during the POR. INA argues that using facts available on all transactions of models that are the same as those imported from third-country affiliates would be unsupported by substantial evidence on the record because it would result in the application of facts available to merchandise imported directly from Germany. INA suggests that the Department apply the adjustment to transactions equaling the quantities imported during the POR from third- country affiliates. INA agrees that application of facts available is necessary for warehousing expenses which the third-country affiliates incurred and suggests that the Department add a factor equal to the amount of U.S. pre- sale warehousing expense to third-country import transactions as the facts available. Department Position: In its section A response dated July 8, 1999, INA reported that it had made no U.S. sales through third-country affiliates. INA also did not report any sales made through third-country affiliates in its section C response dated August 23, 1999. Only after we inquired about the possibility of such sales did INA provide minimal information in its supplemental response dated October 29, 1999. In that supplemental response, INA reported movement and brokerage expenses for these transactions by estimating freight from INA to the third-country affiliates at rates applicable to shipments from INA to other European destinations. It then added these rates to the actual freight expenses from the third-country affiliate to its U.S. affiliate. See INA's supplemental response at 14-15. INA indicated that it did not have sufficient time to provide the foreign-expense data but it did not request that the Department provide additional time for the reporting of these expenses. Section 776(b) states that, if the administering authority finds that an interested party has failed to cooperate by not acting to the best of its ability to comply with a request for information from the administering authority, the administering authority may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise available. Such adverse inference may include reliance on information derived from the petition, a final determination in the investigation, any previous review, or any other information placed on the record. See 62 FR 27296 at 27340 (May 19, 1997). As a result of INA's failure to report certain U.S. transactions and the expenses associated with such transactions, we find that INA has not acted to the best of its ability. Therefore, we have determined to apply facts available consistent with section 776(b) of the Act. In light of the factors we consider in making an adverse facts-available determination in these proceedings, we have determined that making an adverse inference in applying facts available is appropriate. See, e.g., AFBs 6, 62 FR at 2088. First, INA has previously participated in reviews of the AFB orders, which indicates that it has experience with the Department's antidumping reporting requirements. Second, INA had the ability to obtain the necessary sales and expense information because the sales were made through an affiliated party. Finally, even if we were inclined to use the U.S. sales as reported, it is impossible to calculate a margin, given INA's failure to provide a full reporting of expenses related to these transactions. We also can not determine whether a CEP-offset adjustment is appropriate because INA did not provide the selling activities associated with the sales made through third-country affiliates. Therefore, we have applied partial adverse facts available by applying a facts available rate to INA's CEP sales of all models purchased from third- country affiliates. Because we are unable to identify third-country affiliate transactions from among all other CEP transactions, we have applied the partial adverse facts-available rates to all transactions involving sales of models which INA reported were sold by the third- country affiliate. In our application of the partial facts-available rates, we have selected the highest rates INA has ever received for BBs, 49.62 percent, and CRBs, 52.43 percent. The BB rate was a rate previously calculated for INA's sales of BBs for the 1995/1996 review. The CRB rate was previously calculated for INA's sale of CRBs in the LTFV investigation. Because we have calculated these rates for INA in previous segments, we consider these rates relevant for INA. Our analysis for corroborating these rates is based on proprietary information; please see INA's final results analysis memorandum dated July 28, 2000, for a description of our analysis. For SPBs, we did not select the highest rate calculated for INA in any segment, which is 28.62 percent, because we were not able to corroborate it for this review period and an alternative adverse rate is available. Accordingly, we selected 5.02 percent, the SPB rate we have calculated for SKF Germany for this review, which is not secondary information and does not require corroboration. Comment 7: Osaka Pump requests that the Department reconsider its preliminary decision to use facts available as the basis for its weighted- average dumping margin with respect to BBs. The respondent acknowledges that its submissions contained serious errors, including missing sales transactions, but asserts that the omissions were not intentional and that it did not try to withhold information. The respondent requests that the Department apply a lower facts-available rate for the final results. Department's Position: For the reasons detailed in the Preliminary Results, 65 FR at 18036, we affirm our decision to use adverse facts available for Osaka Pump with respect to BBs as the basis for the weighted- average dumping margin. As adverse facts available, we have continued to apply the highest rate we have calculated for companies under review for this segment of the proceeding. The respondent provided no argument on this issue. Specifically, we have applied as adverse facts available a rate we determined for Takeshita for this period to Osaka Pump's exports to the United States during the POR. As the selected rate is derived from information obtained in the course of this review, corroboration is unnecessary. Therefore, we have not made any changes in our analysis of Osaka Pump from the preliminary results of review, but, because the rate for Takeshita has changed, the facts-available rate for Osaka Pump has changed. 2. Revocation Comment 8: TIE argues the Department should revoke the orders as proposed in the preliminary results and as provided under 19 CFR 351.222(b)(3). It therefore argues that revocation should apply to all six factories that supplied bearings to TIE during the period which formed the basis for revocation. Torrington claims that TIE has not met the requirements under the amended 19 CFR 351.222(b)(1) that became effective November 1, 1999 (64 FR 51236, 51240 (September 26, 1999)). Torrington disagrees that any such revocation should apply to all six producers or suppliers that supplied TIE during the period that formed the basis for the revocation because, it asserts, non have met the revocation conditions set forth in 19 CFR 351.222(b)(3). TIE responds that, if a party requesting revocation is a non-producing exporter, the Department normally will revoke an order in part only with respect to the subject merchandise produced or supplied by those companies that supplied the exporters during the time period that formed the basis for the revocation. TIE asserts that, under the amended or the former regulations, it has met all necessary conditions and that revocation of the six producers or suppliers that supplied TIE during the period that formed the basis for the revocation is lawful and factually supported. Department's Position: The Department finds the issues moot for several reasons. First, for this review period, the Department has calculated a de minimis margin for TIE. Therefore, all covered entries will be liquidated without regard to dumping duties. Second, the U.S. International Trade Commission (ITC) determined in a sunset review pursuant to section 751(c) of the Act that revocation of the order on AFBs from Romania would not be likely to lead to continuation or recurrence of material injury to an industry in the United States within a reasonably foreseeable time. See Certain Bearings from China, France, Germany, Hungary, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom, Inv. Nos. AA-1921-143, et al 65 FR 39925 (June 28, 2000). Based on the ITC's sunset determination, revocation of the order on BBs from Romania became effective January 1, 2000. See 65 FR 42667 (July 11, 2000). Thus, a revocation decision in this review would not affect any entries after that date. Third, although TIE may have had exports during the interim period between the end of this review period and the effective date of revocation, (a) entries of such exports would have been made at a zero cash-deposit rate based on previous reviews, (b) the opportunity to request a review of those interim entries has passed without a review request, and (c) those entries are subject to automatic assessment under 19 CFR 351.212 (c) as a result of reason (b). Consequently, there are no entries which would be affected by a revocation decision in this review and, therefore, it is not necessary for the Department to address these revocation issues as to TIE. Comment 9: Torrington alleges that one of TIE's suppliers located in Alexandria was purchased by a Japanese producer, Koyo Seiko, during the POR and sales by this supplier after July 27, 1998, are through Koyo USA and reported by Koyo. Koyo is a market-economy producer that has been found to dump BBs produced in Japan in every review of the order on Japan. Therefore, according to Torrington, TIE did not report sales of subject merchandise at fair value for three consecutive years, as the regulation requires, and neither TIE nor the Department can predict future dumping or deliver an agreement for immediate reinstatement of that certain supplier. TIE reiterates that the order should be revoked for TIE's sales of subject merchandise produced by the supplier in Alexandria. TIE claims that it reported all sales from this supplier for the entire POR appropriately. Department's Position: As indicated in the response to Comment 8, this issue is moot. Comment 10: Torrington argues that the Department recently concluded a sunset review regarding the antidumping duty order on BBs from Romania in which the Department determined that revocation of the order would likely lead to recurrence of dumping due to a significant decline in imports since the imposition of the antidumping order, citing Ball Bearings from Romania; Final Results of Expedited Sunset Review, 64 FR 60313 (November 4, 1999). Thus, Torrington argues that the second condition for revocation has not been satisfied for any of the Romanian exporters or manufacturers, citing 19 CFR 351.222(b)(2). TIE responds that the sunset-review finding in this case has no relevance to revocation and does not support a finding that revocation of the order would likely lead to a recurrence of dumping. TIE believes that the objective and procedure of a sunset review are completely different from those in an administrative review. Department's Position: As indicated in response to Comment 8, this issue is moot. Comment 11: Torrington argues that the Department should rescind its preliminary determination to revoke the antidumping duty order in part with respect to BBs produced and exported by SNFA France. Torrington contends that SNFA France has not satisfied the second requirement for revocation which states that the Department may not revoke an order in part as to an exporter or producer covered by the order unless the Department determines that it is not likely that those persons will in the future sell the subject merchandise as less than normal value, citing 19 CFR 351.222(b)(2). The petitioner argues further that this criterion has not been satisfied based on the recent sunset review regarding the antidumping duty order on BBs from France. Citing Ball Bearings from France; Final Results of Expedited Sunset Review, 64 FR 60321 (November 4, 1999), Torrington asserts that the Department determined that revocation of the order would likely lead to recurrence of dumping by SNFA France and that the dumping margins calculated in the LTFV investigation reflect the level of dumping likely to prevail without the discipline of the order. Therefore, Torrington claims that the Department should determine that the second criterion of section 19 CFR 351.222(b)(2) has not been satisfied. SNFA France argues that the Department should grant its request for revocation with respect to BBs based on the Department's determination that SNFA France met all the requirements for revocation. The respondent contends that no provision specifies that the five-year sunset review supersedes the regulatory principles of company-specific revocation in an administrative review. Department's Position: The Act states that the Department "may revoke, in whole or in part" an antidumping duty order upon completion of an administrative review. See section 751(d) of the Act. Although Congress has not specified the procedures that the Department must follow to revoke an order, the Department has developed a procedure for revocation that is set forth in 19 CFR 351.222. Under subsection 19 CFR 351.222(b)(2), the Department may revoke an antidumping duty order in part if it concludes that (1) the company in question has sold the subject merchandise at not less than normal value for a period of at least three consecutive years, (2) it is not likely that the company will in the future sell the subject merchandise at less than normal value, and (3) the company has agreed to immediate reinstatement of the order if the Department concludes that the company, subsequent to the revocation, sold the subject merchandise at less than normal value. A request for revocation of an order in part must be accompanied by three elements. The company requesting revocation must do so in writing and submit the following statements with the request: (1) the company's certification that it sold the subject merchandise at not less than normal value during the current review period and that, in the future, it will not sell at less than normal value; (2) the company's certification that, during each of the three years forming the basis of the request, it sold the subject merchandise to the United States in commercial quantities; (3) an agreement to reinstatement of the order if the Department concludes that the company, subsequent to revocation, has sold the subject merchandise at less than normal value. See 19 CFR 351.222(e)(1). SNFA France's request meets all of the criteria under subsection 19 CFR 351.222(e)(1). Thus, our analysis turns to whether this company has satisfied the criteria of subsection 351.222(b)(2). The Department first examines whether the requesting company sold the subject merchandise at not less than normal value to the United States in commercial quantities for three consecutive reviews. It then examines whether it is likely that the company would in the future sell the subject merchandise at less than normal value. Our final margin calculations show that SNFA France sold BBs at not less than normal value during the current review period. Furthermore, SNFA France sold the subject merchandise at not less than normal value in the two previous consecutive administrative review periods. See AFBs 8, 63 FR at 33321; and AFBs 9, 64 FR at 35591. Thus, SNFA France had zero or de minimis dumping margins for three consecutive reviews. Moreover, as we found in our Preliminary Results, SNFA has shipped subject merchandise in commercial quantities during these review periods. Thus, in the absence of any other evidence on likelihood, the Department finds that dumping is not likely to resume. As for Torrington's argument that the Department should not revoke the order in part with respect to SNFA France based on the sunset review in which the Department determined that revocation of the antidumping duty orders on certain bearings from France would lead to continuation or recurrence of dumping, we disagree. The Department makes its determination in a sunset review on an order-wide basis and thus it includes consideration of other companies that have continued to dump under the order. See SAA at 889-90. See also Policy Bulletin 98:3 - Policies Regarding the Conduct of Five-year ("Sunset") Reviews of Antidumping and Countervailing Duty Orders; Policy Bulletin; Request for Comments, 63 FR 18871, 18872 (April 16, 1998) (Policy Bulletin). The Department does not calculate new dumping margins in a sunset review and, as detailed in the Policy Bulletin, at 18873, we rely on the margin from the initial investigation unless there is good cause to use a more recently calculated margin for a particular company. In contrast, the Department considers company-specific revocations based on the elements identified at 19 CFR 351.222. The Department may revoke an order in part when the producer or importer has sold the subject merchandise at not less than fair value for three consecutive years, which creates a presumption of no future dumping as to that company. See 19 CFR 351.222(b)(2). Thus, where there is no evidence to the contrary, the Department will normally determine that revocation is warranted. See, e.g., Furfuryl Alcohol From the Republic of South Africa; Preliminary Results of Antidumping Duty Administrative Review and Intent to Revoke Order, 64 FR 10983, 10894 (March 8, 1999), Titanium Sponge From the Russian Federation; Preliminary Results of Antidumping Duty Administrative Review and Partial Revocation, 63 FR 47474, 47475 (September 8, 1998), and Steel Wire Rope From the Republic of Korea; Final Results of Antidumping Duty Administrative Review and Revocation in Part of Antidumping Duty Order, 63 FR 17986, 17988 (April 13, 1998). SNFA France has established to our satisfaction that it has sold subject merchandise in commercial quantities at not less than normal value for three consecutive years. Absent evidence to the contrary, the Department presumes that three years of de minimis or zero margins is indicative of SNFA France's future behavior. See, e.g., Professional Electric Cutting Tools From Japan; Final Results of the Fifth Antidumping Duty Administrative Review and Revocation of the Antidumping Duty Order, 64 FR 71411, 71419 (December 21, 1999); see also, Polyvinyl Alcohol From Taiwan; Preliminary Results of the Third Antidumping Administrative Review and Intent Not to Revoke Order, 65 FR 36900 (June 6, 2000). The general findings of sunset reviews which are made on an order-wide basis, in the absence of additional information, are not sufficient to overcome the presumption that a specific company which has made sales in commercial quantities at not less than normal value for three consecutive years will not dump in the future. Therefore, based on our findings and in accordance with 19 CFR 351.222(b)(3), we have revoked the antidumping duty order covering BBs from France as it pertains to BBs produced and exported by SNFA. We will terminate the suspension of liquidation for any BBs from France that were produced and exported by SNFA France and entered, or withdrawn from warehouse, for consumption on or after May 1, 1999, and we will instruct the Customs Service to refund any cash deposits for such entries made after that date. Comment 12: Torrington argues that the Department should modify its determination in the preliminary results to indicate that two of the applicable criteria for revocation of Somecat's imports of BBs have not been satisfied. Torrington concurs with the Department's conclusion that the existence of dumping by Somecat in this POR, as indicated in the preliminary results constitutes grounds to disallow revocation of the order. However, Torrington suggests that the order also should not be revoked based on the reason set forth in the Department's recent sunset review regarding the antidumping duty order on BBs from Italy. In that determination, the Department concluded that revocation of the order would likely lead to continuation of dumping and that the dumping margins calculated in the LTFV investigation reflect the behavior of exporters without the discipline of the order. Therefore, Torrington asserts, the Department should indicate that the second criterion of 19 CFR 351.222(b)(2) has not been satisfied. Somecat argues that it will satisfy the three-year criterion for revocation in the final results. Somecat claims that, once the Department corrects certain clerical errors in its calculations of Somecat's margin, Somecat will have a de minimis dumping margin and will meet the final condition of revocation. Somecat alleges further that the three-year, company-specific provision for revocation should not be overridden by the Department's determination in the five-year sunset review as Torrington suggests. Department's Position: We have corrected the clerical error Somecat identified in its case brief. As a result of the correction, Somecat's weighted-average margin for the POR is de minimis. Therefore, Somecat now satisfies the final requirement for revocation. For the reasons discussed in response to Comment 11 with respect to SNFA France, we do not accept Torrington's arguments concerning the relevance of the sunset review to this revocation determination and we agree with the respondent that it meets all of the criteria for revocation under 19 CFR 351.222(e)(1). Therefore, based on our findings and in accordance with 19 CFR 351.222(b)(3) of the regulations, we are revoking the antidumping duty order covering BBs from Italy as it pertains to BBs produced and exported by Somecat. We will terminate the suspension of liquidation for any BBs from Italy that were produced and exported by Somecat entered, or withdrawn from warehouse, for consumption on or after May 1, 1999, and will instruct the Customs Service to refund any cash deposits for such entries made after that date. Export Price/CEP 3.A. CEP Offset Comment 13: Torrington argues that claims by Barden and the SKF companies for a CEP-offset adjustment to normal value are not supported by the record. According to Torrington, the annual reports of the FAG Group and SKF AB (the parent companies of Barden and the SKF companies, respectively) indicate that they engage in numerous activities that generate benefits for Barden U.S. and SKF USA, respectively. Torrington asserts that these activities are selling activities for purposes of a level-of-trade analysis since they materially benefit the U.S. affiliates in their efforts to thrive in the United States. Specifically, according to Torrington, the FAG Group engages in activities such as strategic and economic planning, market research, research and development, technical programs, engineering services, and other activities that benefit Barden U.S. directly. Therefore Torrington argues that, since the claims for a CEP offset are not supported by information on the record, the Department should disallow the claims by Barden and the SKF companies for a CEP offset. Barden argues that Torrington misunderstands how the level-of-trade selling-function analysis works in a CEP-sale situation. According to Barden, the Department defines the level of trade of such sales based on the CEP after adjustments required under section 772(d) of the Act. In this regard, according to Barden, the Department compares Barden U.K.'s selling functions to Barden U.S. versus Barden U.K.'s selling functions in selling to its unaffiliated home-market customers. Barden argues that the selling functions provided by Barden U.S. to its unaffiliated U.S. customers are not relevant to a level-of-trade analysis. Further, Barden argues that Torrington's argument is premised on the unfounded proposition that sales support from the FAG Group benefits selling activities in the U.S. market but not in the home market. Barden asserts that, just as these Group selling activities do not directly benefit unaffiliated home-market customers, neither do they directly benefit unaffiliated U.S. customers. Thus, according to Barden, Torrington is attempting to compare apples to oranges by comparing the impact of Group selling functions on a Group affiliate (Barden U.S.) to the impact of these functions on unaffiliated U.K. customers. Barden argues further that, by definition, "Group" sales support suggests that these efforts benefit all members of the Group, not just a particular member of the Group. For these reasons, Barden requests that the Department reject Torrington's arguments regarding the level-of- trade analysis and the CEP offset. The SKF companies assert that the Department should continue to allow the CEP offset for their companies because the SKF AB annual report did not contain any information to contradict the responses to the questionnaire submitted by the companies or the information obtained by the Department at verification in Sweden. The SKF companies note that this information revealed a demarcation between management decisions, which were made at the Group level, and sales activities, which were performed at the local level by the individual SKF sales organizations. They note further that the findings of the SKF Sweden Verification Report reflect this demarcation. In addition, the SKF companies argue that the Department acted properly when it granted the CEP offset based on a comparison of selling activities at the level of the CEP transaction from the foreign producer to SKF USA versus those in the respective home markets. Department's Position: We disagree with Torrington that a CEP offset for these companies was inappropriate. Section 773(a)(7)(B) of the Act provides for an adjustment to normal value if normal value is established at a level of trade that is different from and more remote than the level of trade of the CEP sale and the information on the record does not provide a basis for determining a level-of-trade adjustment. Based on our analysis of all identified market-specific selling functions on the record, we have determined that the differences in selling functions between the CEP and normal-value levels are substantial and support a finding of different levels of trade. Thus, normal value for Barden and for the SKF companies was established at levels different from, and more remote from the factory than, the CEP level of trade. In addition, because there are no home-market levels of trade that are equivalent to the CEP level, we are unable to quantify a level-of-trade adjustment based on home- market sales prices. Therefore, we made a CEP-offset adjustment to normal value, based on our analyses of the information provided by Barden and the SKF companies. We are unpersuaded by Torrington's argument that we should deny these companies a CEP offset merely because their parent companies have established programs to improve global sales and profitability of their Group companies. While the information in the annual reports of the FAG Group and SKF AB appears to demonstrate that generic, cross-cutting administrative assistance is provided at a global management level and benefit operations of the entire organization as a whole, we have no basis to conclude that these activities are linked to market-specific selling activities and are, therefore, relevant to our CEP level-of-trade analysis. The record demonstrates that, for the most part, such market- specific selling functions per se are performed at the local level in each market. There is little indication that the upper managements of the parent companies provide any direct assistance which is related to either the sales made to, or by, U.S.-based affiliates. If this issue had been raised earlier in these current reviews, the Department could have requested further clarification from respondents with respect to Torrington's concerns. In future administrative reviews, we intend to explore the potential sales functions performed by parent firms on behalf of their group affiliates more closely. In conclusion, because normal-value sales are established at a level of trade more remote from the factory than those at the CEP level of trade and because the record does not provide a basis for a level-of-trade adjustment, the Department finds that it is appropriate to make a CEP offset to normal value for these firms pursuant to section 773(a)(7)(B) of the Act and 19 CFR 351.412. Comment 14: Torrington argues that, if the Department rejects its argument to apply facts-available margins to INA's CEP resales of models which its U.S. affiliate obtained from third-country affiliates (see Comment 6 above), it should alternatively disallow level-of-trade adjustments or CEP offsets in comparisons associated with such models. Torrington contends that such sales result in two sets of selling activities at the CEP level of trade, those of INA-Germany and those of the third-country affiliate. Torrington claims that the Department can not make proper comparisons or adjustments for these CEP resales of models since INA has not adequately reported the CEP selling functions associated with the activities performed by INA's third-country affiliates. INA disagrees with the petitioner's argument that the Department should disallow level-of-trade adjustments or CEP offsets in comparisons associated with models which the U.S. affiliate obtained from third- country affiliates. INA claims that both the sales by INA to a third- country affiliate and the sales by a third-country affiliate to the U.S. affiliate are at a different and less-advanced stage of distribution than INA's sales in Germany. Therefore, INA maintains, the Department should not change the application of the CEP offset to sales of models which its U.S. affiliate imported from third countries. INA claims further that, if the Department disallows the CEP offsets in comparisons associated with models its U.S. affiliate obtained from third- country affiliates, it should only make adjustments to the particular quantity of each model that was imported from a third-country affiliate during the POR. Department's Position: We have decided to apply facts available to the CEP resales of models which the U.S. affiliate obtained from third-country affiliates (as set forth in our response to Comment 6). Consequently, Torrington's arguments on this issue are moot and we are not addressing the appropriateness of the party's comments. 3.B. CEP Profit Comment 15: NTN Japan argues that the Department should have excluded export-price sales when it calculated CEP profit. According to NTN Japan, section 772(f) of the Act and the SAA at 824 direct that CEP profit is to be calculated on revenues and expenses incurred on only the sales used to calculated normal value and CEP sales. Torrington contends that the Department's approach is consistent with its policy bulletin regarding the issue and that the Department has rejected this argument in prior reviews. Department's Position: It is our practice to include export-price sales in the calculation of CEP profit. See, e.g., AFBs 8, 63 FR at 33350, AFBs 9, 64 FR at 35622, and Certain Fresh Cut Flowers From Colombia; Final Results and Partial Rescission of Antidumping Duty Administrative Review, 62 FR 53287, 53295, (October 14, 1997)(Flowers 9). In addition, our analysis in these reviews is consistent with Policy Bulletin 97.1 (September 4, 1997). The basis for total actual profit is the same as the basis for total expenses under section 772(f)(2)(C) of the Act. The first alternative under this section states that, for purposes of determining profit, the term "total expenses" refers to all expenses incurred with respect to the subject merchandise sold in the United States (as well as the foreign like product sold in the exporting country). Thus, where the respondent makes both export-price and CEP sales to the United States, sales of the subject merchandise would encompass all such transactions. See Policy Bulletin 97.1 (September 4, 1997). Therefore, because NTN Japan had export-price sales, we have included these sales in the calculation of CEP profit. Comment 16: NTN Japan argues that the Department should calculate CEP profit on a level-of-trade-specific basis. NTN Japan claims that selling expenses alone do not entirely explain the differences in prices between levels of trade and that, to account fully for the differences in prices, profit levels must be taken into consideration. NTN Japan also contends that there is a statutory preference that profit be calculated on the narrowest possible basis and that, by not calculating CEP profit on a level-of-trade-specific basis, the Department ignores that preference. NTN Japan argues further that, because CV profit is calculated on a level-of- trade-specific basis, CEP profit should also be calculated on that basis. Torrington contends that the Department has rejected this argument in prior reviews and that NTN Japan does not explain why the Department should alter its decision. Department's Position: It is not our practice to calculate CEP profit for different levels of trade. See, e.g., AFBs 9, 64 FR at 35621, and 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; Final Results of Antidumping Duty Administrative Reviews, 63 FR 2557, 2583 (January 15, 1998). We believe that NTN Japan's reliance on the term "narrowest" as used in sections 772(f)(2)(C)(ii) and (iii) of the Act is misplaced. While the statute uses the term "narrowest" in describing the second and third alternative methods, those methods are based upon financial reports. For NTN Japan we used the first alternative method since the company provided the necessary data (i.e., U.S. and home-market sales information as well as CV and COP data for the subject merchandise and the foreign like product, respectively). This is consistent with section 772(f)(2)(C) of the Act and the SAA at 824-825. Moreover, regardless of the basis for the CEP-profit calculation, neither the statute nor the SAA suggests that we calculate CEP profit on a basis more specific than subject merchandise and foreign like product. See Toyota Motor Sales, USA v. United States, 15 F.Supp.2d 872, 886 (CIT 1998). Finally, the CIT has upheld our practice in NTN Bearing Corp. of America, et al. v. United States, Slip Op. 00-64 (CIT June 5, 2000), (2000 Ct. Intl. Trade LEXIS 67 at 79). Thus, we have not adopted NTN Japan's suggestion. Comment 17: NPBS claims that the Department distorted the calculation of CEP profit by under-weighting profit on home-market sales. NPBS explains that the Department permitted reporting of a sample of sales in the U.S. and home markets and calculated the total expenses and the total profit in each market to approximate the total profit and loss on all sales for a full year. The Department did this on home-market sales, NPBS continues, by totaling expenses and revenues based on comparison families as opposed to the entire class or kind of merchandise sold in the home market. NPBS argues that, since its U.S. sales are more profitable than its home-market sales, this method has led to an overstatement of the profit ratio. NPBS observes that it is the Department's standard policy to base the CEP- profit calculation on all U.S. and home-market sales of the like product. Citing Color Picture Tubes From Japan; Final Results of Antidumping Administrative Review, 62 FR 34201, 34212 (June 25, 1997), NPBS argues that the calculation of CEP profit should include all home-market sales of the foreign like product during the POR. NPBS argues that section 772(f)(2)(C) of the Act supports this profit calculation because it states that the first preference is for the calculation to be based on "the subject merchandise sold in the United States and the foreign like product in the exporting country..." Thus, NPBS argues, section 772(f)(2)(C) of the Act requires the Department to calculate home-market expenses and revenues based on the ratio of the total revenues for the class of merchandise, which NPBS provided in its section A questionnaire response, to the home-market revenues for the comparison families and sample months. NPBS believes that the Department should correct the calculation of the CEP-profit ratio to include all sales of the like product for these final results. Torrington disagrees with NPBS by stating that the Department followed its established practice for calculating CEP profit. It asserts that the Department properly applied a factor of two for totaling home-market sample sales to account for a full year of data based on the number of months sampled in the home market. Citing the SAA at 824-825, Torrington comments that the profit calculation looks to subject merchandise and the foreign like product and that the foreign like product is family merchandise in the AFB proceedings. Torrington continues that family merchandise is defined by reference to reportable U.S. sales and physical characteristics thereof. It argues that, under this method, it would be inappropriate to look to aggregate information in Section A of the questionnaire response if that information went beyond the relevant families in this review. Since NPBS has not demonstrated that the Department's methodology distorted the CEP-profit calculation, Torrington concludes that there is no reason for the Department to alter its methodology. Department's Position: The first alternative under section 772(f)(2)(C)(i) of the Act for the purpose of determining profit refers to all 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..." However, we did not request that the respondents report all of their home- market sales of the foreign like product partly in order to reduce the reporting burden on respondents. Furthermore, since we have not requested expense data which corresponds to the sales totals respondents reported in response to section A of our questionnaire, we have used the second alternative under section 772(f)(2)(C)(ii) of the Act. That section indicates that we are to use the expenses incurred with respect to the narrowest category of merchandise in the United States and exporting country which includes the subject merchandise for the purpose of determining profit. Therefore, because we did not request information on all home-market sales and expenses, the category of merchandise we have used is the foreign like product reported in the respondents' home-market databases. 3.C. Other Expenses Comment 18: Torrington argues that, with respect to certain sales made through one of FAG Italy's U.S. affiliates, the Department should deduct indirect selling expenses and commissions paid by FAG Italy's affiliate to related warehouses (which serve as selling agents for FAG Italy) to calculate CEP. Torrington comments that, with regard to any sales made by related warehouses in the United States, FAG Italy's U.S. affiliate pays a sales commission and a warehouse commission. Torrington asserts that FAG Italy has not demonstrated that the commissions cover all expenses which its U.S. affiliate incurred. Torrington alleges further that the Department should apply facts-available values in its calculation of U.S. price on sales made through the affiliate in question by using FAG Italy's indirect-selling-expense factor applicable to other U.S. sales. FAG Italy argues that it never claimed that the reported commissions cover "all" expenses which its U.S. affiliate incurred. FAG Italy explains that its U.S. affiliate pays commissions to a related warehouse only on sales for which the warehouse incurred the selling expenses. Thus, according to FAG Italy, for those particular sales on which it paid commissions, an adjustment for both the commission and indirect selling expenses would constitute double- counting. FAG Italy argues that it is not advocating the use of either the sales commission or the indirect selling expense amount; rather, it is noting that adjusting for both the sales commission when applicable and indirect selling expenses would be contrary to the Department's established policy, citing Oil Country Tubular Goods From Mexico; Final Results of Antidumping Duty Review, 64 FR 13962, 13968 (March 23, 1999). FAG Italy argues that there is no need for a commission deduction if the Department accounts for expenses represented by the commission by deducting selling expenses. Department's Position: We have not deducted both commissions and indirect selling expenses in calculating the CEP for FAG Italy's U.S. sales. For the preliminary results we did not deduct from the price used to establish the CEP the warehousing commissions that FAG Italy's U.S. affiliate paid to affiliated warehousing companies or sales commissions it paid to its affiliated warehousing companies. See the FAG Italy March 27, 2000, analysis memorandum at page 5. Rather, we deducted the actual expenses, i.e., all indirect selling expenses that FAG Italy's U.S. affiliates incurred on the sales. We followed this methodology because we generally rely on actual expenses rather than intra-company transfers. See Flowers 9, 62 FR at 53294, and AFBs 8, 63 FR at 33345. Generally, we consider affiliated-party commissions to be intra-company transfers of funds to compensate an affiliate for actual expenses incurred in completing the sale to unaffiliated customers and we do not deduct them from CEP. Comment 19: Torrington argues that the Department should adjust FAG Germany's U.S. indirect selling expenses to include appropriate amounts incurred by an affiliated German corporation, FAG Interamericana GmbH (Interamericanca), which operates within a U.S. foreign trade zone (FTZ). Specifically, Torrington claims that Interamericana sold scope products from the FTZ to FAG's U.S. affiliate during the POR. Torrington argues that, since FAG Germany did not provide information which the Department requested with respect to Interamericana, as facts available, the Department should require FAG Germany to provide total POR expenses of Interamericana. According to Torrington, the Department could then allocate the total amount over all U.S. sales unless FAG Germany can point to data, already on the record, justifying lower amounts. Torrington contends that these expenses are part of total U.S. selling expense incurred in selling in the United States and therefore FAG Germany should have reported them. In addition, Torrington argues that the Department should, as facts available, restate other expenses relating to U.S. sales of products that flowed through Interamericana because such expenses are incurred in re- selling goods in the United States. According to the petitioner, these expenses include inventory carrying costs, pre-sale warehouse expense, and expenses for U.S. inland freight and insurance incurred by Interamericana. FAG Germany argues that Interamericana did not make sales of subject merchandise to unrelated U.S. customers during the review period. In addition, FAG Germany argues that its U.S. affiliate did not have sales to unrelated customers of subject merchandise it purchased from Interamericana. Therefore, according to FAG Germany, expenses incurred by Interamericana are in no way related to reported U.S. sales. Further, according to the respondent, to include any such expenses (e.g., Interamericana selling expenses, inventory carrying costs, warehousing cost, etc.) would be inappropriate in the analysis of its U.S. sales. FAG Germany contends that, if it were appropriate to include such expenses in the calculation of U.S. selling expenses, it would also then be necessary to adjust the allocation of selling expenses to reflect the value of Interamericana sales during the POR. FAG Germany argues that it responded accurately and completely to all relevant questions regarding Interamericana and, therefore, the petitioner's call for facts available has no merit. Department's Position: In its July 30, 1999, submission FAG Germany states that on rare occasions Interamericana may sell to FAG U.S. to fill emergency supply needs. However, on page 4 of its July 30, 1999, submission, FAG Germany makes clear that Interamericana made no sales during the POR, either directly or through FAG's U.S. affiliate, of scope merchandise to unrelated customers in the United States during the POR. Consequently, since Interamericana was not involved in sales of subject merchandise to the United States, we find no reason to request FAG Germany to provide total POR expenses of Interamericana as Torrington suggests. Therefore, we find that an adjustment to FAG's U.S. indirect selling expenses is inappropriate as it relates to Interamericana. 4. Discounts and Rebates Comment 20: Torrington argues that the Department should disallow all home-market rebates claimed by Somecat that were paid on sales prior to the dates on which the rebate terms were set. Torrington alleges that Somecat provided a misleading description of its rebate program and recorded rebates that had not yet been established at the time of sale. Torrington also claims that Somecat misrepresented its rebate program by reporting that the parties contemplated payment of rebates if certain sales quantities were met. However, Torrington continues, the Department's verification report states that ". . . rebates are paid at the end of the year but generally not tied to a specific volume" and that "[the Department] found no reference to volume requirements in the contracts." Therefore, Torrington asserts, the Department should reject the adjustments on the basis of inaccurate reporting of information and failure to demonstrate entitlement to any adjustment decreasing normal value. Torrington argues further that the Department should disallow all home- market rebates Somecat claimed on sales in 1999 which are not covered by actual agreements on record. It observes that, for sales during 1999, Somecat reported anticipated rebates based upon the customers' current level of purchases and projected purchases. Because Somecat has no formal rebate policy and determines its rebates on a customer-specific basis, Torrington argues, Somecat has not proven to the Department that it grants rebates on the basis of quantities sold. Thus, Torrington asserts, the Department should not allow for estimations of sales in 1999. The respondent argues that, as verified by the Department, Somecat reported its home-market rebates correctly and the Department should accept them as it has in all prior reviews. Somecat states that there is no legal requirement that all rebate terms be in writing at the time of sale and that the Department permits an adjustment for rebates when the terms of the rebate "are understood from past dealings of the parties," citing the Department's, antidumping manual, Chap. 8, at 11 (January 22, 1997). Somecat contends that the rebate programs are part of its standard business practice. It states that it reported the rebates on a transaction- specific basis and as a fixed and constant percentage of the sales price which reflect two requirements of the Department's long-standing practice for accepting adjustments to normal value, referring to AFBs 5, 61 FR at 66472. Somecat also comments that the Department has never published any decision requiring a written rebate agreement at the time of sale. Department's Position: Historically, the Department has allowed respondents to claim rebates without providing written agreements. The Department set forth its rebate policy in AFBs 2, AFBs 3, AFBs 4, and AFBs 5 where it stated that it would make adjustments for reported home-market discounts, rebates, and price adjustments if (a) they were reported on a transaction-specific basis, or (b) they were granted as a fixed and constant percentage of sales price on all transactions for which they were reported, as in the case with a fixed-percentage rebate program. See AFBs 5, 61 FR at 66498. Somecat reported its rebates as a fixed and constant percentage of the sales price. Moreover, while our verification report for this review states that there was no reference to volumes in Somecat's contracts, we found no discrepancies with Somecat's calculation methodology and were able to tie the sale price to the percentage of rebate reported on a transaction-specific basis. Therefore, the Department has included all reported rebates as price adjustments to normal value in its calculation of Somecat's margin. Comment 21: Torrington argues that the Department should disallow INA's estimated home-market rebates for 1999 since INA did not provide adequate support to demonstrate its entitlement to these rebate adjustments. Specifically, Torrington claims that INA's decision that a customer was eligible for rebates in 1999 was based solely on a year-to-year comparison of sales by that customer and that INA did not take other information into consideration, such as whether the customer earned rebates previously. Thus, Torrington continues, where the rebate program terms changed in 1999, such a comparison (i.e., year-to-year) would result in an invalid apples-to-oranges comparison. For these reasons, Torrington argues that the Department should reject INA's rebate claims. INA claims that it used a reasonable method to project total qualifying sales for 1999 and properly applied the projected sales amount per customer to the rebate terms applicable to each customer. INA asserts that it properly compared sales to the customer in the first half of the calendar year 1998 with sales to the customer in the first half of calendar year 1999; then it applied that ratio to total 1998 sales in order to project 1999 sales and the terms of the new 1999 rebate agreement with particular customers to determine the rebate amount. Department's Position: INA reported actual rebates granted in 1999 to individual home-market customers during the POR in cases where the agreed period of time expired prior to July 1, 1999. For new 1999 rebate programs, where the agreed period of time expired after June 30, 1999, INA estimated whether the customer would be eligible based on a comparison of sales in the first half of 1998 and sales in the first half of 1999 (see page B-23 of INA's Section B response, August 23, 1999). Furthermore, INA's rebate chart, in exhibit B-7 of its August 23, 1999, submission, outlined the customers which received rebates for 1998/99 purchases, the time period for which INA and its customers agreed on the rebate percentages, and the amount of the rebate granted. We have determined, based on INA's response, that INA did take entitlement into consideration when granting a rebate. Therefore, we find no reason to disallow INA's claimed home-market rebates. Comment 22: Torrington argues that INA did not calculate rebate rates for U.S. sales on a customer-by-customer basis properly and claims that the Department should re-calculate INA's rebate rate by using an average rate from INA's U.S. sales data. Torrington asserts further that, if the Department can not determine from the record which customers are involved, then the Department should resort to facts available for these transactions. INA argues that its rebate methodology uses actual rates from each customer-specific rebate agreement, as indicated by its reference to the specific rebate program in its original questionnaire response. INA claims that it had one customer which had a 1999 rebate program but not a rebate program in the prior year. For this customer, INA applied the lowest rate applicable under the customer's rebate agreement. Therefore, INA asserts that it used a reasonable methodology which was non-distortive and no adjustment to its claimed U.S. rebates is warranted. Department's Position: For these final results and consistent with our past practice, we have accepted claims for rebates as direct adjustments to price if we determined that the respondent, in reporting these adjustments, acted to the best of its ability and that its reporting methodology was not unreasonably distortive. See AFBs 8, 63 FR at 33327. See also Torrington Company v. United States, Slip Op. 00-44, at 14-16 (CIT April 19, 2000). INA provided copies of all of its rebate-program agreements during the POR, (July 8, 1999, at Exhibit A-14). Furthermore, INA also provided a comprehensive chart detailing all of its U.S. customers and the applicable rebate rate during the POR (August 23, 1999, at Exhibit C-4). In our supplemental questionnaire, we requested that INA confirm that its rebates were reported on a customer-specific basis. In its October 29, 1999, supplemental response at 42, INA confirmed that it allocated the reported rebates on a customer-specific basis over the sales during the POR. In all but one instance, INA based this allocation on actual per-customer rebates for the first half of the year. INA explained that for one customer there was an agreement for the 1999 period but no agreement during the 1998 period. Therefore, to estimate the rebate expense for the 1998 period, INA used the lowest rate that could be applicable to that customer during the 1999 period. Therefore, we have accepted the reported U.S. rebates because INA demonstrated that it allocated the U.S. rebates on a customer-specific basis, except in the case of one customer. For that single customer, consistent with our decision in AFBs 8, 63 FR at 33327, we find that INA's rebate rate is not distortive. Comment 23: Torrington asserts that the Department should reject all adjustments SKF Germany claimed as home-market billing adjustment two, which applies to multiple transactions involving the same customer, that result in lowering home-market prices. Torrington contends that it is unclear why SKF Germany can not identify the specific transactions to which these adjustments apply. Torrington argues further that SKF Germany's reporting method is fundamentally flawed as it double-counts adjustments, since amounts involved in "partial"-year reporting are used again for "full"-year reporting in the next annual review. SKF Germany responds that its reporting of billing adjustment two is non- distortive and consistent with how it incurs this expense and, furthermore, that it constitutes a reasonable allocation under U.S. law. SKF Germany asserts further that the Department has accepted this adjustment in the last four reviews, as well as verified it in the 1996/1997 review where it found that transaction-by-transaction reporting is simply not possible because the involved adjustments relate to multiple transactions and, therefore, could not have been reported more specifically. SKF Germany contends therefore, that, disallowing downward billing-adjustment-two amounts would be unwarranted in this case, since it has reported this adjustment in a manner consistent with its business records and the Department found them to be allocated reasonably when it examined these adjustments at verification of a previous review. Department's Position: We have accepted post-sale billing adjustments as direct adjustments to price if we determined that the respondent, in reporting these adjustments, acted to the best of its ability and that its reporting methodology was not unreasonably distortive. While we prefer that respondents report these adjustments on a transaction-specific basis (or, where a single adjustment was granted for a group of sales, as a fixed and constant percentage of the value of those sales), we recognize that this is not always feasible, particularly given the extremely large volume of transactions involved in these AFB reviews and the time constraints imposed by our reporting requirements necessitated by statutory deadlines. The billing adjustments SKF Germany reported were part of credit or debit notes, issued to the customer, that related to multiple invoices, products, or invoice lines, which could not be tied to a single specific transaction. In these cases, the most feasible reporting methodology that SKF Germany could use was a customer-specific allocation, which is not unreasonably inaccurate or distortive. It is inappropriate to reject allocations that are not unreasonably distortive where a fully cooperating respondent is unable to report the information in a more specific manner. Verification in the 1996/1997 review was an opportunity to determine whether billing adjustment two represented a reasonable approximation of SKF Germany's experience in granting this adjustment. Our conclusion was that there was no reason to believe that the actual data would differ significantly. See SKF Germany's Sales Verification Report, public version, for the 1996-1997 administrative review, dated December 12, 1997, at pages 6-7. In the instant review, there is no record evidence to indicate that the bearings included in SKF Germany's current allocations vary significantly, either in terms of value, physical characteristics, or the manner in which they are sold. For this reason, we find that this methodology is not unreasonably distortive and have accepted it for the final results. Comment 24: Torrington asserts that, during verification, the Department discovered that one of FAG Germany's customers calculated such a discount net of billing adjustment, but in its response FAG Germany reported the discount calculated on the gross unit price from which the billing adjustment had not been subtracted. Torrington acknowledges that the Department made further inquiries and found no other discrepancies, but it contends that some upward adjustment to FAG Germany's home-market prices is necessary. According to Torrington, sale checks at verifications are limited to a small sampling of sales, such that each sale weighs heavily in assessing overall reporting accuracy. Torrington contends that, in accordance with statistical theory, the pattern reflected within the sample should be seen as reflecting the same pattern within the entire universe of sales and that the Department should project that pattern into the remaining body of sales. Therefore, Torrington requests that the Department develop a factor based on the sample and apply it to all home-market sales. Alternatively, Torrington suggests, the Department could assume that all sales to the home-market customer in question involved the same erroneous calculation and then adjust such sales accordingly. FAG Germany argues that the Department examined and verified that it reported early-payment discounts correctly. According to FAG Germany, after reviewing additional home-market sales with early-payment discounts, the Department found that the reported early-payment discount corresponded to the actual early-payment discount for all transactions selected. FAG Germany argues further that Torrington's suggestion that this single inadvertence is indicative of widespread inaccuracies is directly contradicted by the Department's verification results. Further, FAG Germany asserts that Torrington provides no factual or legal basis for extrapolating this isolated case to all home-market sales with reported early-payment discounts. Therefore, FAG Germany requests that the Department reject Torrington's argument. Department's Position: This inadvertent error which we found during verification does not indicate a pattern of inconsistencies with regard to FAG Germany's reporting of early-payment discounts. The December 3, 1999, verification report makes clear that, during verification, we examined other transactions with early-payment discounts in order to determine whether this was an isolated case or whether there was a general pattern with the way FAG Germany reported its early-payment discounts. As noted on page 6 of the report, we did not find such a pattern. In addition, there is no evidence on the record to suggest that this was not an isolated case. Therefore, we are satisfied with FAG Germany's reporting of early- payment discounts and have not adjusted the reported information. Comment 25: Torrington argues that the Department should disallow all of FAG Germany's home-market extraordinary discounts. According to Torrington, FAG Germany reported extraordinary discounts as a form of negative other revenue in its home market database. Torrington argues that FAG Germany's reporting is not only confusing, since discounts and revenue have opposite effects in home-market price calculations, but FAG Germany did not explain why "extraordinary" discounts were granted or the conditions under which it allowed such "discounts." Torrington argues that, under the circumstances, the Department should disallow the adjustments in the calculation of normal value for all sales for which FAG Germany reported a value for "negative revenue". FAG Germany argues that the offset to its home-market other-revenue amounts were comprised of certain miscellaneous and extraordinary expenses which it incurred on a number of home-market transactions. According to FAG Germany, the amounts involved in these offsets are insignificant and represent the cost of miscellaneous services it provided to only a few customers. FAG Germany states that these miscellaneous service expenses are appropriately reported as offsets to certain other revenue items (e.g., the cost of certain freight and packing services that are billed back to the customer and reported as income items under other revenue), which are also incurred to facilitate the sale to the home-market customer. According to FAG Germany, the expense amounts involved and the number of transactions affected were so insignificant that it simply did not create a separate data field in which to report them. FAG Germany asserts that Torrington has not demonstrated that its reporting in this regard was in any way improper. Department's Position: Based on the record, we determined that FAG Germany's reporting of other revenue was not improper. The fact that FAG Germany included an extraordinary discount as a form of negative other revenue does not distort our calculation of normal value since discounts offset revenue. In addition, in its October 29, 1999, submission, FAG Germany explains why it included transactions that represent an extraordinary discount in the other-revenue field. We are satisfied with the respondent's explanation and find no reason to disallow all of FAG Germany's home-market extraordinary discounts reported in the field for other revenue. Comment 26: Torrington argues that the Department found during verification that one of FAG Germany's export-price customers claimed a U.S. discount that FAG Germany did not recognize; therefore, an unpaid balance remains on the sale that FAG Germany will presumably never collect. Torrington argues that the Department should adjust the credit expense to account for the unpaid amount or treat it as a direct bad-debt expense and adjust the price accordingly. According to Torrington, a decision not to make either adjustment would not accurately reflect the U.S. price and give FAG Germany benefit for an amount it never will receive. FAG Germany concurs with Torrington's suggestion that it would be appropriate to restate the credit expense for the sale in question, based on the deficiency between the amount paid and the amount billed, for a period of one year from the date of the invoice. Department's Position: Because we found at verification that a customer had claimed a discount and paid less than the original price on a specific transaction, we have deducted the claimed discount from FAG Germany's U.S. gross unit price for the sale in question for the final results. See the analysis memorandum for FAG Germany dated July 28, 2000. Comment 27: Torrington asserts that the Department should disallow FAG Germany's home-market rebates because FAG Germany did not demonstrate entitlement to such adjustments. Torrington argues that the Department's antidumping manual states that rebates must be contemplated at the time of sale to be allowable as home-market adjustments. According to Torrington, FAG Germany provides no evidence that rebates were contemplated at the time of sale or otherwise reflected in past understandings. In addition, Torrington asserts that, since the Department only examined rebate amounts during verification and not dates of agreements or pre-existing understanding, the verification report is unable to cure FAG Germany's inability to show entitlement. Torrington argues that, since FAG Germany has not met the legal standards for allowable rebates, the Department should disallow FAG Germany's home-market rebates. FAG Germany argues that it reported home-market rebates properly and that it and its customers agreed on all rebate percentages prior to the sales on which they were granted. According to FAG Germany, Torrington misconstrues the language in the Department's antidumping manual and overlooks specific evidence on the record that indicates that FAG Germany's rebate percentages were in fact known to the customer at the time of sale, either pursuant to a specific rebate agreement or through prior dealings with FAG Germany. Contrary to Torrington's assertions, FAG Germany claims that, in its response to the Department's questionnaire, it provided specific evidence that indicates that, at the point of sale, the rebate percentages were either (1) delineated to the home-market customer via individual agreement or (2) implicitly understood by the parties through prior business dealings and practices regarding rebates. Moreover, FAG Germany argues that, at the home-market sales verification, the Department verified FAG Germany's rebate programs and percentages in effect for fiscal years 1998 and 1999 as well as all rebate voucher information and the sales data upon which the rebates were based. According to FAG Germany, as in all prior review periods, its reporting of home-market rebates was in accordance with Departmental practice including the statements contained in the Department's antidumping manual. FAG Germany argues that Torrington's argument is not substantiated by the record evidence and should be rejected. Department's Position: In its September 3, 1999, submission at page 17, FAG Germany makes clear that it granted rebates to customers in 1998 pursuant to agreements or understandings that were in effect during that year. FAG Germany also makes clear that it agreed to a certain negotiated rebate percentage which is tied to the customer's achievement of an agreed- upon sales volume during the specified year. Also, in exhibit B-6 of its September 3, 1999, submission, FAG Germany's rebate chart outlines the customers which received rebates for 1998 purchases, dates on which FAG Germany and its customers all agreed upon the rebate percentages, and the amount of the rebate granted. Moreover, we verified FAG Germany's policy and practice for granting rebates and found no discrepancies. See the December 3, 1999, verification report at pages 6-7. Therefore, we are satisfied with FAG Germany's reporting of home-market rebates and find no reason to disallow them. Comment 28: Torrington argues that the Department should reject all of Koyo's downward billing adjustments to home-market prices reported as billing adjustment two because the reporting methodology was incorrect and distortive. Torrington contends that billing adjustment two is distortive because it includes adjustments which Koyo granted on a model-specific basis but allocated over all sales to the customer involved, as well as lump-sum adjustments granted on a customer-specific basis, with the result that the Department makes adjustments to transactions for which no adjustment actually applied. Torrington argues further that direct adjustments may not be allocated and treated as indirect adjustments. In rebuttal, Koyo argues that Torrington has offered no new reason why the Department should not reject its arguments in this review as it has in all the AFB reviews since the passage of the URAA. Department's Position: We are satisfied that Koyo has reported billing adjustment two to the best of its ability. This post-sale price adjustment is comprised of two types of adjustments: (1) Lump-sum adjustments negotiated with customers without reference to model-specific prices, and (2) adjustments granted on a model-specific basis but which Koyo records in its computer system on a customer-specific basis only. Given the large number of sales involved, it is not feasible to report this on a more specific basis. See AFBs 6, 62 FR at 54050-51, and AFBs 8, 63 FR at 33328. Furthermore, we examined this expense closely at verification in the previous review and found no indication that Koyo's methodology would result in distortive allocations. See Koyo Seiko Sales Verification Report, public version, for the 1997-98 administrative review, dated January 14, 1999, at pages 10-11. Also, in this review, there is no evidence on the record to indicate that Koyo's methodology would result in distortive allocations. Therefore, we have allowed Koyo's billing adjustment two as a direct adjustment to normal value. Comment 29: Torrington argues that the Department should deny certain home-market rebates Koyo claimed. Torrington contends that, instead of identifying the sales to a certain distributor and reporting the rebate for these sales only, Koyo allocated this rebate across all sales to the distributor. (The remainder of Torrington's comment refers to business proprietary information.) In rebuttal, Koyo argues that it reported its rebate expenses in these reviews in the same manner as it has in past reviews and that the Department has verified and accepted the claimed expense repeatedly. Koyo contends further that, during the POR, it did not have the capability in its computerized record-keeping system to distinguish between sales of bearings to the certain distributor for a specific application covered by the rebate and sales to the same distributor of these bearing models that, although suitable for the specific application for which the rebate was intended, were sold for different applications that were not covered by the rebate. Koyo argues further that the accuracy of Koyo's reporting methodology is evident by the fact that, as shown in its questionnaire response, the calculated rebate-factor percentage is identical to the actual rebate percentage Koyo granted to this distributor. Finally, Koyo argues that, given that its rebate-calculation methodology in the current review is the same as, and even more accurate than, that which it used and the Department verified in the previous review and accepted in all previous post-URAA reviews, the Department should likewise recognize that once again, in using that methodology, Koyo acted to the best of its ability in reporting the information. For these reasons, Koyo argues that the Department should reject Torrington's arguments and accept its reported home-market rebates. Department's Position: For these final results we have accepted claims for rebates as direct adjustments to price if we determined that the respondent, in reporting these adjustments, acted to the best of its ability and that its reporting methodology was not unreasonably distortive. Given that the calculated rebate-factor percentage is identical to the actual rebate percentage Koyo granted to this distributor, we find that Koyo's claimed rebate is not distortive. Thus, since Koyo has reported this rebate on as specific a basis as possible and the rate is not distortive, we have made a direct adjustment to home- market price for Koyo's rebates. Comment 30: Torrington argues that the Department should make certain to correct NSK's U.S. sales database for unreported rebates that NSK described in its supplemental questionnaire dated October 12, 1999, at page S-22. NSK contends that the Department made these corrections in its preliminary calculations. Department's Position: We made these corrections for our preliminary results. See lines 4940 through 4960 of the margin calculation program, which is attached to the NSK preliminary analysis memorandum, dated March 22, 2000. No change is necessary for the final results. Comment 31: FAG Germany argues that, in the U.S. price calculations, the Department unlawfully limited the adjustment of other revenue to the amount of reported inland-freight expenses. According to FAG Germany, it incurs freight expenses for transporting bearings from its U.S. warehouse facilities to its ultimate U.S. customer and it reported these freight expenses on a transaction-specific basis, taken directly from the invoice to the customer. According to FAG Germany, in some cases the freight revenue it earns will be the same as, less than, or more than the freight expense it incurred on a transaction. Therefore, according to FAG Germany, it may be under- or over-compensated for its direct freight costs. Further, FAG Germany argues that in many cases it does not bill the customer for freight expenses and no amounts appear in the other-revenue field. FAG Germany argues further that, for every administrative review prior to the 1997/1998 administrative review, the Department allowed FAG Germany the full adjustment to U.S. price for freight revenue directly billed and received on reported U.S. sales. Citing Certain Steel Concrete Reinforcing Bars from Turkey; Final Results of Antidumping Duty Administrative Review and New Shipper Review, 64 FR 49150, 49153 (September 10, 1999), FAG Germany argues that it is the Department's long-standing practice not to limit revenue directly incurred on U.S. sales to the amount of the corresponding expense. In addition, FAG Germany argues that the CIT has upheld the Department's past decisions to allow the full amount of freight revenue FAG Germany charged even though it exceeded, in some cases substantially, the amount of allocated freight expenses. FAG Germany, therefore, requests that the Department permit a full adjustment for other revenue reported on U.S. sales as in all reviews prior to the 1997-98 administrative review. Torrington argues that it is unclear from the record why invoice-specific freight revenue may or may not be the same as, less than, or more than the freight expense incurred in the transaction. According to Torrington, FAG Germany does not explain in it submissions why overpayment may occur. Torrington argues that, given the information FAG Germany submitted on the record, it is appropriate for the Department to limit the revenue by the freight amount. According to Torrington, the limit represents the common- sense assumption that customers pay only what they owe. In addition, Torrington argues that the CIT stated in Federal Mogul Corp v. United States, 862 F. Supp. 384 18 CIT 785, 818, (CIT 1994)(Federal Mogul,) that "the ITA is given discretion in its choice of methodology as long as the chosen methodology is reasonable and the ITA's conclusions are supported by substantial evidence on the record." Torrington requests that the Department also apply its method for the final results. Department's Position: We have reconsidered our position and have determined that it is inappropriate to limit the adjustment of FAG Germany's other revenue at the amount of reported inland-freight expenses when such revenues actually exceeded the amount of freight expense. In Federal-Mogul, the court upheld our position allowing the full amount of freight revenue FAG Germany charged. In this case, FAG Germany incurs freight expenses for transporting bearings from its U.S. warehouse facilities to its ultimate U.S. customer. FAG Germany reported these expenses on a transaction-specific basis because they are on the invoice to the customer. Since FAG Germany's reported gross price to its U.S. customers specifically includes amounts on the invoice for expenses and revenues, we find that it is appropriate and consistent with our practice to adjust FAG Germany's gross unit price for such expenses or revenues. See Final Determination of Sales at Less-Than-Fair-Value; Stainless Steel Sheet and Strip in Coils from the United Kingdom, 64 FR 30688, 30703 (June 8, 1999). Similarly, this same issue was argued in a prior review and the court upheld our position allowing the full amount of freight revenue FAG Germany charged. See Federal-Mogul. 5. Price Adjustments 5.A. Indirect Selling Expenses Comment 32: The petitioner argues that the Department should increase U.S. indirect selling expenses to account for advertising and technical services performed by INA on sales to the U.S. market. Torrington claims that INA maintains a website which is a source of advertising and technical service for customers in the United States. Torrington alleges that the expense of maintaining the INA website is an indirect selling expense incurred in connection with CEP sales and, accordingly, the expense is a selling expense which the Department should deduct when calculating CEP. Torrington asserts further that INA has not captured or described the activities related to these expenses in its response and requests that the Department restate these expenses by either requesting further information from INA or developing an upward adjustment for INA's reported amounts for indirect selling expenses. Torrington argues further that on December 21, 1999, it notified the Department about INA's website and asked the Department to issue a second supplemental questionnaire requiring further facts and value data, but, to Torrington's knowledge, the Department has not taken any such action. INA responds that, in accordance with 19 CFR 351.301(b)(2), Torrington's argument with respect to INA's website is untimely. INA asserts that, because the regulations require that submissions of unsolicited factual information in this review were due no later than October 18, 1999, the Department should reject Torrington's submission. INA also states that it will respond to any additional questions that the Department might ask regarding its website. Because the Department has not made additional inquiries, however, INA maintains that the Department should not make an adjustment as Torrington requests. Rather, INA claims, since it is common knowledge that websites inherently are global, the fact that the INA website lists an E-mail address for inquiries originating in the United States is no surprise. Citing its October 29, 1999, supplemental response at 48, INA contends that its U.S. affiliate reimbursed it for certain indirect selling expenses and any reimbursement the affiliate paid to the parent company would be included in the affiliate's claimed indirect selling expenses. Department's Position: We have examined the information on the record and have concluded that the record supports INA's contention that its reported U.S. indirect selling expenses capture the cost of all technical services. In particular, in INA's response to the questionnaire, dated August 23, 1999, at Exhibit C-13, it included a list of the sales overhead expenses shared with its U.S. affiliate and the calculation of the affiliate's portion of these indirect selling expenses. We conducted a detailed review of these overhead expenses and have concluded that the expenses associated with INA's website are most likely captured in the technology cost center. Furthermore, Torrington offered no concrete evidence to cause us to doubt that those expenses were captured within INA's indirect selling expenses. Therefore, we are satisfied that INA has accounted for these expenses and have made no adjustment for the final results. We will examine this issue further in a subsequent review. Finally, because we are satisfied that INA has accounted for any advertising or technical services related to maintaining INA's website, we have not addressed INA's argument regarding the timeliness of Torrington's allegations concerning the website. Comment 33: Torrington argues that NSK incurred numerous selling expenses on behalf of its U.S. affiliate's sales of merchandise which the Department should have deducted in calculating the CEP of NSK's sales, pursuant to section 772(d) of the Act. (The specific nature of the expenses is proprietary.) Torrington also argues that, if the Department does not deduct these expenses from CEP, it should not make a CEP-offset or level-of-trade adjustment to normal value in comparisons with NSK's CEP sales. Torrington claims that, if the Department does not deduct these expenses, there is essentially no difference in the selling functions between the level of the CEP and the home-market levels of trade. NSK contends that the Department's normal practice is to deduct only those selling expenses that are associated with economic activity occurring in the United States and related to the sale to an unaffiliated producer. NSK alleges that the Department has examined this issue with regard to the expenses in question in prior reviews and found that the expenses are not associated with activities occurring in the United States. NSK asserts further that the record demonstrates that NSK performs significantly fewer selling functions on sales to its U.S. affiliate than on sales to unaffiliated home-market customers. Thus, NSK argues, the Department correctly granted a CEP offset on its CEP sales. Department's Position: We deduct from CEP only those expenses associated with commercial activities in the United States that relate to the sale to an unaffiliated purchaser. However, we will deduct such expenses no matter where or when paid. See 19 CFR 351.402(b). After reviewing the record, we find that it is appropriate to deduct from CEP some, but not all, of the expenses Torrington identifies pursuant to section 772(d) of the Act. However, because the information regarding these expenses is proprietary in nature, please see the NSK final results analysis memorandum dated July 28, 2000, for a further discussion of our analysis. With regard to Torrington's contention that there is essentially no difference in the selling functions between the level of the CEP and the home-market levels of trade, the record demonstrates that NSK performed substantially more selling functions for its home-market sales than for its CEP sales. The identification by Torrington of a small number of minor expenses that NSK performed on behalf of its sales to its U.S. affiliate does not alter this fact. Therefore, because NSK's CEP sales are made at a less-advanced level of trade than its home-market sales and a level-of- trade adjustment could not otherwise be quantified, we have made a CEP- offset adjustment to normal value. 5.B. Inventory Carrying Costs Comment 34: NTN Japan argues that the Department should not have recalculated its home-market inventory carrying costs. NTN Japan contends that the Department verified NTN Japan's home-market inventory carrying costs and that the Department accepted NTN Japan's methodology in prior reviews. NTN Japan also claims that the Department did not present any reasons for a change in its methodology in the current review and that the Department should not create a facts-available situation when there is adequate verified data on the record. Torrington contends that the Department correctly recalculated NTN Japan's home-market inventory carrying costs in order to take into account the days in inventory. Department's Position: As we stated in the preliminary results analysis memorandum for NTN Japan dated March 22, 2000, at page 8, and in the verification report dated February 1, 2000, at page 7, NTN Japan did not take into account the days in inventory in its home-market inventory carrying costs calculation. Therefore, NTN Japan's calculation is unacceptable because it does not reflect the opportunity costs associated with the time in inventory accurately. However, for the preliminary results, we attempted to correct the problem by making an adjustment to NTN Japan's reported inventory carrying costs. For the final results, we have disregarded NTN Japan's reported inventory carrying costs entirely and recalculated these costs using our standard formula, which is to multiply the inventory value by the interest rate and then to multiply that product by the ratio of the number of days in inventory to the number of days in a year. Finally, with regard to NTN Japan's argument that we verified these costs, the verification of an expense does not mean that we necessarily accept the methodology the respondent used to report the expense. Comment 35: Torrington argues that the Department should recalculate NSK's U.S. inventory carrying costs using a different interest rate than the rate NSK used. Torrington contends that an examination of NSK's unconsolidated financial statements reveals that the rate NSK used is too low. Torrington suggests that the Department use a rate based on the facts available. NSK argues that Torrington's argument is flawed because it relies on data that is relevant to a time period prior to the POR. NSK contends that it demonstrated how it calculated its short-term borrowing rate in its response and that the Department verified and accepted NSK's calculation methodology in the prior review. Department's Position: NSK calculated its U.S. inventory carrying costs using the correct short-term borrowing rate. NSK demonstrated how it calculated the rate in exhibit B-18 of its September 3, 1999, response. Although we did not verify NSK's calculations for these administrative reviews, we verified NSK's methodology for calculating the rate in previous reviews and find that it has not changed for these current reviews. See NSK Sales Verification Report, public version, for the 1997- 98 administrative review, dated January 29, 1999, at pages 5-6. Therefore, we have not restated NSK's U.S. inventory carrying costs. Comment 36: The petitioner argues that the Department should only accept FAG Italy's reported U.S. inventory carrying costs to the extent that it is satisfied that the correct amounts have been reported. Torrington claims that the Department demonstrated concern in its supplemental questionnaire regarding FAG Italy's calculation of inventory carrying costs. Thus, Torrington contends, if the Department is not satisfied with FAG Italy's reported U.S. inventory carrying costs, it should make appropriate adjustments to those costs. Torrington suggests that the Department either apply a facts-available factor to FAG Italy's reported amounts for inventory carrying cost or select an appropriate factor used by another bearings respondent in any of the bearing reviews to calculate FAG Italy's inventory carrying cost. FAG Italy argues that it reported U.S. inventory carrying costs correctly and contends that Torrington's concern is not warranted. FAG Italy asserts that, if the Department was not satisfied with FAG Italy's response, it could have requested additional clarification. Department's Position: We have no reason to believe and there is no record evidence to support a conclusion that FAG Italy reported its inventory carrying costs for U.S. sales inappropriately. The Department requested further clarification in its supplemental questionnaire regarding FAG Italy's calculation of average time in inventory for its U.S. affiliates. In its supplemental response (October 25, 1999, at 32), FAG Italy provided a further explanation of inventory carrying costs, including its calculation of average time in inventory. FAG Italy has provided a complete and accurate reporting of the inventory carrying costs it incurred in the United States. Thus, Torrington's suggestion to apply a facts-available factor to FAG Italy's reported inventory carrying costs is inappropriate. 5.C. Credit Expenses Comment 37: The petitioner argues that the Department should recalculate the credit expense for a Somecat U.S. sale which had an incorrect payment date. Torrington refers to the Department's verification report of December 17, 1999, in which the Department states that Somecat officials explained that, for sales involving commissions, a sale is not considered closed until the commission is paid. Torrington argues that it is likely that Somecat reported credit expense based on credit terms rather than using the actual payment dates for other sales involving commissions. Thus, the petitioner claims that the Department should apply a facts- available credit amount for all other U.S. sales involving commissions, suggesting that the appropriate amount to apply is the longest credit period for any other reported U.S. sale. The respondent concurs that the payment date needs to be corrected and the credit expense be recalculated for this one U.S. sale. Somecat contends that the use of payment terms was reasonable, however, given that sale had not been paid at the time it prepared its original response to the Department's questionnaire. Somecat also states that this sale was the only U.S. sale involving a commission payment. Department's Position: We used the actual date of payment which we obtained at verification to recalculate the credit expense for this sale for our final results. Since a review of the U.S. sales database confirmed that this was the only U.S. sale where a commission was paid, we are not adopting the petitioner's suggestion concerning the application of facts available. Comment 38: Torrington states that, at verification, the Department found an incorrect payment date for one of SKF Sweden's home-market transactions which resulted in an overstatement of credit expenses for the specific transaction. Torrington also asserts that the payment error is significant from a statistical perspective since that one error was discovered in a sample pool of only eight transactions. Therefore, Torrington argues, the Department should apply an adverse inference that 12.5 percent (i.e., one out of eight sales) of SKF Sweden's home-market sales contain similar errors. Torrington argues that the Department should calculate the amount by which SKF Sweden's home-market credit expense was overstated and then multiply this value by a factor of 0.125 to account for the sampling pool the Department examined at verification. Torrington asserts that the Department should reduce all of SKF Sweden's home-market credit expenses by the resulting value. In rebuttal, SKF Sweden argues that the application of facts available is not warranted with respect to its home-market credit expenses. As SKF Sweden explained at verification, the discrepancy was particular to this transaction and was the result of a faulty bank transmission wherein SKF Sweden reported the date of the bank retransmission rather than the date of payment by the customer to the bank. SKF Sweden states that it corrected the error at verification by establishing the correct number of days to use in the calculation of credit. SKF Sweden argues that, as a result, the standard for applying facts available has not been met. Department's Position: SKF Sweden provided the correct payment date at verification. When we found this discrepancy at verification, we then examined additional transactions which we selected at random. Since we found no other discrepancies or indication of a pattern of misreporting of payment dates, we have no reason to doubt the validity of the rest of SKF Sweden's response. Therefore, the application of facts available is unwarranted. Comment 39: Torrington observes that the Department found that, when a customer paid by promissory note, NTN Japan added the number of days between receipt of the note and the maturity of the note to calculate the number of days between shipment and payment in its home-market credit- expense calculation. Torrington claims that NTN Japan did not describe the promissory notes in its response. Torrington argues that the Department should ensure that these notes did not bear interest or provide other consideration because, if they did, there is no basis to impute an expense during the period of the notes. Torrington suggests that, if the Department cannot determine from the record that the notes did not bear interest, the Department should restate NTN Japan's credit expense to omit all days covered by the promissory notes or, if that is impossible, deny any adjustment for home-market credit expenses. NTN Japan argues that Torrington did not cite any statutory authority in support of its argument nor did it provide any evidence that the Department should assume that interest was paid on the promissory notes at issue. Department's Position: There is no evidence on the record that demonstrates or even suggests that NTN Japan received interest on its promissory notes. Moreover, in our verification report dated February 1, 2000, at 9, we make no mention that NTN Japan received interest on its promissory notes. Therefore, we find that it would be inappropriate to assume that NTN Japan received such interest and we have not recalculated NTN Japan's reported home-market credit expenses. Comment 40: Torrington argues that the Department should calculate credit for all U.S. sales for which NSK had not been paid at the time of filing its supplemental response. Torrington suggests that the Department use customer-specific average credit days in calculating the credit expense for these sales. Alternatively, Torrington suggests using the highest credit factor for any sale to the customer in question. NSK concurs with Torrington's suggestion that the Department could use customer-specific average credit days in calculating the credit expense for the sales in question. NSK also suggests that, if any sales without payment dates remain, the Department could use the date of the supplemental response as the date of payment. Department's Position: We have recalculated the credit expense for the sales in question using the later of 1) the customer-specific average credit days, or 2) the date of the supplemental response as the date of payment. We have chosen this approach because, at the time of NSK's October 12, 1999, supplemental response, it had not received payment for these transactions. Thus, it would be inappropriate to use the customer- specific average credit days if that would result in a payment date prior to the date of the supplemental response. Comment 41: The petitioner argues that the Department should develop and apply a downward adjustment to all of FAG Germany's reported U.S. sales to account for unreported post-POR credit notes of the kind the Department discovered at verification. The petitioner asserts that the Department found during verification that, for one export-price sale, there was a post-POR credit note applicable to the sale for which FAG Germany did not account when it responded to the original questionnaire. According to the petitioner, as far as it can determine, the Department made no further inquiries regarding this reporting mistake. In addition, the petitioner contends that FAG Germany did not provide evidence demonstrating that the mistake was an isolated case or that there were no additional similar errors. Given this, according to the petitioner, the Department should extrapolate this error across all reported U.S. sales and apply the factor developed from the verified sales. Therefore, the petitioner argues, without evidence suggesting otherwise, it is more reasonable to assume that FAG Germany overlooked all credit notes issued after the review period when it prepared its list of U.S. sales. FAG Germany argues that there is nothing in the administrative record that indicates that this single oversight was anything more than an isolated clerical error. According to the respondent, the Department verified other export-price transactions and found no other evidence of misreporting or discrepancies. FAG Germany argues, therefore, that the petitioner's attempt to extrapolate this single reporting inadvertence to all export-price sales is simply not supported by any of the Department's findings or any other record evidence. Department's Position: We examined various randomly selected export-price sales at verification and found no indication of a pattern of discrepancies or misreporting of post-POR credit notes. In addition, there is no evidence on the record to suggest that this was not an isolated clerical error. Therefore, we find no reason to develop and apply a downward adjustment to all reported U.S. sales as the petitioner suggests. 5.D. Commissions Comment 42: NTN Japan argues that the Department's test to determine whether commissions paid to home-market affiliates are at arm's length is unreasonable because it does not account for factors such as the extent of actual services provided, number of sales, and volume of sales. NTN Japan contends that its commission arrangements are complex and that the Department's attempt to reduce its analysis to a bare comparison of average rates and nominal selling functions ignores commercial reality. NTN Japan concludes that the Department should use NTN Japan's commission rates as reported. Torrington contends that NTN Japan provided no concrete evidence that the Department's reliance on a comparison of commission rates is not appropriate to determine whether commissions paid to related sales agents were at arm's length. Moreover, Torrington argues, the Department's test conforms with prior practice. Department's Position: When we asked for the information from which we drew our conclusions about the commissions NTN Japan paid to affiliates, we stated "in order for us to determine whether the commissions you paid to affiliated agents were at arm's length, we require that you submit a chart which lists every commissionaire in the home-market (indicating whether the commissionaire is related or unrelated) and which shows all functions performed by the commissionaire. Without this information, we can not determine that the commissions you paid to affiliated agents were at arm's length by any means other than the commission rates you paid to your commissionaires." See NTN Japan Supplemental Questionnaire, dated October 21, 1999, at 4 (emphasis in original). Thus, it was clear that we were asking for this information in an attempt to make a determination as to whether the commissions NTN Japan paid to affiliates were at arm's length. If NTN Japan had concerns that the data we requested would be insufficient to make a proper determination, NTN Japan should have mentioned this in its response to our supplemental questionnaire and provided the data we would need in order to make such a determination. NTN Japan suggests that the rate-based comparisons we used in making our determination do not account sufficiently for the complexities of its situation. Perhaps our comparison could be refined. However, in the final analysis, in order to make a determination that the commissions the respondent paid were at arm's length, we have to make some kind of comparison of the rates paid to affiliates to the rates paid to unaffiliated commissionaires. NTN Japan provided no suggestions or data that we could have used to refine our analysis. Finally, NTN Japan appears to overlook the fact that, if we are unable to determine whether the commissions paid to affiliates are at arm's length, then, as with the arm's-length test for sales to affiliates, we would have no alternative but to disregard the commissions. It is not our responsibility to prove that the commissions paid were not at arm's length; rather, it is NTN Japan's responsibility to prove that they were at arm's length. See 19 CFR 351.401(b). NTN Japan has not done this with regard to the commissions it paid to the affiliated commissionaires in question and we have made our determination, as described in our preliminary analysis memorandum for NTN Japan dated March 22, 2000, at pages 7 and 8, based on the evidence on the record. Therefore, we have not changed our findings for the final results. 5.E. Advertising Expenses Comment 43: Due to discrepancies the Department found in verifying claimed advertising expenses, the petitioner contends that the Department should adjust SNFA U.K.'s indirect selling expenses to reflect a denial of the claimed amount for home-market advertising expenses. The petitioner also argues that, because SNFA U.K. understated its U.S. advertising expenses, the Department should apply a facts-available rate based on a total of all the reported U.S. amounts plus the total home-market amount. Alternatively, the petitioner proposes, the Department could use the highest U.S. advertising expense rate reported by the other respondents in this review. SNFA U.K. argues that it reported its home-market indirect advertising expenses correctly and that Torrington misconstrued the wording and findings contained in the verification report. SNFA U.K. states that the petitioner misunderstood its methodology for the allocation of indirect selling expenses to the home-market sales. Furthermore, SNFA U.K. remarks, all of these reported discrepancies concerning advertising expenses are related to the allocation of indirect selling expenses in the U.S. market and, in calculating export price, the Department disregards such U.S. indirect selling expenses. Thus, SNFA U.K. concludes, the point is irrelevant. Department's Position: SNFA U.K. calculated its home-market indirect selling expenses by multiplying the total indirect selling expenses incurred in all markets by an allocation ratio based on total home-market sales over total sales to all markets. The advertising expenses were indirect in nature. Thus, the total indirect advertising expenses SNFA U.K. used to calculate this variable included the advertising expenses incurred in all markets. Given SNFA U.K.'s allocation method, the Department is satisfied that the indirect selling expenses the respondent reported are accurate and non-distortive. Moreover, even if some advertising expenses which SNFA U.K. allocated to home-market sales could also have been attributed to U.S. sales, we do not use U.S. indirect selling expenses in the calculation of export price and, in this case, we did not adjust normal value using these expenses. Therefore, the petitioner's remarks concerning indirect selling expenses in the U.S. market are moot and the application of facts available for such expenses is unwarranted. 5.F. Technical-Service Expenses Comment 44: Torrington argues that SKF France's claim that its affiliate, Sarma, did not incur any direct technical-service expenses on its export- price sales is not supported by information on the record. Due to the demanding nature of the market to which Sarma sells (i.e., OEMs in the aerospace industry), Torrington claims that it is likely that Sarma incurs significant direct selling expenses for technical and engineering services. Torrington contends that this is confirmed by SKF France's reporting of a high degree of engineering services which Sarma performed in its selling functions. Citing AFBs 3 and AFBs 4, Torrington argues that, where the Department finds that a respondent has not distinguished between direct and indirect technical-services expenses, it is the Department's policy to treat such expenses as direct in the United States. When such information is lacking, Torrington continues, the Department makes a direct-expense deduction on the basis of facts available. Torrington concludes that the Department should calculate and apply a direct-expense rate based on facts available in this case. In rebuttal, SKF France argues that it reported technical-service expenses for its export-price sales correctly and that its indirect selling expenses are supported by evidence on the record. SKF France asserts that, contrary to Torrington's claim, its selling-functions chart supports the fact that engineering services are not direct expenses and supports its claim that Sarma's technical-service expenses are indirect in nature. SKF France comments that, after verifying Sarma's methodology in the 1997/98 review thoroughly, the Department accepted Sarma's information and, therefore, the Department should continue to accept its reporting for these final results. Department's Position: There is no evidence on the record that SKF France misclassified the expenses in question as indirect selling expenses. Furthermore, the petitioner's allegation alone does not call into question Sarma's responses. In response to the petitioner's reference to AFBs 3 and AFBs 4 and as we stated in AFBs 9, 64 FR at 35605, it is clear that in those reviews the Department found that the respondent did not distinguish direct and indirect expenses. Furthermore, in AFBs 3, 59 FR at 39742, in addition to not distinguishing between direct and indirect expenses, the respondent did not indicate that the expenses were all indirect in nature. However, in this review, SKF France distinguished between direct and indirect expenses and indicated that its technical-service expenses are indirect in nature. See SKF France's September 3, 1999, section B response, appendices B-44 and B-48. Therefore, we have accepted SKF France's expenses as reported. 5.G. Bank Charges Comment 45: NTN Japan argues that the Department should not calculate bank charges applicable to export-price sales and add those charges to normal value. NTN Japan claims that its reported selling prices include bank charges incurred in relation to home-market sales. Therefore, according to NTN Japan, adding these charges to normal value amounts to double-counting of the same charge. Torrington contends that NTN Japan does not address the Department's findings at verification but, rather, merely alleges that it reported these expenses elsewhere in the response. Department's Position: The bank charges NTN Japan incurred on its export- price sales are direct selling expenses. NTN Japan incurred these charges as a result of making export-price sales and the amounts of the charges are tied directly to each of NTN Japan's export-price sales. See the Department's Verification Report dated February 1, 2000, at page 5. Thus, it was proper to add such bank charges to normal value as a circumstance- of-sale adjustment. To the extent that NTN Japan incurred bank charges on home-market sales, it was NTN Japan's responsibility to claim an adjustment to normal value for such expenses. See 19 CFR 351.401(b). NTN Japan's failure to claim such an adjustment for its home-market sales does not excuse it from our requirement that respondents must report all expenses incurred on U.S. sales. Therefore, we have not changed our calculations for NTN Japan with respect to bank charges. 5.H. Repacking Expenses Comment 46: NSK argues that the Department should not include U.S. repacking expenses in its calculation of CEP profit. NSK argues that such expenses are properly classified as movement expenses deducted pursuant to section 772(c)(2)(A) of the Act, not, as the Department stated, selling expenses deducted pursuant to section 772(d)(1) of the Act. NSK contends that repacking is conceptually the same as expenses such as pre-sale warehousing, which the Department treats as a movement expense. NSK also contends that the fact that repacking expenses may be directly related to a particular sale does not make it a selling expense, observing that inland-freight expenses it incurs for delivery to the customer are directly related to particular sales. Finally, NSK argues that repacking expenses do not meet the definitional criteria of section 772(d)(1)(B) of the Act. According to NSK, credit, warranties, guarantees, and every other selling expense do not involve bringing goods from Japan to the U.S. unaffiliated customer, but U.S. repacking expenses do. Torrington observes that the Department has rejected this argument in prior reviews and that the Department should continue to reject this argument. Torrington contends that repacking has nothing to do with transporting goods or keeping them in storage for future transportation. Rather, according to Torrington, the expense bears a direct relationship to particular sales and is akin to credit expenses, guarantees, and warranties. Department's Position: Section 772(c)(2)(A) of the Act covers "transportation and other expenses, including warehousing expenses, incurred in bringing the subject merchandise from the original place of shipment in the exporting country to the place of delivery in the United States." See SAA at 824. As we stated in AFBs 8, 63 FR at 33339, and AFBs 9, 64 FR at 35622, we do not view repacking expenses as movement expenses. The repacking of subject merchandise in the United States bears no relationship to moving the merchandise from one point to another. The fact that repacking is not necessary to move merchandise is borne out by the fact that the merchandise was moved from the exporting country to the United States prior to repacking. We regard repacking expense as a direct selling expense because the company incurred the expense on individual products in order to sell the merchandise to the unaffiliated customer in the United States. See, e.g., AFBs 9, 64 FR at 35622. Therefore, we have deducted this repacking expense pursuant to section 772(d)(1)(B) of the Act, which directs us to reduce CEP by "expenses that result from, and bear a direct relationship to, the sale, such as credit expenses, guarantees, and warranties." Furthermore, because these expenses are direct selling expenses, we attribute profit to them pursuant to section 772(d)(3) of the Act by including them in the calculation of total CEP selling expenses. Comment 47: Torrington argues that the Department should restate FAG Italy's U.S. repacking costs or resort to facts available. Torrington claims that FAG Italy's explanation in its supplemental-questionnaire response is unacceptable and FAG Italy has not isolated or reported packing overhead amounts in its U.S. repacking calculation in conformity with the Department's antidumping methodology. Torrington argues that, if this adjustment cannot be made on the basis of information on the record, the Department should use data of another bearings respondent as facts available and recalculate FAG Italy's U.S. repacking expenses accordingly. FAG Italy claims that it reported all U.S. repacking costs for both materials and labor on a per-unit basis correctly. FAG Italy also asserts that it is impossible to extract the overhead costs related directly to packing operations in the United States. FAG Italy claims further that it captured overhead expenses in either packing-labor calculations or in pre- sale warehousing expense calculations. See FAG Italy's September 3, 1999, section C response at 50-51. FAG Italy claims that the Department has accepted its repacking methodology in prior reviews and, therefore, the Department should reject the petitioner's claim. Department's Position: Torrington's suggestion to apply a facts-available factor to FAG Italy's reported repacking costs is unwarranted. Section 776(a) of the Act, provides inter alia, for the use of facts available where a company fails to provide requested information in the form and manner requested. See, also, SAA at 869 (providing that the Department may use facts available to fill gaps in the record due to deficient submissions). However, in this review, we find no evidence on the record to indicate that FAG Italy has not accounted for repacking overhead expenses. The Department requested further clarification in its September 30, 1999, supplemental questionnaire and FAG Italy provided additional information in its October 25, 1999, supplemental-questionnaire response at pages 2 and 32. Furthermore, FAG Italy provided, in its September 3, 1999, response, a detailed packing-cost worksheet as Exhibit 24 which we find is consistent with our practice. Therefore, we have accepted FAG Italy's repacking methodology without changes for these final results. Comment 48: Torrington argues that the Department should treat certain expenses Nachi incurred at its warehouse in the United States, including the cost of tools, as movement expenses instead of as indirect selling expenses, which Nachi claimed. Torrington contends that tools used in the Nachi warehouse, as well as all other expenses at that site, are expenses incurred in bringing bearings from the factory in Japan to the place of delivery in the United States and should therefore be treated as movement expense. In rebuttal, Nachi states that it reported certain fixed expenses it incurred for repacking bearings at the warehouse of its affiliate as indirect selling expenses. Nachi contends that these expenses consisted primarily of the cost of tools used but not consumed. Nachi argues that Torrington's proposal to re-categorize all expenses incurred at the warehouse, including these tool costs, as movement expenses under 19 CFR 351.401(e)(2) is misguided. Nachi argues that the Department's questionnaire glossary describes indirect expenses as fixed expenses that are incurred whether or not a sale is made and that this is consistent with the one-time purchase of tools used in repacking, an expense independent of the individual sales. Nachi argues further that reclassification of this expense would be unwarranted, as the amounts are insignificant. Nachi contends that the Department, under its authority in 19 CFR 351.413, may disregard "insignificant adjustments" which are defined as an adjustment that has an "ad valorem effect of less than 0.33 percent." Department's Position: The repacking of subject merchandise in the United States bears no relationship to moving the merchandise from one point to another (see our response to comment 46, above). Rather, we view repacking expenses as direct selling expenses respondents incur on behalf of certain sales. We deduct such expenses pursuant to section 772(d)(1)(B) of the statute, which directs us to reduce CEP by "expenses that result from, and bear a direct relationship to, the sale, such as credit expenses, guarantees, and warranties." Nachi reported these expenses as indirect selling expenses. Although we consider these tool costs to be indirect expenses related to repacking the merchandise, after examining the information on the record, we determined that the amount of adjustment for these expenses would have an effect considerably less than 0.33 percent ad valorem. Therefore, we have made no change to our calculations. 5.I. Other Direct Selling Expenses Comment 49: NTN Japan argues that the Department was not justified in reallocating its home-market and U.S. expenses without regard to level of trade. NTN Japan contends that the Department has recognized differences in levels of trade for NTN Japan in its home-market and U.S. sales. NTN Japan also contends that the record demonstrates that selling expenses varied across levels of trade and that the allocation of selling expenses should reflect expenses incurred to accommodate varying levels of trade. Citing 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; Final Results of Antidumping Duty Administrative Review and Revocation in Part of an Antidumping Finding, 61 FR 57629, 57636 (November 7, 1996), NTN Japan argues that the Department has found that its allocation of expenses across levels of trade was acceptable and necessary to prevent distortion. Torrington contends that the Department has rejected this argument in prior reviews and that NTN Japan did not provide evidence that it incurred its selling expenses in the manner in which it allocated the expenses. Department's Position: We do not dispute that selling expenses differ between NTN Japan's levels of trade. Indeed, we reallocated NTN Japan's packing expenses for the preliminary results in order to calculate expenses that more accurately reflect NTN Japan's commercial situation. See Preliminary Results Analysis Memorandum for NTN Japan, dated March 22, 2000, at page 8. However, NTN Japan's methodology for allocating expenses to each of the levels of trade does not bear any relationship to the manner in which NTN Japan incurs the expenses in question at each of the levels of trade. Therefore, we find that the methodology is distortive because it allocates expenses to the levels of trade based on the relative sales values at each level of trade rather than how the expenses are actually incurred. Furthermore, we have addressed this issue in prior reviews. See AFBs 8, 63 FR at 33329, first addressed in AFBs 3, 59 FR at 39750. NTN Japan has not changed the methodology we rejected in these prior reviews nor has it presented any evidence that it incurred the selling expenses in the manner in which it allocated the expenses. 6. Level of Trade Comment 50: NTN Japan argues that the Department should not deduct expenses and profit from CEP sales prior to identifying the level of trade. NTN Japan claims that the statute sets forth no adjustments to price prior to conducting the level-of-trade analysis. Citing Borden v. United States, 4 F. Supp.2d 1221, Slip. Op. 98-36 at 57 (CIT March 26, 1998) (Borden), and Micron Technology, Inc. v. United States, Slip. Op. 99- 29 (CIT March 25, 1999) (Micron), NTN Japan claims that the CIT found that the Department's practice of deducting expenses and profit from CEP sales prior to identifying the level of trade is contrary to law. Torrington contends that the Department has rejected this argument in prior reviews and that the Department has appealed Borden to the CAFC. Torrington concludes that the Department should not modify its position. Department's Position: We believe that the statutory definition of "constructed export price" contained in section 772 of the Act indicates that we are to base CEP on the U.S. resale price adjusted for selling expenses and profit. As such, the CEP reflects a price exclusive of all selling expenses and profit associated with economic activities occurring in the United States. See SAA at 823. These adjustments are necessary in order to arrive at, as the term CEP makes clear, a "constructed" export price. The adjustments we make to the starting price, specifically those we make pursuant to section 772(d) of the Act ("Additional Adjustments for Constructed Export Price"), normally change the level of trade. Accordingly, we must determine the level of trade of CEP sales exclusive of the expenses (and concomitant selling functions) that we deduct pursuant to section 772(d) of the Act. Parties in NTN Bearing Corporation of America, et al, v. United States, Slip Op. 00-64 (CIT June 5, 2000) (NTN Bearing). In NTN Bearing the court held that the Department must deduct expenses pursuant to both sections 772(c) and 772(d) of the Act before determining the level of trade for U.S. and home-market sales, specifically declining to follow the rationale of Borden. In addition, we have appealed the CIT's decision in Borden and the Micron decision is not yet final. Therefore, for these final results, we have adjusted CEP under section 772(c) and 772(d) of the Act prior to conducting the level-of-trade analysis, as articulated in 19 CFR 351.412. When we conducted our analysis for the preliminary results, we, for the most part, did not consider differences in functions and their concomitant expenses deducted under section 772(c) of the Act. We did not consider expenses such as freight or import duties to be relevant to our analysis because they are not "selling functions." When we conducted our preliminary level-of-trade analyses, there were two functions that we did consider relevant whose expenses section 772(c) of the Act directs us to deduct. These expenses were packing for export and freight and delivery arrangements. However, when we deduct the expenses associated with these functions from the price for the level-of-trade analyses, our conclusions regarding the differences or similarities between the respondents' reported levels of trade do not change because the level of trade of CEP sales is still less advanced than the level of trade of home-market sales. Comment 51: NTN Japan argues that the Department must make a price-based level-of-trade adjustment when comparing CEP sales to home-market sales made at different levels of trade. NTN Japan claims that the Department found that levels of trade equivalent to the home-market levels of trade existed in the U.S. market based on sales to the first unaffiliated customer. NTN Japan argues that, because the Department found that a pattern of price differences existed between sales made at its home-market levels of trade, the Department must make a price-based level-of-trade adjustment. Torrington contends that the Department has rejected this argument in prior reviews and that NTN Japan does not explain why the record in this review warrants a different conclusion than that which the Department made in prior reviews. Department's Position: NTN Japan made export-price sales to the United States which were at one of the same levels of trade in the home market. Therefore, we tested for and found a pattern of differences at the home- market levels of trade. As we described in response to comment 50, we must determine the level of trade of CEP sales exclusive of the expenses (and concomitant selling functions) that we deduct pursuant to sections 772(c) and 772(d) of the Act. NTN's assertion that there are home-market levels of trade equivalent to the U.S. levels of trade is predicated on the assumption that we do not deduct expenses pursuant to sections 772(c) and 772(d) of the Act before determining the level of trade. This is not the case. Because no home-market levels of trade reported by NTN Japan were equivalent to the level of trade of its CEP sales after deducting those expenses, we were unable to make a level-of-trade adjustment for such sales. Therefore, we made a CEP-offset adjustment to normal value. Comment 52: Torrington argues that an analysis of the evidence on the record relating to Nachi's selling activities and levels of trade indicates that the Department should reject the respondent's claimed CEP offsets. Torrington agrees with the Department's findings that Nachi performs the same selling functions for all reported home-market sales and, therefore, all Nachi home-market sales are made at the same home- market level of trade. However, Torrington argues that the record does not demonstrate that Nachi engages in more activities when selling to affiliates in the home market than when selling at the CEP level of trade. Torrington contends that common sense suggests just the opposite, i.e., dealing with a foreign affiliate responsible for the U.S. market normally involves extensive interrelations not involved in local sales. Therefore, Torrington concludes, U.S. sales at the CEP level of trade are more remote than home-market sales and the Department should disallow Nachi's claims for a CEP-offset adjustment to normal value. In rebuttal, Nachi argues that the Department has found consistently that one level of trade exists for Nachi in the home market and that the selling activities in which it engaged to sell to that level of trade differ substantially from the selling activities associated with CEP sales. Furthermore, Nachi states that the Department confirmed during verification that its findings regarding levels of trade and channels of distribution were consistent with Nachi's questionnaire response, citing the Department's verification report, dated December 20, 1999. Finally, Nachi argues that, as the record demonstrates, the Department should continue to allow Nachi an adjustment for a CEP offset in accordance with section 773(a)(7)(B) of the Act. Department's Position: We conducted a thorough examination of Nachi's reported selling functions and services at verification and found that, after eliminating the selling functions corresponding to economic activities that occur in the United States pursuant to section 772(d) of the Act (i.e., those performed by Nachi's U.S. affiliates), there were significant differences between the selling activities associated with the CEP level of trade and those associated with the sole home-market level of trade. Therefore, we consider Nachi's CEP level of trade to be different from the home-market level of trade and at a less-advanced stage of distribution than the home-market level of trade. Because Nachi does not have a home-market level of trade which is equivalent to the level of trade of its CEP sales, we were unable to make a level-of-trade adjustment for such sales. Consequently, to the extent possible, we determined normal value for CEP sales at the same level of trade as the U.S. sales to the unaffiliated customer and made a CEP-offset adjustment in accordance with section 773(a)(7)(B) of the Act. Comment 53: NSK argues that the Department should grant a partial level- of-trade adjustment when comparing its CEP sales to home-market aftermarket sales. NSK contends that the record shows that there is a pattern of price differences between its home-market OEM and aftermarket levels of trade and that the aftermarket level is more advanced than the OEM level of trade. While NSK acknowledges that the home-market OEM level of trade is not equivalent to the CEP level of trade, it argues that there is nothing in the law prohibiting the Department from making a partial level-of-trade adjustment in this situation. Torrington observes that the Department has rejected this argument in prior reviews and that the Department should continue to reject this argument. Department's Position: There is no provision in the statute for making such a partial adjustment. We make a level-of-trade adjustment when there is "any difference between the export price or constructed export price and the normal value that is shown to be wholly or partly due to a difference in level of trade between the export price or constructed export price and the normal value." See section 773(a)(7)(A) of the Act. We interpret the statutory phrase "wholly or partly due to a difference in level of trade" to mean that we may make a level-of-trade adjustment only if part of the differences in prices between levels of trade is attributable to the difference in level of trade. Thus, we do not read into this language of the statute the authority to make a level-of-trade adjustment between two home-market levels of trade where neither level is equivalent to the level of trade of the U.S. sale. See 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; Final Results of Antidumping Duty Administrative Reviews, 63 FR 2558, 2578 (January 15, 1998), and AFBs 9, 64 FR at 35609. Further, the CIT has upheld our practice recently in NTN Bearing Corp. of America, et al, v. United States, Slip Op. 00-64, at 44 (CIT June 5, 2000). Comment 54: Torrington Nadellager contends that the CV profit the Department used to calculate normal value for models sold in the United States to "large" OEMs is based on inadvertently and inaccurately reported home-market data. To correct its inadvertent reporting and calculate CV profit properly, Torrington Nadellager asserts that the Department should re-categorize certain sales in the "large" OEM channel of distribution as sales in another channel and, ultimately, as sales at a different home- market level of trade. Torrington Nadellager contends that, if it had not made the errors in its channel-of-distribution reporting, the Department would not have calculated an "extremely high" CV-profit rate on home- market sales to "large" OEMs and, therefore, its margin would be much lower. Torrington Nadellager explains that, when it assigned the channel-of- distribution codes to specific customers, it focused only on the customer's name and the overall activities associated with the customer but not on specific activities surrounding individual transactions. Torrington Nadellager argues that, based on its analysis of each above- cost sale in the "large" OEM channel of distribution, for the following reasons the Department should re-categorize the sales to 1) a division of a "large" OEM that purchased the merchandise and engaged in activities that Torrington Nadellager associates with "small" OEMs; 2) "large" OEMs that consumed the merchandise as replacement parts rather than in the manufacture of original equipment; and 3) "large" OEMs that purchased the merchandise as samples and/or prototypes. Torrington Nadellager also asserts that sales to the same customer category, e.g., "large" OEM, can be classified in different channels of distribution. In support of this argument it cites the Department's decision in Tapered Roller Bearings, Four Inches or Less in Outside Diameter, and Components Thereof, From Japan; Final Results of Antidumping Duty Administrative Review, 57 FR 4975, 4981 (February 11, 1992) where the Department classified certain sales differently, even to the same customer, since the facts of the sales supported such a distinction. In its case brief, Torrington Nadellager provides invoices and purchase orders to support its request that the Department correct the alleged clerical errors. Citing NTN Bearing Corp. v. United States, Slip. Op. 94- 1186 (CAFC 1996) (NTN), the respondent contends that the Court's decision requires the Department to accept the new materials it provided since they are necessary to illustrate the inadvertent error. It also asserts that the facts in NTN are similar to the situation here. Torrington Nadellager contends further that the Department must correct inadvertent errors to avoid manifestly unjust results, even if the errors are discovered subsequent to the deadline for submitting information. In support of this argument the respondent cites the Court's decisions in NTN and Zenith Electronics Corp v. United States, 74 F.3d 1204, 1206-8 (CAFC 1996). FAG Germany, INA, and SKF Germany contend that the Department should reject as untimely the information Torrington Nadellager provided in support of reassigning the distribution channels since Torrington Nadellager submitted the data beyond the regulatory deadline for new factual information. In support of this argument, they cite 19 CFR 351.301(b)(2) which specifies 140 days from the last day of the anniversary month as the cutoff for the submission of new factual information. INA and SKF Germany provide examples of case precedent to support their argument that the Department should reject Torrington Nadellager's request for the correction for clerical errors. INA cites to Certain Fresh Cut Flowers From Colombia; Final Results of Antidumping Duty Administrative Reviews, 61 FR 42833, 42834 (August 19, 1996) (Flowers 5 - 7), and reviews the conditions the Department established in that case for accepting a respondent's correction of alleged clerical errors after the preliminary results. INA contends that Torrington Nadellager's clerical-error submission does not satisfy the first and second of the six conditions outlined in Flowers 5 - 7 which state that 1) the error in question must be demonstrated to be a clerical error, not a methodological error, an error in judgment, or a substantive error and 2) the Department must be satisfied that the corrective documentation provided in support of the clerical error is reliable. With regard to the second of the six conditions outlined in Flowers 5 - 7, INA and SKF Germany argue that Torrington Nadellager has not established the reliability of the new factual information and that, in order for the information to be considered reliable, the Department would need to collect additional information and/or verify Torrington Nadellager's newly submitted data. INA and SKF Germany also assert that the Department would need to review information supporting that Torrington Nadellager classified the channels of distribution for all other sales transactions properly. They assert that it would be distortive to address only the specific misclassification errors that Torrington Nadellager raised without an inquiry into the potential for other coding problems. Furthermore, they assert that the facts on the record do not support reclassifying the transactions. Regarding Torrington Nadellager's request to reclassify certain sales of samples or prototypes to "large" OEMs, INA asserts that there is an apparent conflict between arguments in Torrington Nadellager's case brief and information contained in its section A response. In the Section A response, Torrington Nadellager indicated that sample and prototype sales to "large" OEMs are part of the "large" OEM channel of distribution. SKF Germany also argues that sample and prototype transactions are obviously part of the sales-distribution process with "large" OEM customers and should not be unilaterally reclassified without appropriate documentation and argument. SKF Germany affirms its position that the calculation methodology for CV profit must be flexible and should yield reasonable and representative rates. SKF Germany takes issue, however, with the timeliness of the factual data Torrington Nadellager submitted in support of its clerical- error corrections. Therefore, SKF Germany requests that the Department reject the arguments and data presented in Torrington Nadellager's case brief. Citing the Court's decision in Technoimportexport v. United States, 15 CIT 250, 258-259, 766 F. Supp. 1169, 1178 (1991), SKF Germany asserts that the only exception to the rejection of untimely submitted information is for an error "so egregious that the failure to correct it was an abuse of discretion and undermined the interests of justice." SKF Germany also contends that, in accordance with AFBs 1, 56 FR at 31741, the exception only applies when the Department is "able to assess from information already on the record that an error has been made or that the new information is accurate." SKF Germany asserts that the errors in question are neither obvious from the record nor egregious and, therefore, it requests that the Department reject the factual information submitted by Torrington Nadellager. SKF Germany also contends that Torrington Nadellager's situation is distinguishable from the "obvious" or "clerical- error" mistakes with which the Department was faced in NTN, where the respondent inadvertently miscoded sales to a third country as U.S. sales. SKF Germany asserts that, unlike NTN, which involved the statutorily precluded action of using Canadian sales in the calculation of dumping margins for U.S. sales, Torrington Nadellager's situation involves a subjective matter in which no clear demarcation or definition of channel of distribution exists. In support of its argument that Torrington Nadellager's errors are ones of judgment and data interpretation rather than clerical, SKF Germany notes the extensive arguments and documentation the respondent submitted for the alleged clerical error. Department's Position: On page A-4 of our June 24, 1999, questionnaire, we provided the following instructions to Torrington Nadellager regarding its reporting of information for channels of distribution and sale process: The description you provide of your channels of distribution and sales process (question 4 below) is intended to provide the Department with the information necessary to make appropriate comparisons of sales at the same level of trade or to adjust normal value, if appropriate, when sales are compared at different levels of trade. Your response to this section may be of critical importance to this review. Accordingly, your response should include all the information requested and all information you believe the Department should consider in making a comparison." We relied on Torrington Nadellager's response to the above request and the channel-of-distribution codes it reported to designate the different levels of trade for home-market sales. We then calculated CV profit by multiplying the level-of-trade-specific weighted-average profit rate calculated on home-market sales made in the ordinary course of trade by the cost of production of a specific model. For the reasons described below and based on our response to Comment 58, we have not altered our calculation methodology for these final results. Torrington Nadellager has not satisfied our standard for clerical-error corrections and, therefore, we have not made the requested corrections. As a result of the NTN decision to which Torrington Nadellager cites, we re- evaluated our policy for accepting clerical errors which respondents made in their submissions. See Antidumping Duties; Countervailing Duties; Final Rule, 62 FR 27296 (Final Rule), 27327 (May 19, 1997). In Flowers 5 - 7, we established a new policy for accepting a respondent's clerical errors. Specifically, we established that we would accept corrections of clerical errors under the following conditions: 1) the error in question must be demonstrated to be a clerical error, not a methodological error, an error in judgment, or a substantive error; 2) the Department must be satisfied that the corrective documentation provided in support of the clerical- error allegation is reliable; 3) the respondent must have availed itself of the earliest reasonable opportunity to correct the error; 4) the clerical-error allegation, and any corrective documentation, must be submitted to the Department no later than the due date for the respondent's administrative case brief; 5) the clerical error must not entail a substantial revision of the response; and 6) the respondent's corrective documentation must not contradict information previously determined to be accurate at verification. As indicated above, Torrington Nadellager claims that it made clerical errors in its reporting of channel-of-distribution codes for certain sales. We find, however, that Torrington Nadellager's clerical-error submission does not satisfy the first and second of the six conditions outlined in Flowers 5 - 7. As indicated above, the first condition outlined in Flowers 5 - 7 for correcting a respondent's clerical error is that the party must demonstrate that the error in question is a clerical error, not a methodological error, an error in judgment, or a substantive error. Contrary to Torrington Nadellager's claim, we have not found that the alleged errors are the type normally associated with clerical errors. We discuss clerical errors in our regulations; specifically, at 19 CFR 351.224(f) the definition of "ministerial error" can be found. There we state that "ministerial error means an error in addition, subtraction, or other arithmetic function, clerical error resulting from inaccurate copying, duplication, or the like, and any other similar type of unintentional error which the Secretary considers ministerial." Although Torrington Nadellager raises the coding issue as a clerical error, asserting that it assigned the wrong channel-of-distribution code inadvertently, we do not find that the record corroborates this claim. We analyzed information on the record with regard to each of the following categories of transactions that Torrington Nadellager would like us to re- categorize: 1) sales to a division of a "large" OEM that purchased the merchandise and engaged in activities that Torrington Nadellager associates with "small" OEMs; 2) sales to a "large" OEM that consumed the merchandise as replacement parts rather than in the manufacture of original equipment; and 3) sales to "large" OEMs that purchased the merchandise as samples and/or prototypes. For the reasons described below, we find that the record supports that the alleged miscoding for these transactions is a substantive issue and, if an error at all, an error in judgment. Therefore, we find that Torrington Nadellager's situation is distinguishable from the "obvious" or "clerical-error" mistakes in NTN. As stated above, with regard to all three categories of sales to "large" OEMs that Torrington Nadellager requests that we re-categorize to another level of trade, we do not find that the record supports Torrington Nadellager's request. Specifically, the types of customers that purchased the foreign like product from Torrington Nadellager are essentially the same, the point in the channel of distribution at which the selling activities occurred is identical, and there is no evidence of a significant difference in selling activities. See Torrington Nadellager March 13, 2000, preliminary results analysis memorandum at page 3 for a detailed description of the factors we analyzed in identifying the home- market levels of trade. Therefore, we find no reason to distinguish these transactions from Torrington Nadellager's other sales to "large" OEMs. While we do not dispute Torrington Nadellager's assertion that it is possible to classify sales to the same customer category, e.g., "large" OEM, in different channels of distribution, to support such a classification the respondent bears the responsibility to make the distinction and provide timely information and argument to substantiate the differences. For these reasons, we find that Torrington's failure to analyze the transactions in sufficient detail or give us sufficient information on the difference in selling functions performed and services offered is a substantive matter and an error in judgment, not a clerical error. Furthermore, there is no information on the record that specifically precludes the transactions in question from being categorized as sales to "large" OEMs. Finally, regarding the third category of transactions (i.e., sales to "large" OEMs that purchased the merchandise as samples and/or prototypes), the record supports Torrington Nadellager's original channel-of-distribution coding. The description Torrington Nadellager provided for these sales on pages A-21 and A-22 of its September 3, 1999, Section A questionnaire response supports that Torrington Nadellager developed prototypes and supplied samples to the "large" OEMs. Thus, the information on the record supports that the originally designated channel-of-distribution codes for these transactions are appropriate. The second condition for correction of a respondent's clerical errors also has not been met. Without collecting additional information and documentation on the alleged clerical errors, we cannot ascertain the reliability of the corrective documentation. As specified in Flowers 5 - 7, the second condition for accepting corrections of clerical errors is that the Department must be satisfied that the corrective documentation provided in support of the clerical error is reliable. Torrington Nadellager did not raise the clerical-error argument or provide any documentation in support of such an argument until a very late stage in the administrative process. Essentially, it precluded us from establishing the reliability of the corrective documentation and pursuing the line of inquiry necessary to establish that, in fact, a clerical error had occurred. If the respondent had examined its designation of channels of distribution more thoroughly before verification and then raised the issue, it may have been relatively easy to establish the reliability of any necessary corrections. Moreover, raising the error in a timely fashion would have led to a detailed examination of the channel-of-distribution designations for other transactions and may have confirmed whether any other coding problems exist. During our analysis of the data submitted by Torrington Nadellager on channels of distribution, we have not discovered any evidence of errors in its reporting. Furthermore, we are not satisfied that the corrective documentation provided by Torrington Nadellager in support of its clerical- error allegation is reliable, especially since, as noted above, certain record evidence conflicts with the clerical-error claims. Thus, for the final results, we have not re-categorized Torrington Nadellager's above- cost sales in the "large" OEM channel of distribution to a different level of trade. Finally, the information Torrington Nadellager provided in support of reassigning the distribution channels is untimely since the alleged error has been determined to be non-clerical and the respondent provided the data beyond the regulatory deadline for new factual information. We did not reject its case brief of May 8, 2000, however, so that we could analyze its arguments and the documentation presented fully in order to determine whether, in fact, a clerical error may have occurred. 7. Samples and Sales Outside the Ordinary Course of Trade Comment 55: NTN Japan argues that its home-market sample sales are outside the ordinary course of trade and that the Department should exclude these sales from its calculation of normal value. NTN Japan argues that the purpose of the ordinary-course-of-trade provision of the statute is to prevent dumping margins from being based on sales which are not representative of the practices in the home market. NTN Japan contends that customers of these sales specifically requested samples and that the products are used only for testing purposes as opposed to the normal use of bearings. NTN Japan also contends that the Department's refusal to determine that such sales are outside the ordinary course of trade because there are a large number of such sales is illogical. According to NTN Japan, the Department should examine the ratio of sample sales to total sales rather than consider just the total number of sample sales. Torrington argues that NTN Japan has not substantiated its claim that its home-market sample sales are outside the ordinary course of trade. Torrington claims that the data NTN Japan submitted comparing the price and quantity of sample sales versus "normal" sales does not provide sufficient contrast to support NTN Japan's claim that its sample sales were outside the ordinary course of trade. Torrington also remarks that the Department has rejected this argument in prior reviews. Department's Position: The ordinary-course-of-trade provision of the statute ensures that we base normal value on sales that are representative of the home market. The burden rests on NTN Japan, however, to demonstrate that its sample sales are actually outside the ordinary course of trade. NTN Japan has not met this burden. NTN Japan bases its claim largely on the fact that its customers use sample sales for a purpose other than that for which customers normally use bearings (i.e., for testing purposes rather than in production). However, the fact that these sales are used for testing purposes does not, in and of itself, demonstrate that the sales are outside the ordinary course of trade. As we have stated in prior reviews of these orders, the standard for finding certain sales to be outside the ordinary course of trade is high. See, e.g., AFBs 8, 63 FR at 33342-44. Also, we base our determinations of ordinary course of trade on the totality of circumstances. Id. In this case, NTN Japan has not shown that its sample sales are in any way unrepresentative of its other sales. As Torrington points out, the prices for NTN Japan's sample sales are frequently similar to the prices for its other sales. Because there is no evidence to indicate that these sales are outside the ordinary course of trade, we have included them in our calculation of normal value. Comment 56: NTN Japan argues that sales with abnormally high profits are outside the ordinary course of trade and that the Department should exclude these sales from the calculation of normal value. According to NTN Japan, the Department's regulations indicate that merchandise sold at aberrational prices or with abnormally high profits may be examples of sales outside the ordinary course of trade. NTN Japan contends that its high-profit sales are not representative of other sales in the market and are, therefore, outside the ordinary course of trade. Citing CEMEX, S.A. v. United States, 133 F.3d 897, Slip. Op. 97-1151 (CAFC 1998), NTN Japan alleges that the CAFC upheld a finding by the Department that certain types of cement were sold outside the ordinary course of trade because of significant differences in profit levels. NTN Japan claims that the evidence it provided in its response for these AFB reviews demonstrates that sales with high profits are abnormal and should be considered outside the ordinary course of trade. Torrington contends that the Department has rejected this argument in prior reviews and that NTN Japan did not provide evidence, other than the profit levels, to support its assertion that the sales in question are outside the ordinary course of trade. Department's Position: As we have stated in prior reviews, high profits, by themselves are not sufficient for us to determine that sales are outside the ordinary course of trade. See AFBs 9, 64 FR at 35620, 35621. In order to determine that a sale is outside the ordinary course of trade due to abnormally high profits, there must be unique and unusual characteristics related to the sales in question which make them unrepresentative of the home market. See 19 CFR 351.102. Other than the allegedly high profits, NTN Japan has not provided any evidence suggesting that these sales have any characteristics that would make them unrepresentative of the home market. See AFBs 8, 63 FR at 33344. Furthermore, NTN Japan has not shown that these profit levels are in any way aberrational. Indeed, NTN Japan has not demonstrated or provided any evidence to support its assertion that these sales are somehow unusual and, therefore, outside the ordinary course of trade. Consequently, we have not excluded NTN Japan's so-called "high-profit" sales from our analysis. 8. Cost of Production and Constructed Value 8.A. Profit for Constructed Value Comment 57: FAG Germany, FAG Italy, Barden, INA, Izumoto, and NSK argue that the Department's CV-profit methodology is contrary to law and ignores the statutory definition of foreign like product. The respondents contend that the Act instructs the Department to calculate profit for CV on amounts for the foreign like product of the merchandise. According to FAG Germany, FAG Italy, Barden, INA, and NSK, if the foreign like product of the U.S. sale is an identical bearing, CV profit should be based on profit amounts for identical bearings; if the foreign like product of the U.S. sale is a bearing family, CV profit should be based on the profit amounts for bearings in the same family. Izumoto argues that CV profit should be based on the profit amounts for family matches. FAG Germany, FAG Italy, and Barden cite Mitsubishi Heavy Industries, Ltd. v. United States, 54 F. Supp.2d 1183 (1999). They also argue that the CIT's decision in RHP Bearings, Ltd. v. United States, 83 F. Supp.2d 1322, 1336 (December 16, 1999) (RHP), was decided incorrectly because it upheld the Department's methodology without ever ascertaining the definition of "foreign like product" upon which the Department relied. INA also argues that determining and applying profit on a foreign-like-product basis is not any more complicated than other calculations routinely performed by the Department in a review and that it would generate more accurate results than the Department's aggregated-profit approach. NSK provides an analysis of its data that, according to NSK, rebuts the Department's prior position that NSK's construction of the statute makes the statutorily preferred CV-profit methodology inapplicable to most cases involving CV. NSK suggests that under its construction of the statute the Department is able to make numerous CV matches using the preferred CV- profit methodology. Therefore, NSK claims, this evidence demonstrates that one of the bases by which the CIT sustained the Department in RHP does not exist in this review. FAG Germany, FAG Italy, Barden, and INA contend further that, if the Department calculates CV profit on a class-or-kind basis rather than on a more specific basis, the Department should not exclude below-cost sales when it calculates CV profit. According to the respondents, the Department's methodology for calculating CV profit is not authorized by section 773(e)(2)(A) of the Act but is authorized by section 773(e)(2)(B)(i) of the Act. The respondents contend that, although profit calculated pursuant to section 773(e)(2)(A) of the Act is limited to sales made in the ordinary course of trade, the same is not true of profit calculated pursuant to section 773(e)(2)(B)(i) of the Act. SKF France, SKF Germany, SKF Italy, and SKF Sweden argue that the Department should calculate CV profit on a class-or-kind basis but that the Department should not exclude below-cost sales when it calculates CV profit for their companies. The SKF companies claim that the Department usually resorts to CV when sales of an identical or similar product are unavailable and, thus, it is proper for the Department to calculate CV profit pursuant to section 773(e)(2)(B)(i) of the Act. The SKF companies also argue that the CIT's decision in RHP is not dispositive because it appears to be inconsistent with the earlier-decided Mitsubishi case and the decision in RHP does not mandate that the affirmed methodology be used in this review. The SKF companies argue that, if the Department continues to exclude below-cost sales from its CV-profit calculation, it should at least include the value of the below-cost sales in the denominator when it calculates the rate of CV profit. In the opinion of the SKF companies, the statute gives the Department the discretion to calculate profit in a manner that better reflects a respondent's actual profits by including the total revenues received on unprofitable sales in its calculations. The SKF companies claim that this position is supported by the Department's different methodologies for calculating deposit and assessment rates. Torrington observes that the Department has rejected these arguments in prior reviews and that the Department should continue to reject them for the same reasons on which the Department relied in those reviews. Torrington also contends that the respondents' statutory interpretation would make the statutorily preferred method inapplicable in most cases and would result in an overly complex calculation. Torrington argues further that the comparison of the SKF companies to the Department's assessment- rate calculation is inapposite. According to Torrington, the assessment- rate calculation is designed to allow the Customs Service to collect the proper amount of duties due on the basis of the entered values but that there is no equivalent concern in the calculation of a CV-profit rate. Finally, Torrington argues that the respondents' reliance on Mitsubishi is misplaced because the CIT decided that case based on the context of the specific record involved. According to Torrington, the facts in that case are very different from the facts in this case. Torrington claims that the product involved in Mitsubishi is a very expensive product sold in very small quantities and thus is "the polar opposite" of AFBs. Torrington also asserts that Mitsubishi case is different because the Department did not use "family" comparisons but, instead, relied on a differences-in- merchandise test. Department's Position: Section 773(e)(2)(A) of the Act directs us to include in our calculation of CV "the actual amounts incurred and realized by the specific exporter or producer being examined in the investigation or review for selling, general, and administrative expenses, and for profits, in connection with the production and sale of a foreign like product, in the ordinary course of trade, for consumption in the foreign country" unless there are no sales of the foreign like product. Thus, the preferred method for computing CV profit is for us to use actual profit and selling, general, and administrative expenses incurred on sales of the foreign like product. As we stated in prior reviews (see, e.g., AFBs 9, 64 FR at 35610), an aggregate calculation that encompasses all foreign like products under consideration for normal value represents a reasonable interpretation of section 773(e)(2)(A) of the Act. Moreover, in applying the preferred method for computing CV profit under section 773(e)(2)(A) of the Act, the use of aggregate data results in a reasonable and practical measure of profit that we can apply consistently where there are sales of the foreign like product in the ordinary course of trade. In the preamble to our regulations, we stated: The Department recognizes that there are other methods available for computing SG&A and profit for CV under section 773(e)(2)(A) of the Act, including those suggested by the commenters. We continue to believe, however, that an aggregate calculation that encompasses all foreign like products under consideration for normal value represents a reasonable interpretation of the statute. This approach is consistent with the Department's method of computing SG&A and profit under the pre-URAA version of the statute, and, while the URAA revised certain aspects of the SG&A and profit calculation, we do not believe that Congress intended to change this particular aspect of our practice. Moreover, the Department believes that in applying the preferred method for computing SG&A and profit under section 773(e)(2)(A), the use of aggregate data results in a reasonable and practical measure of profit that the Department can apply consistently in each case. By contrast, a method based on varied groupings of foreign like products, each defined by a minimum set of matching criteria shared with a particular model of the subject merchandise, would add an additional layer of complexity and uncertainty to [antidumping] proceedings without generating more accurate results. See Final Rule, 62 FR at 27359. In addition, we disagree with the respondents' interpretation of the term "foreign like product." In accordance with the definition of foreign like product under section 771(16) of the Act, it is clear that "foreign like product" is not limited to the product which is identical in physical characteristics to the subject merchandise (section 771(16)(A) of the Act) or even to the product that is similar to the subject merchandise (section 771(16)(B) of the Act). Merchandise of the "same general class or kind" as the subject merchandise (section 771(16)(C) of the Act) will qualify as the "foreign like product" in cases where either the identical or the similar merchandise is not available. There is no indication that, by referring to "a foreign like product" in section 773(e)(2)(A) of the Act, Congress intended that profit be calculated upon the basis of merchandise that is identical or most similar to the subject merchandise. If Congress had such intentions, then the "preferred" method provided in section 773(e)(2)(A) of the Act would rarely be applicable since CV ordinarily becomes necessary for determining normal value when identical or similar home-market merchandise is not available for comparison to the U.S. merchandise. Furthermore, as the respondents have noted, the CIT has upheld our practice in RHP, which was the result of litigation of an administrative review of the AFB orders. In RHP, the CIT affirmed the Department's use of aggregate data that encompassed all foreign like products under consideration for normal value for purposes of CV profit because the use of such aggregate data matched the criteria of section 771(16)(C) of the Act - i.e., the same general class or kind of merchandise. Given that the CIT has specifically upheld the same methodology we used in these reviews, we see no reason why the CIT's finding in RHP would not be dispositive. We also disagree with INA that calculating profit on a product-by-product basis is not any more complicated than calculating weighted-average prices or CV for each product. In general, the respondents have reported numerous varieties of bearings which fall into hundreds of product or family categories. Calculating CV profit on a match-by-match basis would require a product-by-product analysis and profit-calculation determination. For certain products, if there were sales (i.e., sales within the ordinary course of trade) of identical or family bearings, we would be able to use the preferred method under section 773(e)(2)(A) of the Act to calculate profit. However, for other bearing families, we would need to determine which of the three alternative methods under section 773(e)(2)(B) of the Act would be appropriate based on the factual situation before us. Given the number of bearing families, this would add layers of complexity which the Department does not face in calculating weighted-average prices or in calculating an aggregate profit figure. In the Department's view, Congress did not intend such a result when it enacted section 773(e)(2) of the Act. With respect to our position that the respondents' construction of the statute makes the statutorily preferred CV-profit methodology inapplicable to most cases involving CV, NSK's claim that one of the bases by which the CIT sustained the Department in RHP does not exist in this review is factually wrong and, in NSK's case, irrelevant. We did not say it would never be applicable but, rather, that it would rarely be applicable. Even in NSK's case, the analysis NSK submitted shows there would be a large proportion of CV comparisons where we would have to use an alternative CV- profit calculation under NSK's construction of the statute. NSK's construction would, therefore, create an additional layer of complexity and uncertainty to antidumping proceedings without generating more accurate results. Finally, with regard to the argument of the SKF companies that we should, at a minimum, include the total revenues respondents received on unprofitable sales in our CV-profit calculation, as we have stated in prior reviews (see, e.g., AFBs 9, 64 FR at 35610), section 773(e)(2)(A) of the Act requires us to use the actual amount for profit in connection with the production and sale of a foreign like product in the ordinary course of trade. Section 771(15) of the Act defines sales outside the ordinary course of trade as those sales disregarded under section 773(b)(1) of the Act because they failed the cost test. Thus, as required by law, the Department has continued to exclude sales that failed the cost test from the CV-profit calculation under section 773(e)(2)(A) of the Act. Including the values of these sales in the denominator of our profit calculation would be the same as not disregarding the sales, even if we were to set the profit level of such sales to zero. Thus, we have not followed the proposal of the SKF companies as it would be contrary to section 773(e)(2)(A) of the Act. Comment 58: As detailed in Comment 54 above, Torrington Nadellager contends that the CV profit the Department used to calculate normal value for models it sold in the United States to "large" OEMs is based on inadvertently and inaccurately reported home-market data. To calculate CV profit properly, Torrington Nadellager asserts that the Department should re-categorize the profitable sales in the "large" OEM channel of distribution as a different level of trade; alternatively, the respondent requests that the Department combine all home-market levels of trade for the CV-profit calculation and use the combined rate to derive the CV for of all home-market levels of trade. Torrington Nadellager asserts that, since neither the statute nor the regulations require the Department to calculate a level-of-trade-specific profit for CV, the Department can depart from its normal practice and apply an alternative methodology when the results are irrational. Torrington Nadellager asserts that a level-of-trade-specific profit promotes precise calculations; however, it contends that applying the methodology under the factual scenario at hand yields irrational results. In support of this argument the respondent cites the Department's Final Rule, where the Department stated that ". . . the sales used as the basis for CV profit should not lead to irrational or unrepresentative results." Final Rule, 62 FR at 27360. In support of its proposed change in methodology, the respondent argues that the number of above-cost transactions used to calculate CV profit for the level of trade at issue is de minimis, the transactions are unrepresentative of other sales it reported for the same level of trade, and the profits are excessive under the circumstances presented. Torrington Nadellager concludes that a combined CV-profit rate for all of its home-market levels of trade would satisfy the Department's calculation requirements, eliminate irrational results, and more closely conform to the methodology applied to it in past reviews (i.e., a single profit rate from its financial statements). FAG Germany requests that the Department not revise the CV-profit calculation for Torrington Nadellager. It argues that high profits or a de minimis database do not render sales transactions unrepresentative. With regard to including high-profit sales in the calculation of profit for CV, FAG Germany cites AFBs 6, 62 FR at 2114, asserting that, in that case, the Department held that there is no statutory provision or SAA reference requiring a profit "cap." In support of using a de minimis database for the calculation of CV profit, FAG cites AFBs 8 at 63 FR at 33334-5, where the Department found that the profit calculation for a respondent was authorized under the statute although it was based on a small number of highly profitable home-market sales. FAG Germany concludes that the Department should reject Torrington Nadellager's arguments since it rejected similar arguments by respondents in other cases based on the same factual scenario. SKF Germany argues that the Department should not calculate the combined profit rate as requested by Torrington Nadellager. It contends that the existence of high-profit sales for a small number of transactions does not justify altering the CV-profit methodology and comments that the Department has applied a level-of-trade-specific profit calculation methodology during all post-URAA reviews. Citing to the arguments on page 5 of Torrington's June 27, 1997, General Issue Case Brief for AFBs 7, SKF Germany asserts that, without evidence that specific sales which pass the cost test are outside the ordinary course of trade, high profits on such transactions are not necessarily abnormal or irrational. Finally, SKF Germany argues that since Torrington Nadellager was aware of its pricing policies it should have known that there would be only a small number of above-cost sales in the level of trade at issue. INA asserts that the methodology applied to Torrington Nadellager for the calculation of CV profit is consistent with the Department's practice and is fair. The respondent contends that it would indeed be unfair if the Department were to depart from its normal methodology by making the change requested by Torrington Nadellager. Department's Position: For the reasons provided in response to Comment 54 above, for the final results we have not re-categorized Torrington Nadellager's profitable sales in the "large" OEM channel of distribution to a different level of trade. Moreover, for the reasons that follow, we have not departed from our normal practice by altering the methodology we used to derive Torrington Nadellager's CV profit. Where we base normal value on CV, the CV figure is intended to approximate what a price-based normal value would be if there were usable sales. Thus, for Torrington Nadellager, we calculated CV by level of trade using the selling expenses and profit determined for each level of trade in the comparison market. As noted by all of the parties commenting on this issue, including Torrington Nadellager, this methodology is consistent with our practice. See, e.g., Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Preliminary Results of Antidumping Duty Administrative Reviews and Partial Rescission of Administrative Reviews, 64 FR 8790, 8795 (February 23, 1999), and Certain Welded Carbon Steel Pipes and Tubes From Thailand: Preliminary Results of Antidumping Duty Administrative Review, 65 FR 18301, 18303 (April 7, 2000). As discussed in the preamble of the final antidumping regulations, the Department affirms that the sales used as the basis for CV profit should not lead to irrational results. See Final Rule, 62 FR at 27360. In support of its assertion that the above-cost sales to "large" OEMs yield irrational results, Torrington Nadellager does not claim that the transactions are outside the ordinary course of trade. Instead, it argues that these above-cost sales are unrepresentative of other sales to "large" OEMs. Specifically, Torrington Nadellager provides the ratio of above-cost transactions to total transactions in the level of trade and asserts that the above-cost sales constitute a de minimis database. It also identifies allegedly atypical facts surrounding the sales and claims that the profits on the transactions are excessive under the circumstances presented. First, we disagree with Torrington Nadellager's claim that the above-cost sales in the "large" OEM channel of distribution constitute a de minimis database. In the preamble to our regulations, we addressed this situation. Specifically, we stated that: ... Where the Department finds a majority of sales of a foreign like product to be at below-cost prices (and, thus, excludes those sales from the calculation of profit), the fact that only a few sales remain at above-cost prices does not, by itself, render such sales outside the ordinary course of trade. Rather, it is the below-cost sales that are outside the ordinary course of trade. Whether the few remaining above-cost sales are also outside the ordinary course of trade is a separate issue that depends on the facts and circumstances surrounding these transactions. See Final Rule, 62 FR at 27360. With the exception of the transactions that Torrington Nadellager would like us to reclassify, all other sales in the level of trade at issue were determined to be outside the ordinary course of trade because they were sold below cost. Lacking evidence that the small portion of above-cost sales should be considered outside the ordinary course of trade, we do not find them necessarily unrepresentative because they are few in number. We also disagree with Torrington Nadellager's allegation that the transactions at issue involve atypical facts. The reasons we relied on in deciding not to re-categorize the transactions to a different level of trade support this position. (See our response to Comment 54, above.) Overall, we found that the record evidence indicates that the characteristics of the sales are typical in the context of the designated channel of distribution. For example, several of the transactions that Torrington Nadellager wants reclassified are non-zero-priced sample transactions. As detailed in the Department's response to Comment 54 above, the record evidence supports that these transactions are part of the normal process for selling to "large" OEMs. Moreover, since Torrington Nadellager did not demonstrate that the sample transactions have "unique or unusual characteristics" in comparison to other home-market sales, we have continued to use them for the calculation of CV profit. High profits alone do not render sales transactions unrepresentative. As we stated in AFBs 8, 63 FR at 33335, with respect to SNFA U.S., "in deciding whether certain sales used as the basis for CV profit lead to irrational and unrepresentative results, we must consider the specific facts and circumstances surrounding the transactions." Similarly, in AFBs 8, 63 FR at 33344, citing CEMEX, S.A. v. United States, 133 F.3d at 900 (CAFC 1998) (CEMEX), we addressed the issue of unrepresentative home- market sales in terms of whether transactions with high profits are outside the ordinary course of trade. Specifically, we stated that: The presence of profits higher than those of numerous other sales does not necessarily place the sales outside the ordinary course of trade. In order to determine that a sale is outside the ordinary course of trade due to abnormally high profits, there must be unique and unusual characteristics related to the sale in question which would make it unrepresentative of the home market. The SAA, at 840, states that "under section 773(e)(2)(A), in most cases Commerce would use profitable sales as the basis for calculating profit for purposes of constructed value." The SAA, at 839-840, also identifies sales with aberrant high profits as an example of the type of transactions that the Department could consider outside the ordinary course of trade (along with off-quality merchandise and sales to related parties at non- arm's-length prices). However, in the preamble to our Final Rule, we specified that the proof that the profits a respondent earned on specific sales are abnormal will depend on a number of factors. These factors include the type of merchandise under review and the normal business practices of the respondent and of the industry in which the merchandise is sold. See Final Rule, 62 FR at 27360. Here, Torrington Nadellager did not provide argument or evidence of "unique or unusual characteristics" to fulfill the evidentiary burden necessary to establish the transactions as outside the ordinary course of trade. Lacking evidence that we should consider the above-cost sales to "large" OEMs outside the ordinary course of trade, we do not find them unrepresentative. Finally, we also disagree with Torrington Nadellager's assertion that we should combine all home-market levels of trade for the CV-profit calculation because it would more closely conform to the methodology applied in past reviews (i.e., a single profit rate from its financial statements). In past reviews, we used an alternative methodology for the calculation of CV profit pursuant to section 773(e)(2)(B) of the Act because we could not calculate profit in accordance with the preferred method provided under section 773(e)(2)(A) of the Act based on the information on the record. However, for the current segment of the proceeding, because we conducted an investigation of sales below cost, we had the necessary information on the record to calculate profit in accordance with the preferred section 773(e)(2)(A) of the Act. Therefore, in accordance with this section of the Act, we based Torrington Nadellager's SG&A expenses and profit on the amounts incurred and realized in connection with the production and sale of the foreign like product in the ordinary course of trade for consumption in the home market. This is the most appropriate methodology to apply under the statute and is in accordance with Department practice. Comment 59: Takeshita argues that the CV-profit rate the Department used for BBs in the preliminary results is an unreasonable value. Citing the financial statements of several Japanese bearing manufacturers, Takeshita states that the CV-profit rate the Department used is too high because no other Japanese bearing manufacturer had such a high profit rate. Furthermore, Takeshita alleges that the CV-profit rate the Department calculated is the result of using the incorrect value in the formula. Takeshita also questions whether the total COP the Department used in the calculation of the CV-profit rate is the total for all eleven Japanese manufacturers. Department's Position: Under our standard methodology for calculating CV profit, we calculate CV profit on a class-or-kind-specific basis. The Japanese bearing manufacturers' financial statements which Takeshita cites are based on the bearing manufacturers' total operations and are not based solely on the production of BBs. Furthermore, we disregarded sales of BBs for several of the Japanese respondents because we found them to be made at prices below the cost of production. The financial statements upon which Takeshita relies include such below-cost sales. However, we discovered an inadvertent error upon re-examining our calculation methodology for CV profit to confirm that we used the correct value in our calculation and accounted for all eleven companies under review. We inadvertently did not include the total home-market value for one company. Therefore, we adjusted the total home-market value to account for all eleven companies and revised our CV-profit rate calculation for these final results. This resulted in a higher CV-profit rate in our calculation for Takeshita. 8.B. Affiliated-Party Inputs Comment 60: NTN Japan argues that the Department should not adjust the costs for affiliated-party inputs by using the higher of the market price or the affiliate's COP when these are higher than the transfer price. NTN Japan claims that the Department offered no explanation or statutory support for not using the transfer prices it reported. NTN Japan argues that there is no evidence that the transfer prices of the affiliated-party inputs did not fairly reflect the amount usually reflected in the sales of merchandise under consideration. It also argues that the Department did not have reasonable grounds to believe the inputs were sold at less than the COP. Torrington contends that the Department has rejected this argument in prior reviews and that NTN Japan did not explain how the evidence differs in this review. Department's Position: Pursuant to section 773(f)(3) of the Act, in the case of a transaction between affiliated persons involving the production of a major input, the Department may consider whether the amount represented as the value of the major input is less than its COP. In addition, 19 CFR 351.407 states that, for purposes of section 773(f)(3) of the Act, the value of a major input purchased from an affiliated person will be based on the higher of (1) the price paid by the exporter or producer to the affiliated person for the major input, (2) the amount usually reflected in sales of the major input in the market under consideration, or (3) the cost to the affiliated person of producing the major input. We have relied upon this methodology in past AFB reviews as well as in other cases. See, e.g., AFBs 6, 62 FR at 2117, AFBs 7, 62 FR at 54065, AFBs 8, 63 FR at 33337, AFBs 9, 64 FR at 35612, and Round Wire from Taiwan. NTN Japan argues that the Department must have reasonable grounds to believe that inputs are being sold at less than COP before it may use COP information. The Department considers the initiation of a cost investigation concerning home-market sales a specific and objective reason to believe or suspect that the transfer price from a related party for any element of value may be below the related suppliers' COP. See 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; Final Results of Antidumping Duty Administrative Reviews and Termination in Part, 63 FR 20585 (April 27, 1998). This practice was affirmed by the CIT in NSK Ltd. v. United States, 910 F. Supp. 663 (CIT 1995). Therefore, based upon prior case precedent, it was appropriate to consider the cost data available on the record in determining how to value major inputs. Comment 61: Torrington argues that the Department should use facts available for certain major inputs that SKF France obtained from affiliated parties and for which it did not provide market prices. Torrington claims that, despite the Department's specific reporting instructions in its original and supplemental questionnaires, SKF France only reported the higher of a) the transfer price from the affiliate, or b) the affiliate's COP, and it did not take into consideration the market price of certain inputs. SKF France asserts that, contrary to Torrington's claim, it has ensured that it used the price from the unaffiliated supplier in its cost calculations if that price was greater than the higher of COP or transfer price. SKF France points to Appendix D-14 of its supplemental response as evidence of its reporting and argues that, since it has complied fully with the Department's questionnaire, the Department should continue to accept its cost reporting. Department's Position: In a supplemental questionnaire, we asked SKF France about certain major inputs that it purchased from both affiliated and unaffiliated suppliers. Specifically, we asked that the respondent provide a chart listing the cost, transfer price, and market-price data for these inputs on a comparable basis. SKF France provided this information at Appendix D-14 of its supplemental questionnaire response and stated in its narrative that, in cases where it purchased the identical input from an unaffiliated source, "SKF has ensured that the price from the unaffiliated supplier was used in its cost calculations if it was greater than the higher of cost or transfer price." See SKF's supplemental questionnaire response dated October 15, 1999, at page 40 and Appendix D-14. Since SKF France has provided this information in compliance with our supplemental questionnaire, the use of facts available is not warranted in this case. Therefore, we have not made any changes to SKF France's valuation of major inputs for these final results. Comment 62: Torrington argues that the Department's questionnaire instructed the respondents to provide detailed and specific comparison data for major inputs purchased from affiliates, including transfer prices, market prices, or, alternatively, COP data for components. Torrington states that NMB/Pelmec did not comply with these instructions for a number of major inputs. Torrington claims that, in the case of rings purchased from affiliated suppliers, NMB/Pelmec reported transfer prices only. In the case of balls, Torrington claims that NMB/Pelmec reported the affiliate's COP data only when the model had been in production during the POR while, for models last produced prior to the POR, Torrington claims that NMB/Pelmec valued the input of these other components at the transfer price. Torrington argues that NMB/Pelmec's explanations for not reporting the information requested is insufficient. In the case of rings, Torrington argues that there is no evidence to support NMB/Pelmec's explanation that, because unfinished inner and outer rings were at various states of completion and were further processed by NMB/Pelmec, "it is impracticable (if not impossible) to make comparison of the varying COPs of such rings and their transfer prices." Torrington also argues that NMB/Pelmec provided no explanation for not reporting the affiliated parties' COP where the product was produced prior to the POR. Citing AFBs 9, 64 FR at 35599, Torrington asserts that the Department, in a similar situation involving SKF France, applied partial facts available to SKF France's purchases of major inputs from affiliated suppliers. Torrington argues that, in this case, the Department should also apply partial facts available to NMB/Pelmec's reported transfer prices for inner and outer rings and that the Department should calculate an appropriate factor by which to increase NMB/Pelmec's reported costs. NMB/Pelmec responds that there is no justification for use of facts available. NMB/Pelmec argues that, with the exception of unfinished rings and pre-POR inputs, it reported COP for the major inputs obtained from affiliates for production during the POR. NMB/Pelmec states that, under the circumstances and within the time limits of this administrative review, it was unable to report the COP information for the inputs at issue and that, according to the Department's precedent, when respondents act to the best of their ability in attempting to obtain cost information pursuant to the major-input rule, the Department does not resort to adverse facts available as proposed by Torrington. NMB/Pelmec argues that, under the circumstances of this case, the Department should use the information reported by NMB/Pelmec as the most reasonable "gap-filling" data in those exceptional cases where it was unable to report COP. NMB/Pelmec argues that market prices do not exist for many of NMB/Pelmec's components because NMB/Pelmec's affiliates do not sell such components to outside parties and NMB/Pelmec did not purchase such components from unaffiliated parties. In instances where NMB/Pelmec did purchase from unaffiliated suppliers, NMB/Pelmec states that it reported market prices. NMB/Pelmec also states that it made extensive efforts to satisfy the Department's affiliated-party-input reporting requirements, including deviating from its normal cost-accounting system exclusively for purposes of responding to the Department's affiliated-party-input questions. Instead of using its standard method, NMB/Pelmec explains that it reported the higher of market price, COP, or transfer price, where possible, for over one hundred types of parts and components. NMB/Pelmec states that this was an extraordinarily complicated process because it purchases these parts from affiliated parties in numerous countries. Also, with the exception of unfinished or semi-finished inner and outer rings, for which NMB/Pelmec noted in its September 3, 1999, questionnaire response that it was unable to obtain the COP information in time, NMB/Pelmec states that it provided worksheets comparing the COP and transfer price for each component part used in the CV and COP listings for BBs. NMB/Pelmec states that it made clear in its response that it purchases a minimal number of rings from affiliates, whether finished or unfinished, and claims that Torrington is seeking to inflate an issue of negligible impact in order to impugn NMB/Pelmec's overall reporting methodology for all affiliated-party inputs. Department's Position: Pursuant to section 773(f)(3) of the Act, in the case of a transaction between affiliated persons involving the production of a major input, the Department may consider whether the amount represented as the value of the major input is less than its COP. In addition, section 351.407 of the Department's regulations states that, for purposes of section 773(f)(3) of the Act, the value of a major input purchased from an affiliated person will be based on the higher of (1) the price paid by the exporter or producer to the affiliated person for the major input, (2) the amount usually reflected in sales of the major input in the market under consideration, or (3) the cost to the affiliated person of producing the major input. We have relied upon this methodology in past AFB reviews as well as in other cases. See, e.g., AFBs 6, 62 FR at 2117, AFBs 7, 62 FR at 54065, AFBs 8, 63 FR at 33337, AFBs 9, 64 FR at 35613, and Round Wire from Taiwan. In recent cases, when a respondent has been unable to obtain COP data for a major input from an affiliate and we have determined that the respondent has acted to the best of its ability to get that data, we have looked to the record and used non-adverse facts available. For example, in the case of a cost allegation, we might use the cost data supplied in the allegation if no other data is available on the record. In these situations, we do not draw an adverse inference but rather use this information as "gap-filling" facts available because it is the only information available on the record. See Notice of Final Determination of Sales at Less Than Fair Value: Certain Polyester Staple Fiber From the Republic of Korea, 65 FR at 16682 (March 30, 2000)(the Department's policy is clear that, when COP is not provided and the Department has determined that the respondent has acted to the best of its ability to obtain the information, the Department will look to the record to determine what information will replace the missing COP information). In cases where we do not have COP data on the record, either from the respondents or from the petitioner, we have used the higher of the transfer price or the market price because there was no indication on the record that the COP would be higher than the transfer or market price. With respect to NMB/Pelmec, the record for this review includes cost data for most major inputs it purchased. In its September 3, 1999, Section D Response at 25, NMB/Pelmec stated that, with respect to BB components, market prices do not exist because NMB/Pelmec's affiliates do not sell such components to outside parties and NMB/Pelmec does not purchase such components from unaffiliated parties. For components that were also purchased from unaffiliated suppliers, NMB/Pelmec reported market prices. For COP information on inputs from affiliates, NMB/Pelmec provided information for the majority of inputs. In the case of unfinished inner and outer rings purchased from affiliates, NMB/Pelmec stated that these parts were at various stages of completion when transferred and were further processed by NMB/Pelmec. NMB/Pelmec stated that, as a result, it was impracticable to make comparison of the varying COPs of such rings and their transfer prices, and therefore it used the transfer price in reporting costs for these parts. These parts accounted for an insignificant number of total parts NMB/Pelmec used which it purchased from affiliates. See NMB/Pelmec's Section D response dated September 3, 1999, at Attachment D-2. NMB/Pelmec also stated that, among major inputs, the overwhelming majority it purchased from affiliates are products other than rings and it provided support for this statement in its response. See NMB/Pelmec's Section D response dated September 3, 1999, at Attachment D-2. Given the insignificant number of parts for which information is missing and because we did not pursue this information further, the use of facts available is not justified in this situation. We do not agree with Torrington that our use of facts available for SKF France in AFBs 9 is similar to this situation. In that case, we specifically requested additional information pertaining to market prices by unaffiliated parties but found that the information SKF France provided in response to the request did not allow us to determine whether the market prices were higher than the reported COP or transfer prices. In this case, NMB/Pelmec explained in its September 3, 1999, Section D response that it purchased unfinished inner and outer rings that were at various stages of completion from some affiliates and that market prices did not exist. NMB/Pelmec also stated that the unfinished rings are further processed by NMB/Pelmec and that a comparison of the varying costs of such rings and their transfer prices was impracticable, so it used transfer prices to value such rings. Although we asked NMB/Pelmec to clarify its response with respect to instances in which it used market prices for valuing inputs in our supplemental questionnaire, we did not ask for additional information with respect to its methodology for valuing rings because these parts accounted for an insignificant number of parts. Therefore, we find that transfer prices are an acceptable "gap-filling" measure for the few components for which COP data was not provided and have used these prices for our final results. 8.C. General, Selling, and Administrative Expenses Comment 63: Torrington argues that, based on Nachi's income statements, the Department should revise Nachi's reported COP and CV by including losses on the disposal of fixed assets and the retirement of fixed assets. Nachi argues that it categorizes its losses as a result of the disposal of fixed assets and the retirement of fixed assets as extraordinary losses and, therefore, they are not part of the company's current-period operating costs. Nachi asserts, that, if the Department decides to include those amounts in COP, it should use Nachi's fiscal year as this period most closely approximates the POR. Department's Position: We regard losses as a result of the disposal of fixed assets and retirement of fixed assets as a normal cost of production (see AFBs 6, 62 FR at 2118). Based on our review of Nachi's response, we have found no evidence to suggest that Nachi's losses were unrelated to the general production activity of Nachi overall. Therefore, we have included these losses as a general expense and recalculated Nachi's reported COP and CV accordingly. Comment 64: Torrington argues that certain residual expenses, which SKF Germany incurred in closing a factory which produced non-subject merchandise, should be classified as restructuring costs and included in SKF Germany's COP/CV calculation under G&A expenses. SKF Germany argues that, because these expenses specifically and solely relate to former employees who worked at a now-closed factory making non- subject merchandise, it would be improper to categorize them as and include them in its reported G&A expenses. Department's Position: The costs incurred from the closure of a plant, despite whether that plant produced merchandise subject to the order, are a general cost of doing business. Given that the costs in question are general in nature, they should be captured under the general and administrative expenses (G&A) reported by the respondent. Accordingly, we have revised SKF Germany's reported GS&A factor to include its costs related to factory closure. See Brass Sheet and Strip From Canada; Final Results of Antidumping Duty Administrative Review, 60 FR 49582, 49583 (September 26, 1995). Comment 65: Torrington asserts that SKF Germany did not report certain accrual amounts of its small non-manufacturing related extraordinary expense and points out that SKF Germany did not provide an adequate description of these expenses nor an explanation of why such expenses should be excluded from its reported costs. SKF Germany contends that, as the expenses in question were overstated accrual costs, it was proper to exclude them from the reported costs. However, SKF Germany points out that it included in its reported costs the actual expenses related to the expenses in question. Department's Position: Based on the record evidence that SKF Germany submitted in its supplemental questionnaire response (dated November 12, 1999, at page 2) and in its rebuttal brief (dated May 15, 2000, at page 14), we are satisfied that SKF Germany reported the small non- manufacturing related expenses in question. Comment 66: Torrington asserts that SKF Germany did not report certain accrual amounts of its small, non- manufacturing-related extraordinary expenses and points out that SKF Germany did not provide an adequate description of these expenses nor an explanation of why such expenses should be excluded from its reported costs. SKF Germany contends that, as the expenses in question were overstated accrual costs, it was proper to exclude them from the reported costs. SKF Germany states that it did include the actual expenses related to the expenses in question in its reported costs. Department's Position: Based on the evidence that SKF Germany submitted in its supplemental questionnaire response (dated November 12, 1999, at page 2) and in its rebuttal brief (dated May 15, 2000, at page 14), we are satisfied that SKF Germany captured the actual expenses in question rather than of an accrual amount which SKF Germany knew to be overstated at the time it prepared its response. 8.D. When to Use CV Comment 67: NTN Japan argues that the Department must use CV rather than attempt to find a family match when it disregards contemporaneous sales of identical merchandise because they fail the cost test. NTN Japan alleges that the Department's current practice is based on an erroneous reading of the determination in CEMEX. NTN Japan contends that the statute directs that the identification of the foreign like product to match to a U.S. sale must occur prior to the determination of whether the sales of the foreign like product are outside the ordinary course of trade. According to NTN Japan, section 773(b)(1) of the Act directs further that, once the foreign like product has been identified, the normal value for those sales which are disregarded as below-cost sales is to be based on CV. Torrington contends that the Department has rejected this argument in prior reviews and that NTN Japan has not demonstrated why the CEMEX decision should not be followed. Department's Position: The Court stated in CEMEX that "[t]he language of the statute requires Commerce to base foreign market value on nonidentical but similar merchandise..., rather than constructed value when sales of identical merchandise have been found to be outside the ordinary course of trade." See CEMEX, F.3d at 904. Pursuant to section 771(15)(A) of the Act, below-cost sales in substantial quantities and within an extended period of time are outside the ordinary course of trade. Therefore, in order to be consistent with CEMEX, when making comparisons in accordance with section 771(16) of the Act, we considered all products sold in the home market that were identical or similar to merchandise within the scope of each order and which were sold in the ordinary course of trade for purposes of determining appropriate product comparisons to U.S. sales. Where there were no sales of identical merchandise in the home market made in the ordinary course of trade to compare to U.S. sales, we compared U.S. sales to sales of the most similar merchandise made in the ordinary course of trade. Only where there where no home-market sales of comparable merchandise in the ordinary course of trade did we resort to CV. 8.E. Inventory Write-offs Comment 68: Torrington argues that NTN Japan's response to the Department's question about inventory write-offs was so vague that the Department should deem it a non-answer which does not allay its concerns that the write-offs may not be captured in NTN's reported costs. Torrington contends that the Department should apply facts available by increasing NTN Japan's costs using the largest write-offs reported by any Japanese respondent. NTN Japan contends that it answered the Department's question on the matter and that the Department did not show dissatisfaction with NTN Japan's response. Thus, NTN Japan argues, the use of facts available is not necessary. Department's Position: Contrary to Torrington's assertion, it is clear from NTN Japan's response where the respondent included write-offs in its reported costs, and we are satisfied that they are captured in the reported costs. Because of the proprietary nature of this information, please see the NTN Japan final results analysis memorandum dated July 28, 2000, for a description of where NTN Japan included the costs of inventory write-offs. Thus, no change to NTN Japan's reported costs is necessary. Comment 69: Torrington argues that the Department should revise NSK's reported costs to include all inventory write-offs. Torrington contends that the Department normally includes inventory write-offs as a normal cost of production. NSK argues that the Department has rejected Torrington's argument in a prior review. Department's Position: We determined that NSK accounted properly for inventory write-offs in its reported costs in the prior review. See AFBs 9, 64 FR at 35615. NSK did not change its practice regarding inventory write-offs for this review. See NSK's Section D questionnaire response dated September 3, 1999, in which NSK explains that the only changes from the prior review concerned the useful life of buildings and depreciation methodology. No adjustment is necessary. 8.F. Allowance for Doubtful Accounts Comment 70: Torrington argues that the Department should ensure that all allowances for doubtful accounts are included in NTN Japan's COP and CV costs. Torrington claims that NTN Japan's response to the Department's question about the impact of a change in Japanese law concerning allowances for doubtful accounts on its reported costs is unclear. NTN Japan argues that Torrington's argument is based on a misunderstanding and that it included all expenses in its reported costs. Department's Position: NTN Japan's response about the impact that the change in Japanese law has on NTN Japan's reported costs was clear. Furthermore, NTN Japan's response was subject to verification. Based on our examination of the record, we find that it is not necessary to change NTN Japan's reported costs. Because of the proprietary nature of this information, please see the NTN final results analysis memorandum dated July 28, 2000, for a description of our analysis. 8.G. Marketable Securities Comment 71: Torrington argues that the Department should restate NTN Japan's COP and CV to include losses on the revaluation of marketable securities. Torrington contends that such losses are not, as NTN Japan claims, extraordinary expenses, but are normal and foreseeable. Thus, Torrington concludes, such losses are part of fully loaded costs under Department practice and rules. NTN Japan contends that its treatment of losses on the revaluation of marketable securities is in accordance with generally accepted accounting principles in Japan. NTN Japan claims further that the Department was fully satisfied with NTN Japan's responses on the matter and did not ask for further information regarding these expenses. Department's Position: We find that NTN Japan excluded its losses on the revaluation of marketable securities from its reported costs properly. First, such losses are manifestly not a part of COM. Second, we do not include losses (or gains) on the revaluation of marketable securities as a part of G&A expenses because such expenses are related to investment activities which are not associated with the core business of NTN Japan. This practice has been upheld by the CIT. In U.S. Steel Group a Unit of USX Corporation, USS/Kobe Steel Co., and Koppel Steel Corp. v. United States, 998 F. Supp. 1151 (CIT February 25, 1998), the CIT upheld our remand determination, in which we stated "where, . . . items of income and expense are most closely related to the general operations of the company (all general activities associated with the company's core business), it is appropriate to treat those items as part of G&A." The question as to whether such losses are extraordinary, in this instance, is irrelevant. Therefore, we have not recalculated NTN Japan's reported costs. 9. Packing and Movement Expenses Comment 72: The petitioner argues that the Department should disallow an adjustment for FAG Germany's home-market packing expenses because it did not provide the Department with requested information. According to the petitioner, the Department in its supplemental questionnaire directed FAG Germany to create an additional data field to re-state the expenses in question and to calculate new amounts by multiplying units by packing time. In addition, according to the petitioner, the Department further directed FAG Germany to justify its new methodology which the company used for the first time in this review. The petitioner asserts that, since FAG Germany did not provide any of the requested information, the Department should, as facts available, disallow any adjustment for home-market packing expense. FAG Germany argues that its packing-labor calculation conforms with the Department's practice and preference. It asserts that, as preferred by the Department, it calculated a packing factor based on total actual material and labor costs allocated over total weight. In addition, FAG Germany contends that, at verification, the Department verified its packing methodology fully and found no discrepancies. According to FAG Germany, against this background it is clear that its packing-cost methodology was proper and that the petitioner's claim is without merit. Furthermore, according to the respondent, the supplemental question to which the petitioner refers in its case brief bears no relation to the packing-cost methodology FAG Germany actually used in this review period. Therefore, it asserts, it did in fact respond adequately to the Department's supplemental question regarding packing cost given that the question was inapplicable to this review period and that the petitioner's position that FAG Germany was unresponsive is simply not valid in light of this fact. Department's Position: Based on the record, we have determined that FAG Germany's packing-cost methodology is reasonable. Our regulations provide that we may consider allocated expenses and price adjustments when transaction-specific reporting is not feasible if we are satisfied that the allocation method used does not cause inaccuracies or distortions. See 19 CFR 351.401(g). We find that FAG Germany's packing methodology, in which it calculated its packing factor based on total actual material and labor costs, is reasonable because it reflects the basis on which costs were incurred. In addition, we verified FAG Germany's packing cost methodology and found no discrepancies. See FAG Germany December 3, 1999, Verification Report at page 10. Therefore, we find no reason to disallow an adjustment for FAG Germany's home-market packing expenses. Comment 73: Torrington argues that, although SKF Sweden reported air and ocean freight separately, it did not report transaction-specific air- freight expenses, as required by the Department's instructions. The petitioner asserts that the Department should reject SKF Sweden's reporting methodology because air freight is generally more expensive than ocean freight and apply a facts-available rate for SKF Sweden's international freight on all U.S. sales. It claims that it is inexcusable in the tenth administrative review to use alleged inconvenience in reporting as a reason not to report these freight expenses on a transaction-specific basis, particularly when other respondents have been able to report air-freight expenses as such. SKF Sweden contends that there is no justification to apply facts available for SKF Sweden's reporting of international freight. SKF Sweden states that it reported ocean freight and air freight expenses separately in its sales files, which, it argues, does not reflect business reality. However, SKF Sweden states that it is unable to report, or bill, international freight expenses on a transaction-specific basis because they are not incurred as such. SKF Sweden rebuts the petitioner's assertion that SKF Sweden did not report these expenses because of inconvenience; rather, SKF Sweden contends, transaction-specific data were not available. The respondent also states that the fact that other respondents may be able to report these expenses on a transaction-specific basis is irrelevant to the situation of SKF Sweden. Department's Position: We have found in previous administrative reviews that it is sometimes not feasible for the respondents to report air and ocean freight on a transaction-specific basis in these proceedings where they do not incur these expenses on a transaction-specific basis in their normal course of business. See AFBs 8, 63 FR at 33340, and AFBs 7, 62 FR at 54081. This is consistent with our regulations, which provide that we may consider allocated expenses and price adjustments when transaction- specific reporting is not feasible and we are satisfied that the allocation method used does not cause inaccuracies or distortions. See 19 CFR 351.401(g). Also, in the preamble to the Department's regulations, in discussing the allocation of expenses and price adjustments, the Department agreed with commenters that "allocated expenses or price adjustments may not be as exact as expenses or prices adjustments reported on a transaction-specific basis," but it added, "in our view, the drafters of the URAA and SAA could not have intended that all allocations are inherently distortive or inaccurate for purposes of the antidumping law." See Final Rule, 62 FR at 27346. The SAA supports this position, stating that "[t]he Administration does not intend to change Commerce's current practice, sustained by the courts, of allowing companies to allocate these expenses when transaction specific reporting is not feasible, provided that the allocation method used does not cause inaccuracies or distortions." See SAA at 823-824. In SKF Sweden's questionnaire response, it indicated that its business records do not link specific freight expenses with specific U.S. sales to unaffiliated customers, and during verification we did not find evidence that it could make such a link. See SKF Sweden's Section C response to the questionnaire at 49 and our March 21, 2000, verification report. Moreover, during the verification we did not find any inaccuracies in SKF Sweden's reporting of international freight. The respondent used an aggregate reporting methodology that reflects the manner in which it incurs international freight expenses accurately. Specifically, it sampled data covering the entire POR by totaling monthly expenses and associated shipping weights which it used to calculate an ocean-freight factor and an air-freight factor based on ocean-freight and air-freight invoices. It then applied rates to the per-unit shipping weights to yield the reported amount for international freight. It would be inappropriate to reject these allocations without evidence that they are unreasonably distortive. Therefore, we have accepted SKF Sweden's data concerning these expenses. Comment 74: Torrington argues that, although NMB/Pelmec incurred separate charges for air freight and ocean freight, it reported only aggregate freight expenses for international freight and did not explain why these expenses could not be reported separately and applied to appropriate sales or groups of sales. Torrington asserts that the air-freight expenses were very likely incurred pursuant to long-term requirements contracts. Torrington argues that the Department should require NMB/Pelmec to report separate air-freight and ocean-freight expenses prior to the final results or apply an appropriate facts-available rate for NMB/Pelmec's international freight expenses, given that NMB/Pelmec did not report separate freight expenses. The respondent argues that Torrington's claim is without merit, that the Department verified NMB/Pelmec's reporting of both types of freight and found that freight could not be reported on a transaction-specific basis, and that the methodology used was reasonable and accurate. Department's Position: We agree with the respondent that there is no reasonable way for it to trace a specific re-sale in the United States to a particular method of transport and we have commented on this issue in previous reviews. See, e.g., AFBs 7, 62 FR at 54081, AFBs 8, 63 FR at 33340, and AFBs 9, 64 FR at 35616. During verification, we spent considerable time in examining sales-trace documents for shipments and payments to vendors who provided international transport, and we discussed with company officials how merchandise enters into the inventory of NMB/Pelmec's U.S. affiliate. Our regulations state specifically that we may consider allocated expenses when transaction-specific reporting is not feasible, provided we are satisfied the allocation method does not cause inaccuracies or distortions. See 19 CFR 351.401(g). The preamble to the regulations, in discussing factors to examine in assessing distortion or inaccuracies, comments on the issue of out-of-scope merchandise in the calculation of allocated expenses and states that even the inclusion of out-of-scope merchandise is not inherently distortive. See Final Rule, 62 FR at 27348. In the case of NMB/Pelmec, in its calculation of international freight, shipment-quantity figures only included in-scope merchandise. See NMB/Pelmec's supplemental questionnaire response dated November 17, 1999, at 13. Accordingly, we are satisfied that the allocation method the respondent used is reasonable and not distortive. Comment 75: Torrington argues that the Department should not accept Koyo's combined reporting of air-freight expenses and ocean-freight expenses for shipments to the United States. Torrington argues further that, given the relative cost of air freight versus ocean freight, the Department should require that Koyo provide separate amounts for the aggregated ocean-freight and air-freight costs reported in its questionnaire response. Finally, Torrington contends that Koyo should revise its response to add an allocation for the air-freight expenses to a subset of sales where they are most likely incurred. In rebuttal, Koyo states that the Department should reject Torrington's argument and accept its international-freight factor as reported. Koyo argues that is has used the same methodology in this review, by which it calculates a single factor based on the total of ocean-freight and air- freight expenses, as in all prior AFB reviews. Koyo contends further that the Department uniformly accepted its freight calculations in prior reviews, finding that there was nothing in the record to lead to the conclusion that Koyo could have linked air-freight shipments to specific sales, and the Department's acceptance of Koyo's international-freight calculation was affirmed by the CIT in NSK Ltd. v. United States, 995 F. Supp. 123 (CIT 1995). Finally, Koyo argues that Torrington has raised no new arguments or pointed to any new facts in this review that should lead the Department to reject this well-established practice. Department's Position: We find no new information on the record that would indicate that Koyo has changed the manner in which it records these expenses in its accounting system and is now able to determine a direct link between a sale and an air shipment. We have discussed this issue extensively in previous reviews. See AFBs 4, 60 FR at 10942, AFBs 5, 61 FR at 66510, AFBs 6, 62 FR at 2121, AFBs 8, 63 FR at 33340, and AFBs 9, 64 FR at 35617. Because of the standard ways these firms maintain their records, we have found that it is generally not possible for respondents to report air freight and ocean freight on a transaction-specific basis in these proceedings. Therefore, we have accepted Koyo's reporting of these movement expenses for the final results. Comment 76: The petitioner argues that NMB/Pelmec combined expenses for U.S. movement expenses in one field instead of reporting them separately as the Department requested in the original questionnaire and in the supplemental questionnaire. Torrington asserts that NMB/Pelmec's excuses for combining these expenses ring hollow and that alleged inconvenience in reporting is not a valid excuse. Torrington recommends that the Department apply facts available to these three expenses. The respondent argues that the Department's questionnaire instructs respondents not to separate expenses that have been combined in a single fee paid to transport companies and that it reported its combined movement expenses correctly, which the Department verified. Department's Position: Our questionnaire's instructions clearly permit respondents to report certain movement expenses in a single field if they are combined in a single billing paid to a transport company. Moreover, 19 CFR 351.401(g) provides that we may consider allocated expenses and price adjustments when transaction-specific reporting is not feasible if we are satisfied that the allocation method used does not cause inaccuracies or distortions. We examined NMB/Pelmec's method of allocation in its response to the questionnaire and in more detail at verification. As we stated in response to Comment 74, the preamble to the regulations states that an important factor to consider in examining whether an allocation method is reasonable is the extent to which out-of-scope merchandise is included in the calculation. See Final Rule, 62 FR at 27348. In this case, NMB/Pelmec explained in its response that it included in the field of international freight the cost of inland freight from plant to port of exit because they were combined in one invoice. See NMB/Pelmec questionnaire response at page C-19. In addition, the respondent explained in its supplemental response that it only included in-scope merchandise in the shipment quantity for its calculation of international freight. See NMB/Pelmec's supplemental response dated November 17, 1999, at 13. Thus, we are satisfied that the allocations are on as specific a basis as is feasible and do not cause distortions, and we have accepted NMB's allocation methodology for this review. Comment 77: Torrington alleges that NTN Japan did not respond to the Department's request to demonstrate that its allocation methodology concerning home-market freight expenses was not distortive. Torrington contends that NTN Japan's claim that the Department has accepted its methodology in prior reviews is not responsive to the Department's request. Therefore, Torrington argues, the Department should restate NTN Japan's home-market freight expenses, relying on facts available or denying the adjustment, if necessary. NTN Japan argues that it fully and accurately responded to the Department's question by explaining how the expenses are incurred and that the Department accepted NTN Japan's methodology and resulting data at verification. NTN Japan also claims that the Department found NTN Japan's allocations to be non-distortive at verification. Department's Position: The regulations provide that we may consider allocated expenses and price adjustments when transaction-specific reporting is not feasible, provided we are satisfied that the allocation method used does not cause inaccuracies or distortions. See 19 CFR 351.401(g)(1). The regulations also provide that "any party seeking to report an expense or a price adjustment on an allocated basis must demonstrate to the Secretary's satisfaction that the allocation is calculated on as specific a basis as is feasible, and must explain why the allocation methodology used does not cause inaccuracies or distortions." See 19 CFR 351.401(g)(2). Moreover, with respect to home-market adjustments, in accordance with the CAFC's decision in Torrington Company v. United States, 82 F.3d 1039, 1047-51 (CAFC 1996), we do not treat improperly allocated home-market price adjustments as indirect selling expenses, but we instead disallow downward adjustments in their entirety. Finally, although 19 CFR 351.402(g)(4) directs that we will not reject an allocation method solely because the method includes expenses incurred with respect to sales of merchandise that does not constitute subject merchandise, the preamble to our regulations state that ". . . in the case of these types of allocation methods, it will be particularly important that a party claiming an adjustment provide the explanation required under paragraph (g)(2) as to why the allocation method used is not inaccurate or distortive." See Final Rule, 62 FR at 27348. NTN Japan did not report home-market freight expenses on the basis on which it incurred them, explaining that it was impossible for it to do so. See NTN Japan's supplemental questionnaire response dated November 12, 1999, at page 7. Moreover, NTN Japan included expenses incurred on both in- scope and out-of-scope merchandise in its allocation of these expenses. See worksheet 4 of exhibit B-4 of NTN Japan's questionnaire response dated September 3, 1999. In a supplemental questionnaire, we asked NTN Japan "[f]or all expenses ... which are reported as a percentage of price (e.g., freight, packing, warehousing), please indicate whether that is the basis on which you actually incurred them. If they are not incurred on that basis, please revise your response to reflect the basis on which the expenses are incurred or, if this is not possible, explain why such a recalculation is impossible and demonstrate that your allocation methodology is not distortive." See NTN Japan supplemental questionnaire dated October 21, 1999, at page 3. Although we had accepted NTN Japan's allocation methodology in prior reviews, we had never asked NTN Japan in a supplemental questionnaire to demonstrate that its allocation methodology was not distortive and believed in conducting earlier reviews that it would be improper to deny the adjustment without allowing NTN Japan to show that its methodology was not distortive. Therefore, we asked this question in these current reviews to alleviate our concern that NTN Japan's allocation methodology might shift expenses incurred on non- subject merchandise to sales of subject merchandise. NTN Japan did not attempt to demonstrate that its allocation methodology was not distortive. Rather, it merely stated that we had accepted its methodology in prior reviews. We find this answer to be non-responsive. The fact that we have accepted NTN Japan's allocation in prior reviews does not relieve the firm of the responsibility to demonstrate that its claimed adjustment to normal value is not distortive. Each review stands alone and our acceptance of a methodology in a prior review does not indicate acceptance of the methodology in the current review. The CIT has affirmed that we do not have to "adhere to [a] prior [methodology], especially where [we are] striving for more accuracy." See, e.g., NSK Ltd. v. United States, 19 CIT 1013, __, 896 F. Supp. 1263, 1275 (1995). In this case, in seeking to ensure that NTN Japan's allocation methodology was reasonably accurate, we asked NTN Japan specifically in our supplemental request to demonstrate that its allocation methodology was not distortive. Although NTN Japan explained that it could not report these expenses on the basis on which they were incurred, NTN Japan still had the responsibility to demonstrate that the allocation methodology it used was not distortive. Because NTN Japan did not respond to our request that it demonstrate that its methodology was not distortive in the context of this review, we can not regard the reported expenses as a reliable or reasonable indicator of what those expenses would be had NTN Japan reported them on a transaction-specific basis. Thus, we are not able to determine that NTN Japan acted to the best of its ability in selecting the allocation methodology it used. Therefore, we consider the use of adverse facts available to state NTN Japan's home-market freight costs to be warranted. We find that disallowing the entirety of the respondent's claimed home- market freight expenses would be inappropriate because this is a movement expense and the record shows that the company actually incurred these freight costs on behalf of its sales of subject merchandise during the POR. See NTN Japan's questionnaire response dated September 3, 1999, at page B-16. In selecting the adverse facts available, we calculated the freight rates for each cost center based on the worksheets NTN Japan submitted in its questionnaire response dated September 3, 1999, at exhibit B-4 and we selected the lowest freight rate from those worksheets. We consider this rate to be appropriate because it is based on NTN Japan's reported data in the context of this review. Furthermore, we verified the data upon which this freight rate is based. See NTN Japan Verification Report dated February 1, 2000, at page 6. Therefore, for the final results, we calculated NTN Japan's home-market freight by applying the rate we selected to NTN Japan's sales. See the NTN Japan final results analysis memorandum dated July 28, 2000, for our calculation of this rate. Comment 78: Torrington alleges that NTN Japan refused to provide specifically requested weight information regarding U.S. inland-freight expense. Torrington argues that the Department should address this refusal by applying adverse facts available. NTN Japan argues that facts available is warranted only when the necessary information is not available on the record, when a party withholds information requested by the administering authority, misses a deadline, significantly impedes a proceeding, or provides information that can not be verified. NTN Japan claims that it provided information available on the record in a timely fashion and that the Department has verified this information. NTN Japan argues that it acted to the best of its ability to comply with the Department's request for information. NTN Japan argues that the Department can not reject information which may not be ideal if the party has acted to the best of its ability. Department's Position: As we did for home-market freight expenses, we asked NTN Japan to demonstrate that its allocation methodology for U.S. freight expenses was not distortive. See NTN Japan supplemental questionnaire dated October 21, 1999, at page 6. Here, too, NTN Japan did not respond adequately. It merely stated that we had accepted its methodology in prior reviews. We find this answer to be non-responsive. Thus, as with NTN Japan's home-market freight expenses, we were unable to determine that its allocation methodology was not distortive. We determine that the use of adverse facts available in stating the company's U.S. freight expenses is warranted for the same reason we found it warranted in the case of the firm's home-market freight expenses. As adverse facts available, we assumed that all freight expenses which NTN Japan incurred during the POR were incurred on behalf of sales of subject merchandise and allocated the total freight expense to those sales. We find that this is reasonable because it is based on NTN's reported data. See the NTN Japan final results analysis memorandum dated July 28, 2000, for our calculation of the freight expense we applied to NTN. Comment 79: The petitioner disagrees with INA's reporting methodology to use sales value rather than a per-unit value in the allocation of INA's home-market freight costs. Torrington asserts that INA calculated a weight- based allocation for freight costs in the U.S. market and, therefore, argues that INA should use a weight-based calculation to allocate home- market freight costs. INA claims that there was no inconsistency in INA's response between the method used to allocate freight cost in the home market and in the U.S. market and states that, in both instances, the cost of freight to the customer was allocated on the basis of sales value. INA argues that it used a non-distortive allocation methodology appropriate to its circumstances in this review. INA asserts further that the Department has accepted its freight methodologies in prior reviews, citing AFBs 1, 56 FR at 31715, and AFBs 3, 58 FR at 39769-70, and that nothing has changed that would warrant requiring a different method of allocation in this review. Department's Position: While we prefer allocations of freight costs based on volume, weight, distance, or a combination of these factors, we will accept allocations based on sales value, if the firm's records do not permit the use of a preferred allocation method and the allocation constitutes a reasonable alternative methodology. See 19 CFR 351.401(g)(2). Based on INA's supplemental-questionnaire response, dated October 29, 1999, at 20, we are satisfied that INA was unable to provide its domestic inland-freight expense according to the preferred allocation method. From the evidence on the record, we have no basis to conclude that the use of sales value in this case is an unreasonable allocation methodology. INA presented evidence of the non-distortive nature of the allocation and Torrington did not provide evidence to the contrary. Also, we have accepted the methodology that INA used in past reviews. See AFB 8, 63 FR at 33340. Therefore, we have accepted INA's claimed adjustment for the final results. Comment 80: The petitioner argues that FAG Italy did not report "finished condition" packing as a packing expense in Section C of its questionnaire response but instead included those expenses as elements of COP, thus causing the U.S. packing expense to be understated in its U.S. sales listing. Torrington claims that FAG Italy stated in its October 25, 1999, supplemental questionnaire response that its parent company, FAG Germany, requests on occasion that bearings manufactured by FAG Italy at its Somma Vesuviana plant (Somma) be packed in finished condition for resale instead of in bulk for final packing in Schweinfurt. Torrington also asserts that FAG Italy has not reported the bulk-packing cost related to these resales to the United States in its total packing expenses incurred at the Somma plant and that FAG Italy treated these expenses improperly as COP. Torrington proposes that the Department reject FAG Italy's reporting of these packing expenses and apply facts available if the Department is unable to restate costs on the basis of information on the record. Torrington suggests, as facts available, that the Department double the U.S. packing expense incurred at Schweinfurt, since Torrington claims that two factories are involved in packing shipments to the United States and FAG Italy apparently reported amounts associated with only one. FAG Italy argues that Torrington is incorrect in its allegation that two factories are involved in packing shipments to the United States and that FAG Italy has reported expenses for only one factory. The respondent maintains that it explained in both its original September 3, 1999, Section C questionnaire response and its October 25, 1999, supplemental- questionnaire response at 2 that all U.S. sales of Italian bearings have associated packing costs incurred in Italy and material and labor repacking costs incurred in Schweinfurt. FAG Italy also claims that it provided a detailed explanation of packing operations at the factory at Somma and that those costs are considered part of the production expenses by the company and included in total COM. FAG Italy explains further that at the time FAG Germany makes a sale to the United States of an Italian bearing, it does not know whether the bearing originating in Somma was packed as part of an industrial bulk shipment and thus repacked in Schweinfurt or was packed in a finished state for resale in Somma. Accordingly, the respondent states, the Schweinfurt facility incurs the packing expenses associated with the bearings destined for the United States market via Schweinfurt. Thus, FAG Italy asserts that all U.S. sales include the product-specific finished packing costs in Germany, which account for all expenses incident to placing the merchandise in condition packed ready for shipment to the United States in accordance with sections 772(c)(1)(A). FAG Italy claims further that the Department has accepted this same methodology for reporting packing expenses in prior reviews and that the Department verified its accuracy and completeness. Department's Position: Based on information FAG Italy submitted on the record, we have no reason to question FAG Italy's accuracy in reporting U.S. packing expenses. In its supplemental response (see October 25, 1999, supplemental-questionnaire response at 2), FAG Italy provided a detailed explanation of its two separate types of packing operations. FAG Italy also provided sufficient information on the record for us to determine that the two types of packing expenses were both reflected accurately in FAG Italy's reported packing expense. We accept FAG Italy's explanation that, at the time that FAG Germany makes a sale to the United States of an Italian bearing, it does not know whether the bearing originating in Somma was packed as part of an industrial bulk shipment and thus repacked in Schweinfurt or was packed in a finished state for resale in Somma. Thus, the Department has no reason to believe that FAG Italy did not report the U.S. packing expenses in the COM in question accurately, and we therefore accept that FAG Italy reported the product-specific finished packing costs in Germany for U.S. sales accurately. In addition, we are satisfied that FAG Italy accounted for all expenses incident to placing the merchandise in condition packed ready for shipment to the United States in accordance with sections 772(c)(1)(A) and 773(a)(6)(A) of the Act. In addition, section 776(a) of the Act provides for the use of facts available where a company fails to provide requested information in the form and manner requested. Since there is no evidence on the record to indicate that FAG Italy has not accounted for packing expenses properly, the Department is satisfied with FAG Italy's packing methodology and accepts that it is reasonable and not distortive. Therefore, Torrington's suggestion to apply a facts-available factor to FAG Italy's reported U.S. packing costs is unwarranted, and we have made no changes for this adjustment for the final results. 10. Romania-Specific Issues Romania-Specific Issues Comment 81: Torrington argues that the Department should correct its factors-of-production calculation for the material costs for balls. Torrington claims that, where an NME respondent purchases all or a portion of its requirements from a market-economy source, the Department will normally use these actual costs rather than relying on a surrogate value, citing 19 CFR 351.408(c)(1). In this case, Torrington states, the Department used the surrogate value for alloy rod rather than the actual prices of imported bearing-quality steel rod. Torrington asserts further that two producers that supply TIE and Koyo Romania's factory imported bearing-quality steel rod from market-economy countries that was ultimately used in the production of balls. Thus, pursuant to the Department's normal methodology, Torrington asserts that the material cost for balls should have been calculated on the basis of the reported market values rather than Indonesian surrogate values. Torrington also contends that the Department relied on a surrogate value consisting of Indonesian unit import values for alloy-steel rod rather than the actual prices of the imported bearing-quality steel rod. The Romanian respondents did not rebut this argument. Department's Position: According to 19 CFR 351.408(c)(1), the Department will normally use actual costs, rather than relying on surrogate values, where an NME producer purchases all or a portion of its requirements from a market-economy source. Therefore, the Department has revised its calculations and used market-economy actual costs to value this input for the final results where the producer purchased the input material from a market-economy supplier. Comment 82: Torrington states that, in the preliminary results, the Department calculated the cost of rings using the Indonesian import values for alloy-quality steel bar. The petitioner argues that this is inappropriate for two reasons. First, Torrington claims that the alloy- quality steel import category encompasses a broad range of steel-bar products and may not include bearing-quality steels. Therefore, it maintains that it would be more appropriate for the Department to use bearing-quality steel values in determining the costs of rings. It also states that, while the bearing-quality steel rod may vary in size from alloy-steel bar, the size distinction is irrelevant and the quality similarities are more significant. Second, Torrington argues that the Department should use actual costs from market-economy imports where such data are more accurate. Torrington suggests further that the Department, as an accurate measure of the cost of such steel, examine the reported purchases of bearing-quality steel rod by the producers that supply TIE and Koyo Romania's factory from a variety of market-economy suppliers. TIE and Koyo Romania disagree with the petitioner that the Department should use the value for bearing-quality steel rod as the surrogate value for steel bar used in the production of rings. TIE and Koyo Romania argue that steel bar and steel rod are different products with different physical characteristics and different cost structures. TIE and Koyo Romania state further that the Department has routinely used distinct surrogate values for steel bar and steel rod in these reviews, citing Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al.; Preliminary Results of Antidumping Duty Administrative Reviews, 62 FR 31566, 31573 (June 10, 1997), and Tapered Roller Bearings and Parts Thereof, Finished and Unfinished from the People's Republic of China; Final Results of Antidumping Administrative Review, 62 FR 61276, 61283 (November 17, 1997). TIE and Koyo Romania claim that, contrary to Torrington's assertion, publicly available U.S. import data on the record indicates that steel bar and steel rod have different prices. Moreover, TIE and Koyo Romania claim that, if steel bar and steel rod were the same product, that the Department would not require separate factor data for rings and balls. Finally, TIE and Koyo Romania disagree with the petitioner's contention that the Indonesian import data includes steel which is not bearing quality and assert that the Department uses basket-category data routinely. TIE and Koyo Romania also contend that the steel which certain Romanian factories use is an alloy steel which has applications other than the bearing industry, just as is the case with the basket category the Department used. Therefore, TIE and Koyo Romania claim that the Indonesian Customs category which the Department used is appropriate and consistent with the quality of steel used for products sold in the United States. Thus, TIE and Koyo Romania assert that the Department's decision to value steel bar based on Indonesian import data is reasonable. Department's Position: According to 19 CFR 351.408(c)(1), the Department normally will use actual costs rather than relying on surrogate values only where an NME respondent purchases all or a portion of its requirements from a market-economy source. However, the producers that supply TIE and Koyo Romania did not purchase bearing-quality steel bar from a market-economy country. In addition, the country we selected as the surrogate, Indonesia, did not import steel bar identified as of bearing quality. We were able to confirm at verification that the grade of alloy steel TIE's supplier used for production for certain models was purchased only from NME sources. See Verification Report, March 2, 2000, at Exhibit 5. Moreover, we are satisfied that evidence on the record supports that the respondents are using alloy-steel bar to produce rings. Furthermore, in previous reviews we used Indonesian imports of alloy-steel bar to value the factor for rings, and we have found no convincing argument to change this practice. Therefore, for TIE and Koyo Romania, we are satisfied that the record demonstrates that the use of a value for alloy-steel bar in calculating the cost of rings is not distortive and have continued to use alloy-steel bar surrogate values from Indonesia for the final results. Comment 83: Torrington argues that the Department should correct its calculation of the material costs for grease. Torrington claims that, where an NME producer purchases all or a portion of its requirements from a market-economy source, the Department will normally use these actual costs rather than relying on surrogate values, citing 19 CFR 351.408(c)(1). In this case, Torrington states that the Department used the surrogate value for grease rather than the actual prices of imported grease. Torrington argues further that all of the producers that supply TIE and Koyo Romania's factory purchased grease from market-economy suppliers during the POR. Therefore, Torrington maintains that the Department erred in using surrogate-country values for this input in the preliminary results. Torrington also comments that, for a number of models where grease is listed as an input, TIE's factor worksheet for the Barlad producer that supplies TIE contained no quantity or value for grease. Thus, Torrington suggests that the Department recalculate the material cost of grease for all of the producers that supply TIE in accordance with 19 CFR 351.408(c)(1) for the final results. The respondents made no rebuttal comments. Department's Position: In accordance with 19 CFR 351.408(c)(1), we normally use actual costs, rather than relying on surrogate value, where an NME producer purchases all or a portion of its requirements from a market-economy source. Therefore, the Department has revised its calculations and used market-economy actual costs to value this input for the final results for TIE. With respect to Torrington's argument that certain quantities and values are missing grease values from the Barlad producer's factor worksheet, we have reviewed the record. TIE reported that zero units of grease were used in the production of some models. We have no reason to believe that TIE has misreported these grease values. The Department has verified TIE's reported input values in prior reviews and has found no reason to suspect the reliability of TIE's reporting of factor information. See TIE's Sales Verification Report, for the 1997-1998 administrative review, dated February 12, 1999, at 8 (placed on the record of this review). Therefore, we have accepted TIE's reporting of the factor values for grease. Comment 84: Torrington alleges that the Department should correct its calculation of material costs for rubber for one of TIE's producers. Torrington claims that the specific producer imported rubber from a market- economy source and therefore the Department should use the actual costs paid by that factory for rubber to value rubber purchases made by all producers that supply TIE. TIE disagrees with petitioner that the Department should use the specific producer's import prices for rubber from market-economy suppliers as the surrogate value for rubber for all producers that supply TIE. TIE argues that the Department has no authority to value an input in this manner, also citing 19 CFR 351.408(c)(1). TIE argues that, if the Department values a factor based on purchases from a market-economy country, then that "purchaser" must be the producer under review. TIE claims that only one of its producers purchased rubber from a market-economy supplier during the period of review and that the Department has no authority to use that producer's data as a surrogate value for rubber inputs for the other producers. Department's Position: If an individual NME producer sourced some of the input from market-economy suppliers and the remainder from domestic sources, then the value for the inputs should be based on the price paid by the individual NME producer to a market-economy supplier. See Final Rule, 62 FR at 27319; 19 CFR 351.408(c)(1). In this case, TIE purchased bearings from different producers whose purchasing habits differed. For the final results, in accordance with 19 CFR 351.408(c)(1), where the producer purchased grease as an input from a market-economy supplier and paid for it in a market-economy currency, we used the actual price paid to the market-economy supplier to value the input. In those instances where the producer purchased grease as an input from a non-market economy supplier, we used a surrogate value to value the input. Comment 85: Several issues were raised with respect to the manner in which the Department calculated SG&A, profit, and overhead ratios using the Indonesian financial statements of P.T. Jakarta Kyoei (Kyoei), P.T. Jaya Pari Steel Corporation Ltd. (Jaya Pari), and P.T. Krakatau (Krakatau). In general, the respondents objected to the following: (1) the Department's treatment of the foreign-exchange losses in the calculation of SG&A; (2) double-counting labor through the inclusion of employee wages and benefits and miscellaneous expenses in overhead; (3) the inclusion of freight and warehouse expenses in SG&A; (4) not including negative profit in the profit calculation. The petitioner agrees with the Department's overall calculation of the Indonesian financial ratios. Department's Position: In our preliminary results, we used surrogate data from financial statements of the above-referenced Indonesian steel companies in order to determine the factory overhead, SG&A, and profit rates for the Romanian respondents. We followed the same preliminary methodology to calculate the overhead, SG&A and profit rates that we used in the Preliminary Determination of Sales at Less Than Fair Value and Postponement of Final Determination: Certain Small Diameter Carbon and Alloy Seamless Standard, Line and Pressure Pipe From Romania, 65 FR 5594, 5599 (February 4, 2000). However, since the issuance of our preliminary results, the Department published its Final Determination of Sales at Less Than Fair Value and Postponement of Final Determination: Certain Small Diameter Carbon and Alloy Seamless Standard, Line and Pressure Pipe From Romania, 65 FR 39125 (June 23, 2000) (Steel Pipe). In Steel Pipe, the Department, after further analysis of the financial statements of the three Indonesian companies, found that there were problems with the use of the financial statements of Kyoei and Jaya Pari; therefore, in Steel Pipe the Department relied exclusively on the financial statements of Krakatau. Thus, for purposes of our final results and consistent with Steel Pipe, we have revised our methodology to calculate the surrogate values for overhead, SG&A, and profit and have relied exclusively on the financial statements of Krakatau. See Memorandum to The File from Suzanne Brower and J. David Dirstine dated July 21, 2000, Attachment at 6-13. There are essentially three major problems with the use of Kyoei and Jaya Pari's financial statements. First, Jaya Pari's financial statements do not identify separate energy costs; thus, we are unable to compute an accurate overhead rate for this company. Second, it is unclear whether the cost of materials, as reported on the financial statements for Kyoei and Jaya Pari, includes amounts for foreign-exchange losses. The footnotes to each of these two companies' financial statements indicate that they capitalized foreign-exchange losses related to the materials in inventory. While there is still a possibility that these companies included all foreign-exchange losses expensed in the fiscal year in the other income/expense section of the income statement as opposed to the COM, the fact that they capitalized exchange losses to inventory raises some doubt. In the case of Krakatau, however, there is no indication that any of its foreign-exchange losses are included in inventory or COM. Finally, when selecting among potential sources for financial ratios, we prefer, if possible, to use the same source for computing the surrogate factory overhead, SG&A, and profit rates. Accordingly, following the methodology used in Steel Pipe, as the financial statements for Krakatau are the only ones on the record with usable surrogate information for all three rates, we have relied on Krakatau's financial statements exclusively in order to compute the surrogate values for factory overhead, SG&A, and profit. Thus, in addressing interested-party comments regarding the computation of these three financial ratios, we have focused on those related only to Krakatau. Comment 85.a. Foreign-Exchange Gains/Losses and Imported Raw Materials. Both respondents argue that the Department's inclusion of foreign- exchange losses as part of SG&A is contrary to law and the Department's practice. The respondents contend that it is the Department's practice that foreign-exchange gains and losses on the purchase of raw materials used in the production of the subject merchandise must be related directly to the acquisition of input materials and, therefore, should be included in the COM. Specifically, the respondents maintain that, because the Indonesian steel industry imported raw materials for the production of steel during the Asian monetary crisis in 1997, it is reasonable to assume that the majority, if not all, of the net foreign-exchange losses relate to the importation of raw materials. The respondents argue further that, while their position supports the inclusion of foreign-exchange losses in COM, if the Department disagrees, those losses should still be excluded from SG&A. The respondents contend that, because Krakatau is involved in diverse activities and the evidence on the record is unclear as to which activities the losses relate, attributing all of the net foreign- exchange losses as part of SG&A is inappropriate. According to the respondents, the inclusion of foreign-exchange losses artificially inflates the dumping margin in violation of the mandate of the statute to calculate accurate margins. The respondents argue that, if the Department does not exclude foreign-exchange losses from the SG&A calculation, it would, in effect, apply an adverse inference to the respondents, contrary to the evidence on the record and the law. For these reasons, the respondents request that the Department exclude any foreign- exchange losses not included in the COM from its calculation of SG&A. In the alternative, the respondents request that the Department amortize any remaining portion of losses the Department includes in the calculation of SG&A. According to the respondents, if the Department determines that the evidence on the record is insufficient to amortize the losses, the losses must be excluded from SG&A. In support of their argument, the respondents refer to the Department's practice of recognizing only the current portion of foreign-currency losses associated with the debt. The respondents argue further that the extraordinary foreign-exchange losses reflected in the Indonesian financial statements qualify as aberrant data that should be excluded. The respondents state that, during fiscal year 1997, the Bank of Indonesia middle rates of exchange for rupiahs and U.S. dollars increased by almost 50 percent. According to the respondents, this extreme currency devaluation resulted in significant net foreign-exchange losses which, when included in the SG&A calculation, result in a distortion in the calculation of SG&A, as evidenced by very high SG&A ratios. Consequently, the respondents request that the Department, pursuant to its policy to exclude aberrant data and use good data, exclude the net foreign-exchange losses from its calculation of SG&A. Additionally, the respondents argue that, without knowing with absolute certainty the source of the net foreign-exchange losses (other than those attributable to COM because they are related to the importation of raw materials), the Department's inclusion of any portion of the net foreign- exchange losses in SG&A unlawfully applies adverse facts available to the respondents. According to the respondents, the facts of this case provide no support for the use of an adverse inference because the respondents have already provided all data requested by the Department and have cooperated fully with this review. Consequently, according to the respondents, the inclusion of the foreign-exchange losses in the calculation of SG&A is unsupported by substantial evidence on the record and is not in accordance with law. The petitioner counters the respondents' claim that the Department should revise its valuation of SG&A, which included an amount for foreign- exchange losses. Torrington disagrees with the respondents' argument that the included losses are "aberrant" because of the near-100-percent devaluation of the Indonesian rupiah between 1996 and 1997. Torrington argues that the Department should reject this argument because the respondents' arguments are based on facts not previously put on the record and that the inclusion of foreign-exchange losses due to currency devaluation in this case is highly appropriate. Department's Position: We disagree with the respondents that record evidence supports the proposition that Krakatau's net foreign-exchange gains and losses resulted solely from imported raw materials. As noted by the respondents in their case briefs, foreign currency gains and losses arise as follows: transactions involving foreign currencies are recorded at the rates of exchange prevailing at the time the transactions are made. At the balance-sheet date, assets and liabilities denominated in foreign currencies are translated to rupiah based on the prevailing exchange rates at such date. As the rupiah lost value in comparison to the U.S. dollar and most other currencies during the year, we consider it reasonable to assume that the Indonesian company's foreign-exchange losses resulted from the foreign-denominated liabilities. As a result, we consider the ratio of each type of foreign-denominated liability to total foreign-denominated liabilities a reasonable means for determining what gave rise to the net foreign-exchange losses reported on the company's audited financial statements. As can be seen from Krakatau's financial statements at footnote 31, approximately 69 percent of its foreign-denominated liabilities relates to loans while 31 percent relates to operating liabilities (i.e., purchase of inputs). We do note, however, that even if the net foreign-exchange losses were related solely to imported raw material purchases, we would still include these net losses in the SG&A rate computation, as discussed below. The respondents assume, incorrectly, that the Department automatically includes foreign- exchange gains and losses on raw materials purchases in the COM for the merchandise under consideration. Often a company incurs foreign-exchange gains and losses on raw materials that relate to both subject merchandise and non-subject merchandise. Unless we can specifically identify those gains and losses that relate only to those inputs used to produce the merchandise under consideration, to include the full amount of the net gain or loss in the COM of the merchandise under consideration results in an unreasonable assignment of these amounts. As it is very difficult to match each gain or loss specifically to the exact input used to produce each type of finished product, frequently we include the net foreign-exchange gain or loss associated with raw-material purchases in the SG&A expense computation allocated over the company-wide cost of sales (see Certain Cut-to-length Steel Plate From Mexico: Final Results of Antidumping Administrative Review, 65 FR 8338 (February 18, 2000) (Comment 5)). Alternatively, when we were able to trace the foreign- exchange losses to input purchases, we included such losses in the COM (see Round Wire from Taiwan, (Comment 4)). In the instant case, however, we were not able to identify specifically which finished products' inputs generated the net exchange loss. Since we have no reason to believe that the foreign-exchange losses attributable to raw-material purchases relate to inputs used to produce one type of finished product versus another, we consider it reasonable to allocate the company-wide net foreign-exchange loss over the operations of the company as a whole. We disagree with the respondents that the foreign-exchange losses reported in the Indonesian company's financial statements are aberrant and artificial losses. Specifically, the foreign-exchange losses for Krakatau are real costs to the company and are reported in the company's audited financial statements in accordance with Indonesian Generally Accepted Accounting Principles; these losses are a normal part of doing business. Any company, such as Krakatau, which engages in international transactions is susceptible to incurring foreign-exchange gains or losses. Several factors, such as the significance of assets and liabilities denominated in foreign currencies, the level of transactions with foreign customers and suppliers, the length of time the company takes to satisfy its foreign- denominated liabilities, and the volatility of the company's home currency in relation to that of its trading partners, contribute to the magnitude of these gains and losses. We agree with the petitioner that, while the value of the rupiah in relation to the U.S. dollar dropped more significantly this year than in past years, we do not consider this fact alone sufficient grounds to conclude that the foreign-exchange gains and losses reported on Krakatau's financial statements are aberrant. Our position that foreign-exchange losses in Indonesia were not aberrant is consistent with the recent investigation of steel plate from Indonesia (see Notice of Final Determination of Sales at Less Than Fair Value: Certain Cut-to-Length Carbon-Quality Steel Plate Products from Indonesia, 64 FR 73164 (December 29, 1999)). In that case, we included the foreign- exchange losses incurred by the respondents even though the rupiah, in relation to the U.S. dollar, dropped in value more significantly than in past years. We also disagree with the respondents that the inclusion of the foreign- exchange losses unlawfully applies adverse facts available. These foreign- exchange losses are reported on Krakatau's financial statements and we can reasonably approximate what generated the losses. Thus, the amounts used reflect Krakatau's actual costs, the necessary information for our analysis. Accordingly, the inclusion of these losses does not amount to use of facts available. Finally, we agree with the respondents that, for foreign-exchange gains and losses attributable to debt, we should only include gains or losses related to the current portion of the foreign-denominated loans. As such, we have excluded the portion of Krakatau's net foreign-exchange loss attributable to non-current debt from the SG&A rate computation. In addition, we revised Krakatau's profit-rate calculation to exclude the foreign-exchange loss related to the non-current portion of the foreign- denominated debt (see Notice of Final Determination of Sales at Less Than Fair Value: Hot-Rolled Flat-Rolled Carbon Quality Steel Products from the Russian Federation, 64 FR 38626 (July 19, 1999)). Comment 85.b. Double-counting labor, inclusion of employee wages and benefits, and miscellaneous expenses in overhead. TIE and Koyo Romania argue that, in calculating the overhead ratio for Krakatau, the Department included a line item for "employee wages, salaries, and benefits" which effectively double-counts labor costs. According to the respondents, they had accounted for labor costs in the direct and indirect labor hours they provided in their questionnaire responses. The respondents contend that the Department should remove these labor expenses from its overhead calculation. In addition, the respondents argue that the "professional services" and "exclusion for worn-out inventory" categories do not appear to be related to production and should also be excluded from the surrogate-value overhead calculation. Furthermore, the respondents claim that the Department should not have included "miscellaneous expenses" in Krakatau's overhead calculation. The respondents allege that there is no indication in Krakatau's financial statement as to what these expenses constitute and, therefore, argue that, if the Department includes these miscellaneous expenses in overhead, it raises the possibility that such expenses are misallocated. Therefore, the respondents argue that the Department should remove these expenses from its overhead calculation and, at the very least, allocate these miscellaneous expenses to other parts of the COM, not just overhead, for the final results. The petitioner disagrees with the respondents' argument that the electricity or certain labor expenses might still be included in the Krakatau overhead valuation and that the miscellaneous-expense category might have raw material, labor, electricity, or SG&A expenses, which, if true, would result in double-counting. Torrington argues that these expenses are valued separately and that the respondents' claims are based on unsupported speculation about Krakatau's financial statement. Thus, Torrington claims no change is warranted in the Department's overhead calculation for the final results. Department's Position: In its financial statements, Krakatau segregated labor costs into three categories. For our preliminary results we inadvertently classified certain labor categories as overhead expenses when developing our overhead ratio. For these final results, we have ensured that the overhead ratio reflects the correct use of labor figures. Concerning the expenses for "professional services" and "exclusion for worn-out inventory", however, these costs are in Krakatau's audited financial statements as manufacturing expenses for the steel division. As it appears that the inventory adjustment relates to maintenance-type inventory as opposed to raw materials, we consider it appropriate to treat these costs as a factory-overhead expense. In addition, while it may be more common for professional services to be categorized as an SG&A expense item on a company's financial statement, we do not believe that it is unreasonable for professional services to be used in manufacturing. Since Krakatau's audited financial statements categorized these costs as part of the COM and we do not believe professional services to be comparable to the type of labor costs included in the respondents' submitted labor factors, we consider it appropriate to include these costs in the factory- overhead computation for Krakatau. Thus, for the final results, we continued to include the costs for "professional services" and "exclusion of worn-out inventory." Comment 85.c. The Inclusion of Freight and Warehouse Expenses in SG&A The respondents argue that, in calculating SG&A expenses, the Department improperly included freight and warehouse expenses in Krakatau's SG&A. The respondents contend that such expenses are direct in nature and either were reported (such as freight) or would have been reported had they been incurred by the respondents. The respondents comment that note 34 of Krakatau's financial statements includes an amount for "sales transportation costs" which they contend should be deducted from the SG&A calculation. The respondents claim that removing transportation charges from the surrogate SG&A calculation is mandatory and is supported by case precedent. The petitioner counters that, should the Department remove the freight and warehousing expenses from the surrogate SG&A expense, the profit amount becomes understated. The petitioner contends that an amount equal to these items should be added to Krakatau's profit in calculating a surrogate profit ratio. Department's Position: We agree with the respondents that, because we accounted for freight costs elsewhere in our calculations, we should deduct freight costs from Krakatau's SG&A rate computation so as to not double-count these costs. We also agree with the petitioner that we must recalculate the surrogate profit ratio as a result of removing freight expenses from the surrogate SG&A expense rate. However, we disagree with the petitioner that we should add an amount equal to the freight expenses to the numerator in the profit- ratio calculation. Since Krakatau's sales revenue used to determine the profit includes a portion related to freight services provided, the total fully absorbed cost relating to this sales revenue must include the cost of freight. On the Romanian side, however, we treat freight expense as a sales-price adjustment as opposed to being included in the build-up of the factor-of-production cost. Thus, since the total COP to which we apply the profit rate excludes freight, we have excluded freight from the denominator in the profit-rate calculation. Comment 85.d. Profit Ratio. The respondents argue that the Department's calculation of profit using only Krakatau's financial statements is contrary to law and that a "positive" amount for profit is not required by statute. They object to the Department's reliance in the preliminary results on Certain Fresh Cut Flowers From Ecuador; Preliminary Results and Partial Rescission of Antidumping Duty Administrative Review, 64 FR 18878 (April 16, 1999) (Flowers from Ecuador), in applying the "profit cap" provision as a means of disregarding the financial statements of the Indonesian surrogate companies (i.e., Kyoei and Jaya Pari) which incurred losses. The petitioner counters that the respondents' argument contradicts the statute, the SAA, and the Department's established practice of requiring that profit be a positive amount, not a loss or a zero. Specifically, the petitioners cite Flowers from Ecuador which, in turn, followed the determination in Silicomanganese from Brazil; Final Results of Antidumping Duty Administrative Review, 62 FR 37869, 37877 (July 15, 1997), in which the Department calculated "a positive amount of profit" by "disregard(ing) financial statements of producers that incurred losses." Department's Position: As indicated above, we found problems with the financial statements of Kyoei and Jaya Pari, and, therefore, we relied exclusively on the financial statements of Krakatau to calculate SG&A, profit, and overhead for purposes of the final results. Consequently, this issue pertaining to the profit calculation is moot. 11. Miscellaneous A. Programming and Clerical Errors FAG Germany, FAG Italy, IJK, NMB/Pelmec, NTN Japan, SKF France, SKF Germany, SKF Italy, SKF Sweden, Somecat, Takeshita, and the petitioner have alleged that we made certain programming and/or clerical errors in our calculations for the preliminary results. Where we and all parties agree that a programming or clerical error occurred, we have made the necessary correction and addressed the comment only in the final-results analysis memoranda. See company-specific final results analysis memoranda, dated July 28, 2000. The comments included in this decision memorandum address situations where parties alleged that we made a programming or clerical error but either we disagree or a party to the proceedings disagrees with the allegation. Comment 86: SKF Sweden asserts that, due to a clerical error in the preliminary calculations, the Department assigned incorrectly the level of trade for U.S. sales to unaffiliated customers. According to SKF Sweden, that is because it based its assignments on customer categories rather than channels of distribution. Torrington agrees with SKF Sweden that there is an error in the preliminary calculations with respect to assigning level of trade but disagrees with the respondent as to the nature of the error. Torrington contends that in order for the Department to assign a level of trade for both U.S. and home-market sales in the calculations correctly it should define the home-market level of trade by customer category. Torrington also requests that in the final results analysis memorandum for SKF Sweden the Department accurately describe how it assigns level of trade to U.S. and home-market sales. Department's Position: Both parties' clerical-error allegations concern our level-of-trade designations in the preliminary calculations - SKF Sweden addresses U.S. sales and Torrington addresses home-market sales. We agree with both parties that certain errors occurred in our preliminary calculations with respect to the designation of level of trade for matching purposes. We have corrected these errors to ensure that our final calculations reflect the level-of-trade analysis of SKF Sweden accurately. See the final analysis memorandum dated July 28, 2000. Comment 87: Somecat contends that the Department made a clerical error that resulted in the inclusion of zero-priced shipments of "sample" merchandise to U.S. customers in the margin calculation. The respondent proposes computer-programming language to correct this error. Torrington takes no position as to the merits of Somecat's claim. However, Torrington disagrees with the methodology Somecat proposes for excluding the transactions and suggests that the Department use alternative language.. Department's Position: We agree with Somecat that a clerical error occurred which resulted in the improper inclusion of zero-priced shipments of "sample" merchandise to U.S. customers in the margin calculation. However, we have not relied on the programming language either party suggested. We have developed language which corrects the original error. To ensure that we do not collect duties on shipments of zero-priced "sample" merchandise to U.S. customers, we included the entered values and quantities of the sample transactions in our calculation of the assessment rates, and we set the antidumping duties due for the transactions to zero. We have done this because U.S. Customs will collect the ad valorem (or per- unit, where applicable) duty-assessment rate on all entries of subject merchandise regardless of whether the merchandise was a sample transaction. However, to ensure that those sample transactions do not dilute the cash-deposit margin, we excluded both the calculated U.S. prices and quantities for sample transactions from our calculation of the cash-deposit rates. Comment 88: SKF France claims that the Department's analysis memorandum states incorrectly that it assigned the level of trade for home-market sales by customer category. Instead, the respondent asserts that in the margin calculation the Department assigned the level of trade for home- market sales by channels of distribution correctly. Torrington disagrees with SKF France's assertion that the Department's preliminary results analysis memorandum is incorrect. It contends that the analysis memorandum correctly states that in the margin calculation the Department assigned level of trade for home-market sales by channel of distribution. Department's Position: We agree with Torrington that the preliminary results analysis memorandum states correctly that in the margin calculation we assigned level of trade for home-market sales by channel of distribution. Comment 89: FAG Germany and FAG Italy contend that the Department made a clerical error in the preliminary calculation when it attempted to set the warehousing commissions their U.S. affiliate paid to an affiliated warehousing company to zero. They request that the Department correct the clerical error to avoid double-counting that expense by deducting both warehousing commissions and pre-sale warehousing expenses for these sales. Torrington disagrees with the respondents. It claims that the record is unclear about whether the margin calculation results in double-counting by deducting both warehousing commissions and pre-sale warehousing expenses. Torrington contends that the respondents have not demonstrated the alleged error adequately and states that it only concurs with their argument to the extent that the Department is satisfied that the margin calculation results in double-counting. Department's Position: We agree with the respondents that a clerical error occurred. In the margin calculation we intended to deduct the actual expenses (i.e., pre-sale warehousing expenses and indirect selling expenses) incurred in the United States from CEP and to make no deduction for commissions. For the final results, we have modified our calculations to achieve this calculation. We deduct the actual expenses, instead of the commissions paid to an affiliate, because we considered the latter to be an intra-company transfer of funds. As detailed in response to Comment 18 above, affiliated- party commissions are an intra-company transfer of funds that serve to compensate an affiliate for actual expenses incurred in completing a sale to the unaffiliated customers. Such a transfer does not constitute a proper adjustment to price and, therefore, it was our intention to set the commissions that the respondents' U.S. affiliate paid to affiliated warehousing companies equal to zero. Therefore, we are satisfied that the respondents have demonstrated that deducting both the warehousing commissions and pre-sale warehousing expenses results in double-counting and have made the necessary corrections to our calculations. Comment 90: FAG Italy claims that due to a clerical error in calculating CEP profit, the Department treated CEP sales in the U.S. market as though the respondent reported these sales on a sampled basis. FAG Italy requests that the Department modify its calculations to correct this problem. The respondent proposes a calculation methodology that it says will correct the problem. Torrington agrees with the clerical error alleged by FAG Italy. However, it disagrees with the computer-programming language which the respondent proposes for correcting the problem. Torrington provides its own proposed revision to the calculations to correct the error. Department's Position: We agree with FAG Italy that we should not use the weight factor when calculating CEP profit FAG Italy did not report its CEP sales on a sampled basis. However, we have used Torrington's suggestion on how to address this clerical error because it achieves the correction without introducing additional errors. B. Date of Sale Comment 91: Torrington argues that the Department should use a different date than the date NSK reported as the date of sale for certain home- market transactions. (Torrington's argument involves business proprietary information.) NSK contends that Torrington's argument is based on unsupported speculation. NSK also contends that the Department has rejected Torrington's argument in a prior review and that nothing on the record of this review warrants a different conclusion. Department's Position: We find that NSK reported the invoice date as the date of sale for the transactions in question correctly. Due to the proprietary nature of this issue, please see the final results analysis memorandum for NSK dated July 28, 2000, for a complete discussion of our analysis of this issue. Comment 92: NSK argues that the Department should use the date of invoice rather than the date of shipment as the date of sale for certain transactions (other than the transactions covered by Comment 91 above). Citing NSK Ltd. v. United States, 115 F.3d 965, 974 (CAFC 1997) (NSK), and J.H. Cottman & Co. v. United States, 20 C.C.P.A. 344, 356 1932, cert. denied, 289 U.S. 750 (1933), NSK claims that the Department must apply the ordinary meaning of sale to the antidumping law. Citing Black's Law Dictionary at 291-92(5th ed. 1979), NSK contends that the sales in question do not become "sales" under the definition of "sale" until there is payment, or promise, of consideration. In the case of the sales in question, NSK contends it does not receive promise of consideration until the date of invoice. Citing the Department's regulations and Emulsion Styrene-Butadiene Rubber from Mexico, 64 FR 14872 (1999), NSK also alleges that the Department departed from its normal practice in defining the date of sale for these transactions. Torrington argues that the Department based the date of sale for these particular transactions on the shipment date properly. Torrington argues further that NSK is inapposite in this situation because that case involved the question of whether giving free samples is selling goods when the giver receives no consideration. In this case, according to Torrington, the Department found as a practical matter that the parties exchanged consideration. Torrington alleges that the Department has the authority to deem when such transactions become sales and, therefore, NSK is not controlling. Department's Position: Upon reconsideration of this issue, we find that NSK's invoice date, rather than the date of shipment, is the proper date to use as the date of sale for the transactions in question. Therefore, we have changed our calculations for NSK to use the sales NSK originally reported on the basis of invoice date. Due to the proprietary nature of this issue, please see the finals results analysis memorandum for NSK dated July 28, 2000, for a complete discussion of our analysis of this issue. C. Sample Weeks Comment 93: Asahi contends that the Department acted arbitrarily and unfairly in its selection of the sample weeks for CEP sales. Asahi asserts that some of the sample weeks the Department selected had exchange rates that were the most conducive for calculating margins under the antidumping law. The respondent acknowledges that section 773A of the Act and 19 CFR 351.415 regulations require the Department to ignore exchange-rate fluctuations unless there is a sustained movement increasing the value of the foreign currency relative to the U.S. dollar. However, Asahi contends that, in a review where the Department chooses the sample weeks, it must take care to use weeks that are representative and not act arbitrarily. The respondent asserts that by arbitrarily selecting several sample weeks that had the lowest Yen-per-dollar values during the POR the Department has acted contrary to law. As a remedy in this case, Asahi requests that the Department make an adjustment for sales during the second half of the POR (e.g., average the exchange rates over a three- to six-month period). Department's Position: In this proceeding, we permit respondents to report their CEP sales on a sample-week basis if the firm made more than 2,000 CEP sales (not including sales of AFBs further processed in the United States) to the United States of merchandise subject to a particular AFB order during the POR. Similarly, if the firm made more than 2,000 home- market (or third-country) sales of merchandise subject to a particular AFB order during the POR, then we allowed it to report on a sample-month basis the AFB families and part types (e.g., inner races, rolling elements) corresponding to those AFBs and part types reported in the U.S. sales listing (both export-price and CEP). We developed this reporting option for the antidumping duty orders on AFBs due to the extremely large volume of transactions and the resulting administrative burden involved in calculating individual margins for all these transactions. Selecting the sample weeks first entails breaking the POR into six two- month segments. Then we pick one week randomly from each of these segments for a total of six weeks. We choose the sample weeks by a random selection in order to prevent any capricious results. Furthermore, the fact that the sample weeks we select apply, without restriction, to all qualifying respondents subject to the antidumping duty orders on AFBs from eight different countries, including Japan, supports the fact that our selection is not arbitrary. Moreover, as noted by Asahi, the exchange-rate methodology we have used in this case is consistent with the statutory requirements and our regulations. Employing an averaging methodology, like the one proposed by Asahi, would violate section 773A of the Act and 19 CFR 351.415(c), which require that we use the rate on the date of the U.S. sale and that we ignore "fluctuations" in the exchange rates. Since our exchange-rate methodology fulfills the statutory requirements and ensures that all exporters know with certainty, when they set their U.S. prices, the daily exchange rate which we will use in our dumping analysis, we have not altered our calculation of Asahi's margin for these final results of review. D. Clerical Errors in a Respondent's Data Comment 94: Tsubaki-Nakashima contends that it misplaced the decimal in reporting the price for one model it sold to the United States and that this error distorted the margin calculations. In its case brief, the respondent provides documentation (i.e., invoices) and argument to support its view that the Department should correct the alleged clerical error. Department's Position: As detailed in response to Comment 54 above, in Flowers 5 - 7, we issued a policy for accepting a respondent's clerical errors. This policy established six conditions under which we would accept corrections of clerical errors. See also Final Results of the Antidumping Duty Administrative Review; Heavy Forged Hand Tools, Finished or Unfinished, With or Without Handles, From the People's Republic of China, 63 FR 16758 (April 6, 1998). Tsubaki-Nakashima's alleged clerical error satisfies the conditions referenced above as follows: 1) Tsubaki-Nakashima provided documentation (i.e., invoices) for sales of this model at prices which, when compared to the sales database that it submitted in response to the Department's questionnaire, demonstrated that it made a clerical error by misplacing a decimal point in its sales database; 2) we reviewed the record and found the submitted documentation in its case brief to be reliable, as similar models in the sales database had similar price-to- entered-value ratios to that of the subject model on the submitted invoices; 3) there is no information or argument on the record that indicates Tsubaki-Nakashima did not avail itself of the earliest reasonable opportunity to correct the error; 4) Tsubaki-Nakashima provided a clear explanation of the nature of its clerical error with supporting documentation in its May 11, 2000, case brief; 5) our correction of this clerical error did not involve a substantive revision of Tsubaki- Nakashima's response; and 6) Tsubaki-Nakashima's corrective documentation does not contradict information previously determined to be accurate at verification. Therefore, we accept Tsubaki-Nakashima's request that we revise this error and have used the information in the respondent's case brief in our final margin calculations. Recommendation Based on our analysis of the comments received, we recommend adopting all of the above positions and adjusting all related margin calculations accordingly. If these recommendations are accepted, we will publish the final results of reviews and the final weighted-average dumping margins for all reviewed firms in the Federal Register. Agree _________ Disagree _________ ____________________ Troy H. Cribb Acting Assistant Secretary for Import Administration ____________________ Date