66 FR 36551, July 12, 2001 A-100-001 AR:99/00 Public Document G1O3: AFB Team MEMORANDUM TO: Faryar Shirzad 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, Sweden, and the United Kingdom - May 1, 1999, through April 30, 2000 Summary We have analyzed the case and rebuttal briefs of interested parties in the May 1, 1999, through April 30, 2000, administrative reviews of the antidumping duty orders covering antifriction bearings (other than tapered roller bearings) and parts thereof from France, Germany, Italy, Japan, 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 for which we received comments and rebuttal comments by parties: 1. Facts Available 2. CEP Profit 3. Price Adjustments A. Inventory Carrying Costs B. Commissions C. Bank Charges D. Other Direct Selling Expenses E. Other 4. Resellers 5. Level of Trade 6. Arm's-Length Test 7. Prototypes and Sales Outside the Ordinary Course of Trade 8. Further Manufacturing 9. Cost of Production and Constructed Value A. Profit for Constructed Value B. Affiliated-Party Inputs C. When to Use CV 10. Packing and Movement Expenses 11. Miscellaneous A. Clerical Errors B. Scope C. Other Background On February 5, 2001, the Department of Commerce (the Department) published preliminary results of the administrative reviews of 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 (66 FR 8931, Preliminary Results). These reviews cover 56 manufacturers/exporters. The periods of review are May 1, 1999, through April 30, 2000, for ball bearings and May 1, 1999, through December 31, 1999, for cylindrical roller bearings and spherical plain bearings. We invited interested parties to comment on the preliminary results. At the request of certain parties, we held hearings for Germany- specific issues and reseller issues for France, Germany, Sweden and Italy on March 22, 2001, and for Japan-specific issues on March 26, 2001. Company Abbreviations Bearing Discount - Bearing Discount International GmbH Cerobear - Cerobear GmbH EuroLatin - EuroLatin Export Services, Ltd. INA - INA Walzlager Schaeffler oHG Koyo - Koyo Seiko Co. Ltd. Motion Bearings - Motion Bearings Pte Ltd. NSK - NSK Ltd. NTN - NTN Corporation NSK/RHP - NSK Bearings Europe, Ltd., and RHP Bearings Ltd. RIRSA - Representaciones Industriales Rodriguez, S.A. de C.V. Sapporo - Sapporo Precision Inc. SNFA France - SNFA S.A. SNFA U.K.- SNFA (U.K.) Bearings, Ltd. SNR - SNR Roulements Timken - Timken Aerospace UK, Ltd. Torrington - The Torrington Company Torrington Nadellager - Torrington Nadellager GmbH Yoo Shin - Yoo Shin Commercial Co., Ltd. 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 LTFV - Less Than Fair Value NME - Non-Market Economy POR - Period of Review SAA - Statement of Administrative Action accompanying the URAA, H.R. Doc. 103-316, Vol. 1 (1994) SPB - spherical plain bearing URAA - Uruguay Round Agreements Act AFB Administrative Determinations 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). AFBs 10 - Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews and Revocation of Orders in Part, 65 FR 49219 (August 11, 2000). Discussion of the Issues 1. Facts Available Comment 1: Torrington Nadellager argues that the Department should not assign it a margin based on adverse facts available. Torrington Nadellager states that, because the Department has given it an opportunity to respond to the Department's questionnaire, the Department should calculate its margin based on its response to the questionnaire. Department's Position: Torrington Nadellager requested originally that it be excused from responding to the questionnaire in the review concerning the order on BBs from Germany. Because we did not grant that request and because the firm had not responded to the questionnaire, we preliminarily based Torrington Nadellager's rate based on adverse facts available (see Preliminary Results, 66 FR at 8931). However, because we did not deny its request in a timely manner, we provided the respondent an opportunity following the preliminary results to respond to the questionnaire. Torrington Nadellager has since responded to our requests for information. We calculated its margins based on its responses to our questionnaire and supplemental questionnaire and disclosed the recalculations to interested parties. We received no comments. Thus, for these final results, the margin for Torrington Nadellager reflects the use of its information and we have not applied facts available. Comment 2: NTN argues that the Department should not apply adverse facts available in calculating normal value for sales NTN made to certain affiliated resellers in the home market. NTN observes that it reported the downstream sales data for sales through most of its affiliated resellers in the home market. NTN contends that it was unable to provide the downstream sales data for the other affiliated resellers and that it documented this inability to the Department. NTN asserts that, because of this, it acted to the best of its ability. NTN also asserts that, when determining whether a respondent has responded fully to requests for information, the Department must view the situation as a whole. According to NTN, because it provided the vast majority of the downstream sales made by affiliated resellers and because it acted to the best of its ability with regard to the other affiliated resellers, NTN's response does not meet the statutory requirements for applying facts available. NTN observes further that the Department's regulations indicate that a small percentage of sales to affiliated customers can not have a significant impact on margin calculations. NTN contends that the sales through the affiliated resellers in question amounts to less than the five- percent threshold set forth in 19 CFR 351.403(d). NTN also argues that, even if the use of facts available is warranted, the use of an adverse inference with regard to NTN is inappropriate. NTN observes that an adverse inference is warranted only if an interested party has failed to cooperate by not acting to the best of its ability to comply with a request for information. NTN contends that the CAFC has ruled in NTN Bearings Corp. v. United States, 74 F. 3d 1204, 1208 (1995), that the Department should analyze a party's errors or omissions "in light of its overall conduct, the importance of the information, the particular time pressures of this investigation, and any other information that will bear on the determination, of whether this was an excusable inadvertence on [the party's] part or a demonstration of lack of due regard for its responsibilities in the investigation." NTN concludes that, because it acted to the best of its ability to respond within the time limits set by the Department and because the sales in question account for a small portion of its total sales data, an adverse inference is not warranted. Finally, NTN argues that, even if an adverse inference was warranted, the rate assigned to NTN was not reasonable and not rationally related to NTN's actual dumping margin. NTN contends that the Department's methodology for selecting adverse facts available in this review is unlawful for several reasons: it resulted in transaction margins that are abnormally high; it did not bear a rational relationship to the current level of dumping in the industry; and it resulted in a higher rate than the rate the Department applied to respondents which provided no information whatsoever. NTN suggests that, if the Department continues to find that the use of adverse facts available is warranted, at worst, the Department should assign NTN its previous highest rate from the original investigation for such sales. Torrington contends that NTN's failure concerns missing sales data and, by definition, the impact of such a failure is difficult to ascertain. Torrington also argues that a review of a respondent's overall efforts may include the respondent's past efforts regarding the same issue. Torrington observes that the Department has requested similar data from NTN in prior reviews and that its lack of success in obtaining the data has led to the application of partial facts available in such reviews. Torrington also argues that the Department should select different information on the record to use as facts available that would be more reflective of NTN's failure to respond to the best of its ability. Torrington observes that the Department replaced the price of sales to these resellers of certain models with the highest price for that model sold to other customers at the same level of trade or, if the model was not sold to other customers at the same level of trade, with an upward adjustment based on the weighted-average difference between models sold to the resellers and models sold to other customers. Torrington argues that the Department's methodology is insufficiently adverse because it does not reflect NTN's refusal to supply the alternative data which the Department requested and that the use of other NTN data from sales which did not involve these resellers is unlikely to produce an incentive for NTN to report the missing data. Torrington suggests that the Department use a dumping margin based on adverse facts available for all sales of the affected models. Torrington also argues that, should the Department continue to rely upon substitute prices observed on sales of the same model to other customers, then the Department should not limit substitute values to prices observed in the same level of trade. Torrington contends that, although the sales to the resellers are presumably all made at the same level of trade, the unreported downstream sales could have been made to any level of trade. Thus, Torrington argues, there is no intrinsic rationale for limiting the selection of substitute values to sales in the same level of trade. NTN rebuts Torrington's arguments by asserting that it did not withhold information requested by the Department and that it responded to the best of its ability. NTN also contends that, contrary to Torrington's assertion that it did not provide certain alternative data requested by the Department, all of the information the Department requested is on the record. Therefore, NTN argues, the use of facts available is unwarranted. NTN also alleges that Torrington's claim that the facts available the Department selected are not "sufficiently adverse" is baseless. NTN claims that the Department's choice of facts available are not probative and ignore other, better, information available to the Department. NTN argues further that Torrington did not provide any authority for why the facts available that it suggests the Department use would result in a more probative margin. Department's Position: We have reexamined our treatment of NTN's sales to the affiliated resellers in question. For these final results, we have applied the arm's-length test to sales made to these resellers and found that they were made at arm's-length prices. To test whether sales to affiliated resellers are made at arm's-length prices, we compare, on a model-specific basis, the prices of sales to affiliated customers with sales to unaffiliated customers net of movement charges, billing adjustments, discounts, direct selling expenses, and packing expenses. Where, for the tested models of the foreign like product, prices to the affiliated party are an average of 99.5 percent or more of the price to unaffiliated parties, we determine that such sales are made at arm's- length prices. See Notice of Preliminary Determination of Sales at Less Than Fair Value: Certain Hot-Rolled Carbon Steel Flat Products from Thailand, 66 FR 22199, 22202 (May 3, 2001); see also 19 CFR 351.403(c) and Antidumping Duties, Countervailing Duties Final Rule, 62 FR 27295, 27355 (May 19, 1997). NTN's sales to the affiliated resellers in question fell within the acceptable price range and, therefore, we found them to be made at arm's-length prices. Because they were made at arm's-length prices, they formed a reasonable basis for and we used them as reported by NTN. As a result, we have not addressed the issues surrounding the selection of facts available. Comment 3: Torrington asserts that NSK did not report information concerning downstream sales in the home market for one of its affiliates. Torrington argues that, while NSK states that its total sales to affiliated resellers accounts for a de minimis amount of NSK's sales, i.e., less than five percent, and is therefore exempt from reporting, NSK ignores ownership issues with regard to this particular reseller. The petitioner argues that NSK has not offered an explanation as to why it could not report resales data and asserts that, if the Department does not obtain resale data, it should substitute adverse facts-available data to address the omission of information. NSK argues that it reported sales to certain affiliated resellers correctly. NSK states that it advised the Department that its sales to affiliated parties constituted less than five percent of its total sales. NSK asserts that, because NSK's sales to affiliates collectively accounted for less than five percent of its total sales, as it has done in the previous five reviews, NSK instead continued to report, as in past reviews, the specific percentages of foreign like product sold to each of its affiliates, including those affiliates for which NSK could not report downstream information. NSK argues that it has relied on the Department's well-established practice of accepting sales to a certain affiliated reseller when a respondent is unable to obtain downstream sales. NSK argues that, after several years of consistent determinations accepting NSK's position, a change in the Department's position at this time and its reliance on facts available at this point would be unlawful because the Department provided no notification that there was any "deficiency" in its response. Department's Position: NSK's sales of the foreign like product to affiliated parties in the home market constituted less than five percent of the value of all of NSK's sales of the foreign like product in the home market. Section 351.403(d) of our regulations states that "the Secretary normally will not calculate based on the sale by an affiliated party if sales of the foreign like product by an exporter or producer to affiliated parties account for less than five percent of the total value (or quantity) of the exporter's or producer's sales of the foreign like product in the market in question." Therefore, in accordance with our regulations, we did not require NSK to report sales of the foreign like product by the affiliates. Moreover, the sales NSK made to this affiliate were at arm's-length prices. See NSK final analysis memorandum dated July 2, 2001. As a result, these sales constitute an appropriate basis for the calculation of . Therefore the use of facts available, adverse or otherwise, is unwarranted. Comment 4: Torrington alleges that, according to excerpts from Koyo's website, which Torrington attached to its factual submission filed October 18, 2000, Koyo introduced several new bearing products during the POR. However, Torrington argues, during the U.S. and home-market sales verifications, the Department did not address directly whether Koyo properly reported all new products within the scope of the order on AFBs from Japan. Torrington requests that, prior to the completion of the review, the Department obtain the relevant model numbers of all new models and determine whether these new products constitute in-scope merchandise, were sold during the POR in the home market and the United States, and were reported completely and accurately in Koyo's database. To the extent appropriate, Torrington suggests that the Department apply facts available. Koyo responds that Torrington can not make a specific claim regarding the completeness of Koyo's sale reporting because there is simply no issue. It comments that the Department reviewed thoroughly the completeness of Koyo's sales reporting during the U.S. and home-market verifications and did not suggest in any way that Koyo neglected to include any of its recently introduced bearing models. Koyo states further that the Department received pre-verification comments from Torrington, which it discussed with Koyo company officials during the verifications. Although the verification reports do not mention the topic of the new products, Koyo asserts that the silence in the reports signifies that nothing of interest was discovered on this topic. Department's Position: Koyo's sales listing and product catalogs do indicate that new products were reported in the universe of sales made during the POR. During our U.S. and home-market verifications of Koyo's sales responses, we examined the completeness and accuracy of Koyo's reported quantity and value of merchandise sold during the POR and found no discrepancies. Although we did not specifically address Torrington's pre-verification comments in our verification report, we were satisfied that Koyo reported the universe of sales made during the POR accurately. Our regulations at 19 CFR 351.307(b) and (d) provide us with significant flexibility in conducting verifications by permitting the examination of a sample of expenses, adjustments, and other topics that we consider relevant to factual information submitted. However, verification is not intended to be an exhaustive examination of every topic. See Certain Internal-Combustion Industrial Forklift Trucks from Japan; Final Results of Antidumping Duty Administrative Review, 62 FR 5592, 5602 (February 6, 1997). Therefore, because there is no specific evidence on the record of missing sales, we have accepted Koyo's sales database as complete. Therefore, we determine that there is no need to address the issue of new- product sales by Koyo further for purposes of these reviews. Comment 5: Torrington alleges that, according to the Department's U.S. sales verification report for Koyo, the Department was unable to verify Koyo's reported entered value during the POR. Specifically, Torrington contends that Koyo officials did not have on hand, during verification, any customs documentation for the POR. Torrington argues that, if Koyo reported its entered values incorrectly, the Department can not calculate Koyo's duty-assessment rate accurately. Therefore, Torrington concludes, the Department should apply adverse facts available for the purpose of the final results. Koyo responds that Torrington's argument is based on a misunderstanding that arose during the Department's verification of Koyo's U.S. sales. Koyo explains that, with respect to the sales transaction selected by the Department to verify entered value, the bearing model in question had been imported prior to the POR and had not been imported since 1998. Therefore, Koyo explains, it was only able to produce Customs entry documents dating back to 1998 and was unable to produce more current documents for that model. However, with respect to sales of any other models imported during 1999 or 2000, Koyo states, Customs entry documents were made available to Department officials. Department's Position: Although we commented in our verification report dated February 6, 2001, that Koyo was unable to provide copies of Customs documentation for 1999 or 2000, we made this statement with respect to the sales transaction we used to verify entered value (see Exhibit E-8 of Koyo's U.S. verification report dated February 6, 2001) which contained a bearing model that entered the United States in 1998. However, we were able to verify completely the average entered value that Koyo reported in its questionnaire response. Furthermore, as stated in the verification report, we did not find any discrepancies with the calculation of this average value. Therefore, we find that Koyo reported its entered values correctly and have not applied facts available for these final results as suggested by Torrington. Comment 6: Torrington argues that, since the Department found, during verification, that Koyo reported its U.S. rebates incorrectly, the Department should apply partial facts available to Koyo's reported rebate amounts. Specifically, Torrington refers to an example for which the Department found an overpayment of a rebate amount to a customer and a decimal-point error reported throughout the field for one type of rebate. To support its argument, Torrington cites to Monsanto Co. v. United States, 698 F. Supp. 275, 281 (CIT 1988), where the CIT stated that the Department must judge the effect on the unexamined portion of the response in cases where it finds discrepancies in a subset during verification. Koyo responds that the first error to which Torrington refers did not arise in the reporting of the rebate amount but in the granting of the amount by Koyo Corporation of U.S.A. (KCUS) (Koyo's U.S. affiliate) to its customer. Thus, Koyo contends, it reported the full amount of the rebate paid to the customer in its U.S. sales database. Koyo continues that the second error to which Torrington refers involves the misplacement of a decimal point in one of the two U.S. rebate fields. In this case, Koyo contends, the Department acted appropriately by requesting that Koyo correct the error. Koyo asserts that the Department should reject Torrington's unsupported arguments in preparing the final results of reviews. Department's Position: Although we referred to some discrepancies as a result of reviewing Koyo's reporting of rebates during the verification of U.S. sales, we do not find that the nature of these discrepancies warrant the application of partial facts available as suggested by Torrington. We described the first error identified by Torrington in our verification report for Koyo as a mistake made between KCUS and its customer. See the U.S. Sales Verification Report dated February 6, 2001, at 12. The amount of the error was very small. Also, the rebate KCUS actually paid to its customer was reported in Koyo's U.S. sales database to the firm's disadvantage and not to its advantage. Furthermore, we described the second error identified by Torrington in our sales verification report for Koyo as a consistent decimal-point error in one of the rebates data fields and we requested that Koyo make the correction and resubmit its rebate figures for that field. After examining Koyo's re-submission, we are satisfied that it has corrected this error fully. In response to Torrington's reference to Monsanto Co., we agree that it is our responsibility to judge the effect on the unexamined portion of a respondent's response where we find discrepancies in a subset at verification. We did so in this case by requesting that Koyo correct the clerical error we found during verification to ensure the accuracy of the rebate data in Koyo's sales information. Further, as evident in prior reviews, it is neither unusual nor impermissible under the Act for the Department to allow respondents to correct clerical errors discovered during verification. See AFBs 10 and accompanying Decision Memo at Comment 38 and AFBs 9 64 FR at 35604. Therefore, we have not altered our treatment of Koyo's rebates for these final results. Comment 7: Torrington argues that the Department should apply partial facts available to account for discrepancies found at verification with respect to Koyo's home-market re-worked products (re-worked products were one of the groups of sales representing revenue earned from a service but recorded in Koyo's sales records as a sale excluded from the total quantity and value of sales in order to derive the universe of home-market sales Koyo reported to the Department). Torrington points to the Department's home-market verification report for Koyo, dated February 12, 2001, which states that, "concerning one item for which [the Department] requested detailed sales documentation, Koyo Seiko officials could not reconcile the figures recorded in the detailed sales record and the figures on the purchase order." In light of the Department's findings, Torrington suggests that one method the Department could use to apply partial facts available would be to include the re-worked products in the home-market sales database and increase the gross unit prices of these products by appropriate amounts. Koyo responds that this issue does not involve discrepancies in the sales values it reported, but, instead, it relates to the exclusion of revenue earned for re-work services from the universe of sales reported in its home-market database. Koyo explains that it was appropriate to segregate the value it charged to its customer for re-work services from the full amount of the transaction. However, Koyo contends, the question arose at verification over the portion of this value that involved the replacement sales of bearing parts. It was unable to locate the relevant paperwork, Koyo continues, because it was generated by a salesman at another branch and did not accompany the papers obtained from that branch during the verification. Although it would have been preferable to have found the missing paperwork during verification, Koyo asserts, it was clear from the papers that were made available and reviewed by the Department that the portion of the transaction in question did, in fact, represent the sale of replacement parts. Furthermore, Koyo argues, Torrington's proposed "solution," which uses partial facts available, that re-worked products should be included in Koyo's home-market sales database is nonsensical. Koyo asserts that, because the transactions in this situation do not involve the sale of merchandise, the inclusion of such transactions that do not have the normal attributes of sales in Koyo's database would be meaningless. Koyo concludes that the Department should recognize the insignificance of this issue and continue to accept Koyo's home-market response. Department's Position: We do not agree that we should include the re- worked products in the home-market sales database and increase the gross unit prices of these products by appropriate amounts. The sales in question were not sales of merchandise but reflected revenue earned on services rendered. Dumping determinations only cover sales of subject merchandise to which we compare sales of the foreign like product. Because the sales in question were not sales of the foreign like product, it would be inappropriate to use values in our determination of that do not reflect sales of the foreign like product. Therefore, we have not adopted Torrington's suggestion for these final results. Comment 8: Torrington argues that, with respect to Koyo's failure to report sales by its affiliated resellers in the home market, the Department should apply facts available. Specifically, Torrington claims that Koyo did not report sales by several resellers although Koyo has a 50- percent or more ownership of some of these resellers. For these resellers, Torrington contends, Koyo did not submit relevant comparison data, such as models sold to the resellers, the net price for each model, and the average net price of the same model resold to others or the average net price of the same family products resold to others, which are critical to the Department's determination of whether to accept Koyo's sales to resellers. Torrington suggests that the Department apply partial facts available as it did for NTN in AFBs 10, 65 FR at 49219 or, at a minimum, continue to exclude such sales from the calculation of normal value. Koyo argues that Torrington asserts falsely that Koyo did not supply relevant comparison data to the Department. Koyo states that it notified the Department on July 6, 2000, that it intended to report sales to its small affiliated resellers rather than the sales by these resellers. Koyo contends it has done this in the past five administrative reviews. Furthermore, Koyo continues, it reported in its questionnaire response the share of total sales made to each affiliated reseller and demonstrated that these sales fell under the Department's five-percent sales-reporting ceiling. In the supplemental questionnaire, Koyo contends, the Department requested only that Koyo describe the affiliation between Koyo and each of the resellers. However, Koyo states, the Department never objected to Koyo's reporting of sales to the resellers rather than sales by the resellers; nor did it insist that Koyo provide the affiliated resellers' sales. Furthermore, Koyo contends, the Department never requested the data Torrington claims that Koyo did not report. Instead, Koyo argues, Torrington had suggested that the Department collect this information. Koyo explains further that it has used the same methodology in this review as in prior reviews and that Torrington has only challenged the Department once in AFBs 8, 63 FR at 33342, in which the Department disagreed with Torrington and the petitioner did not appeal the decision to the CIT. Because Torrington does not provide any new bases for revisiting this issue, Koyo urges the Department to continue to accept Koyo's reported sales to its affiliated resellers. Department's Position: Koyo's sales of the foreign like product to affiliated parties in the home market constituted less than five percent of the value of all of Koyo's sales of the foreign like product in the home market. Section 351.403(d) of our regulations states that "the Secretary normally will not calculate based on the sale by an affiliated party if sales of the foreign like product by an exporter or producer to affiliated parties account for less than five percent of the total value (or quantity) of the exporter's or producer's sales of the foreign like product in the market in question." Therefore, in accordance with our regulations, we did not require Koyo to report sales of the foreign like product by the affiliates. Rather, we tested whether these sales were made at arm's-length prices using the methodology described in response to Comment 1, above. Where we found the sales to be made at arm's-length prices, we used the sales in our calculation of . Where we found the sales to be made not at arm's-length prices, we excluded the sales from our normal-value calculation. Therefore the use of facts available, adverse or otherwise, is unwarranted. 2. CEP Profit Comment 9: NTN argues that the Department should have excluded export- price sales when it calculated CEP profit. According to NTN, section 772(f) of the Act and the SAA at 824 direct that CEP profit be calculated on revenues and expenses incurred on only the sales used to calculate 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. Torrington also states that the Department's methodology was affirmed by the CIT in NTN Bearing Corp. of America v. United States, 104 F. Supp. 2d 110, 135 (CIT 2000) (NTN Bearing). 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, and AFBs 9, 64 FR at 35622. The CIT recently upheld our practice of including export-price sales in the calculation of CEP profit. See The Torrington Co. v. United States, Slip Op. 01-56 at 61 (CIT May 10, 2001) (The Torrington Co.). 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. Because NTN had export-price sales, we have included these sales in the calculation of CEP profit. Comment 10: NTN argues that the Department should calculate CEP profit on a level-of-trade-specific basis. NTN 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 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 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 does not explain why the Department should alter its decision. Torrington also states that the Department's methodology was affirmed by the CIT in NTN Bearing. Department's Position: We calculate CV profit on a level-of-trade- specific basis so that CV most accurately reflects a sale at a single, home-market level of trade. Our practice of calculating CV profit has been upheld by the CIT in RHP Bearing Ltd. v. United States, 83 F. Supp. 2d 1322 (1999). It is not our practice to calculate CEP profit on a level-of-trade- specific basis. 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's reliance on the term "narrowest", as used in sections 772(f)(2)(C)(ii) and (iii) of the Act, is misplaced. The statute uses this term in describing the second and third alternative methods for determining total expenses. Both of these methods are based upon using information contained in financial reports. For NTN, 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 method 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), The Torrington Co., Slip Op. 01-56 at 66, May 10, 2001, and in NTN Bearing, Slip Op. 00-64 (CIT 2000). Thus, although our practices for calculating CEP profit and CV profit differ, they have both been upheld by the CIT and we have employed both for calculating our final results. Comment 11: SNR states that, contrary to prior reviews, the Department did not include in this review home-market sales from the "window periods" before and after the period-of-review months in the calculation of CEP profit. SNR asserts that, since the Department matches U.S. sales to home- market sales made during the extended period, there is no basis to exclude these sales made during the extended period from the calculation of CEP profit. Torrington did not comment on this issue. Department's Position: We have reexamined this matter and have decided that, since we do match U.S. sales to home-market sales made during the extended period, we should include home-market sales from the extended period in the calculation of CEP profit. We have made this change for all companies included in these final results of reviews. Comment 12: SNR France states that the CIT, in SNR Roulements v. United States, F. Supp. 2d 1333 (CIT 2000) (SNR Roulements), ruled recently that the Department is required to include imputed expenses in the calculation of CEP profit. SNR also cites the Department's November 27, 2000 Analysis Memorandum for Draft Results of Redetermination (Redetermination) in that case, in which the Department stated that it had "complied with the Court's instructions by including imputed credit expense and imputed inventory carrying costs in the calculation of total expenses in (its) calculation of CEP profit for SNR." SNR requests that the Department correct this "clerical" error by applying the methodology of the Redetermination to the calculations for the current review. Torrington disagrees with SNR. Torrington asserts that the Department is correct in excluding imputed expenses in its calculation of CEP profit and observes that SNR Roulements is subject to appeal. Torrington states that the Department's methodology is reasonable and that the statute does not specify whether imputed expenses should be included in total expenses to calculate total profit and the CEP-profit ratio. Torrington makes the following arguments: total expenses which the Department uses to calculate total profit already include actual interest expenses as reported by the company; inclusion of imputed expenses in the same calculation results in double-counting; and reliance on actual expenses conforms with ordinary accounting methods. Torrington asserts further that the statute specifies the elements the Department should include in U.S. expenses when calculating CEP profit and that the Department's inclusion of imputed expenses in U.S. expenses to calculate the CEP-profit amount conforms with the statute. Torrington also asserts that there is no logical inconsistency in including imputed expenses in U.S. expenses and excluding them in the figure for total expenses. Torrington states that the Department's method recognizes that total expenses are best captured by reliance on actual expenses. However, according to Torrington, the relative weight of U.S. expenses might be influenced by the U.S. company's affiliation with the parent firm and should, therefore, include imputed expenses. Department's Position: As we stated in our Redetermination, available on our website at http://www.trade.gov/IAFrameset.html, we believe that it is appropriate to base the CEP-profit ratio on actual expenses as indicated in the wording of section 772(f)(1) of the Act, which directs us to calculate CEP profit on the basis of "total actual profit." We believe, as discussed below, that our practice with respect to imputed costs is reasonable. Furthermore, recent court decisions support the Department's interpretation concerning the calculation of the CEP-profit ratio. Normal accounting principles only permit the deduction of actual booked expenses, not imputed expenses, in calculating profit. Inventory-carrying costs and credit expenses are imputed expenses, not actual booked expenses, so we have established a practice of not including them in the calculation of total actual profit. See, e.g., Antidumping Duties; Countervailing Duties; Final Rule, 62 FR at 27317, 27354 (May 19, 1997), Import Administration Policy Bulletin number 97.1 at 3 and note 5; AFBs 6, 62 FR at 2113 (January 15, 1997), and Notice of Final Results of Antidumping Duty Administrative Review; Canned Pineapple Fruit from Thailand, 63 FR 7392, 7395 (February 13, 1998). Likewise, since the cost of the U.S. and home-market merchandise includes the actual booked interest expenses, the inclusion of imputed interest amounts as well in total expenses would result in double-counting of this expense to a certain extent and overstate the cost attributed to sales of this merchandise. This overstatement of cost understates the ratio of U.S. selling expenses to total expenses and consequently understates the amount of actual profit allocated to selling, distribution, and further- manufacturing activities in the United States. In addition, the use of imputed credit expenses in the calculation of CEP profit would result in the distortion of the ratio of U.S. expenses to total expenses. The change we made in our Redetermination pursuant to the CIT's order in SNR Roulements results in the addition of imputed expenses incurred on sales of the subject merchandise in the United States, but it does not result in the addition of imputed expenses incurred on sales of the foreign like product sold in the exporting country (i.e., the latter expenses are not included in "total United States expenses"). This cumulation of expenses on sales of the subject merchandise sold in the United States in a manner that is inconsistent with the cumulation of expenses on sales of the foreign like product sold in the exporting country is distortive. Finally, we are appealing the CIT's ruling in SNR Roulements, and therefore the ruling is not binding. Further, as discussed below, the CAFC has ruled in favor of the Department's efforts to avoid double-counting. In SNR Roulements, the CIT ruled that "Commerce improperly excluded imputed inventory and carrying costs from 'total expenses' when it had included these expenses in 'total United States expenses'." The CIT concluded that, "since Commerce determined that imputed inventory and carrying costs were to be included in 'total United States expenses', they must be included in 'total expenses' as well." The CAFC ruled recently, however, that the statute "does not require or even vaguely suggest symmetry between the definitions of 'U.S. expenses' and 'total expenses'." In fact, the CAFC stated that the statutory definitions themselves "undercut symmetrical treatment of 'total U.S. expenses' and 'total expenses'." See U.S. Steel Group v. United States, 225 F.3d 1284, 1290 (CAFC 2000); see also Thai Pineapple Canning Industry Corp., Ltd. v. United States, 2000 CIT LEXIS 17 (CIT September 10, 2000) (affirming the Department's method of avoiding double-counting). For the above reasons, we disagree with SNR's arguments. As instructed by the court, we included imputed expenses in our Redetermination in SNR Roulements. In doing so, however, we stated that we respectfully disagree with the CIT's instructions to include imputed expenses in "total United States expenses" in the calculation of "total expenses" and we are appealing that decision. Since there is not a final court decision on this matter, we have not changed our calculation of CEP profit with respect to imputed expenses for these final results of review. Therefore, we have not included imputed expenses in our calculation of total actual profit for purposes of calculating CEP profit. Comment 13: SNFA UK claims that the Department inappropriately applied the CEP-profit adjustment to its CEP sales made through an unaffiliated U.S. consignee. SNFA UK states that this contravenes the clear statutory language regarding CEP profit, as well as long-standing Departmental practice not to apply a CEP-profit adjustment where CEP sales are made through an unaffiliated U.S. consignee and commissions are paid to that consignee. Citing Notice of Preliminary Determination of Sales at Less Than Fair Value: Certain Non-Frozen Apple Juice Concentrate from the People's Republic of China, 64 FR 65675 (November 23, 1999), SNFA UK argues the Department did not adjust for CEP profit where CEP consignment sales were made through unaffiliated consignees. Further, SNFA UK contends, in Certain Fresh Cut Flowers From Colombia; Final Results and Partial Rescission of Antidumping Duty Administrative Review, 62 FR 53287 (October 14, 1997), the Department enunciated clearly that a consignment sale does not imply affiliation between the exporter and the consignment importer. Torrington did not comment on this issue. Department's Position: We agree with SNFA UK that we should not make a CEP-profit adjustment to U.S. price when U.S. sales were made through a consignee not affiliated through stock ownership. Since an amount for profit is already included in the arm's-length commissions SNFA UK paid to the consignee, which we deducted from U.S. price, making a CEP-profit adjustment to U.S. price would result in double-counting. We have made the correction for the final results of review. 3. Price Adjustments Inventory Carrying Costs Comment 14: Torrington argues that the Department should deny NTN's home- market inventory carrying costs rather than recalculate them as it did for the preliminary results. Torrington contends that NTN did not recalculate such costs on a class-or-kind basis as the Department requested or, alternatively, demonstrate that such a recalculation was impossible because it does not maintain inventory by class or kind and does not keep records on that basis. Moreover, Torrington states, NTN provided no evidence that the manner in which it reported its inventory carrying costs was non-distortive. According to Torrington, NTN asserted that such a recalculation was impossible. Torrington also contends that NTN did not comply with the Department's request that it recalculate inventory carrying costs to take into account the average number of days in inventory. Torrington argues that, instead of recalculating such costs to take into account the average number of days in inventory, the Department should deny the adjustment as adverse facts available because NTN did not report such costs to the best of its ability. NTN contends that it responded to the best of its ability and that the data on the record was sufficient to allow the Department to recalculate these costs on a basis which the Department deemed appropriate. Therefore, NTN argues, the denial of its claim and the use of facts available are not warranted. Department's Position: We are satisfied that NTN was unable to recalculate inventory carrying costs on a class-or-kind-specific basis because of the manner in which the firm keeps its records. For further discussion, see NTN's supplemental response dated January 5, 2001, at page B-6. In NTN's response, it indicated that the manner in which it maintains its records does not lend itself to recalculation of inventory carrying costs on class-or-kind-specific bases. See id. While we did not verify in the instant review, this information was subject to verification. NTN had no knowledge whether the Department would conduct a verification. Finally, because NTN provided the average number of days in inventory for its sales, we had the information we needed to recalculate NTN's inventory carrying costs accurately. We found that NTN had responded to our questionnaire adequately; therefore, the use of facts available, adverse or otherwise, is not warranted. Commissions Comment 15: NTN argues that the Department's test to determine whether commissions paid to home-market affiliates were at arm's length ignores important factors other than price affecting commission rates (i.e., the Department's test does not consider the percentage of the quantity or price of the affected sales in terms of total sales which NTN made in the home market). According to NTN, it negotiates commission rates individually with each selling agent, affiliated or unaffiliated, and there are many factors that affect commission rates. NTN claims that, because of the complexity and number of factors it uses to determine commission rates, it would be impractical for NTN to make a comprehensive chart explaining the commission rate for each agent. NTN contends that the more accurate method would be for the Department to review commission rates on an individual basis, which is the manner in which they are negotiated. Torrington contends that the Department has rejected this argument in prior reviews. Department's Position: When we asked for the information from which we drew our conclusions about the commissions NTN paid to affiliates, we stated, "[i]f other factors might be relevant to our analysis of the home- market commissions NTN paid to affiliated parties, please explain what those factors are and how we might account for them in our analysis. Please be aware, however, that if we are not able to make a comparison between the commissions paid to an affiliate and commissions paid to one or more unaffiliated parties, we will not be able to find that the commissions paid to the affiliated party were at arm's length and we will deny the adjustment for that affiliate." See NTN Supplemental Questionnaire dated January 5, 2001, at page B-5. NTN responded that "[c]ommission rates vary significantly between agents according to the services provided by each agent as noted on the attachment," which we took to be attachment B-7 of NTN's questionnaire response dated September 11, 2000, as that was the only data NTN provided on the record regarding the commission rates and services provided. If NTN had concerns that the data in that attachment would be insufficient for us to make a proper determination, NTN had the opportunity to clarify this in its response to our supplemental questionnaire and provide the data we would need to make such a determination. It did not do so. It is a longstanding Department practice that, when a respondent makes a claim for a favorable adjustment, it must demonstrate that it is entitled to that adjustment. See, e.g., 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 Revocation in Part, 65 FR 11767 (March 6, 2000), at Comment 1. In this case, it is incumbent on NTN to demonstrate that the commissions it paid to affiliates in the home market were at arm's length. NTN's claim that it would be impractical for it to try to make a comprehensive chart explaining the commission rate for each agent does not absolve it of its responsibility for demonstrating that its commissions paid to its affiliates in the home market are at arm's length. Because NTN did not demonstrate that the commissions paid to certain affiliates in the home market were at arm's length, we denied the adjustment for sales to those affiliates. Finally, we did review commission rates on an individual basis as NTN suggested. We examined the commission rate it paid to each individual affiliate and the services that affiliate performed and compared this information to the services rendered and rates paid to the unaffiliated commissionaire we found to be closest to the affiliated commissionaire in terms of services rendered. See NTN preliminary analysis memorandum dated January 30, 2001, at page 8. When we analyzed the data on the record, we found that commissions paid to certain affiliated parties were not made at arm's length. NTN appears to ignore the fact that we need some basis for comparison when determining whether the commissions paid to affiliated parties were made at arm's length. If there were no bases for comparison, we would have to deny the adjustment because it would be impossible to determine whether the commissions were at arm's length. However, there was such information and we made our determination based on the evidence on the record. Therefore, we have not changed our findings for the final results. Bank Charges Comment 16: NTN argues that the Department should not calculate bank charges applicable to export-price sales and add those charges to normal value. NTN claims that its reported home-market selling prices included bank charges. Therefore, according to NTN, adding these charges to normal value amounts to double-counting of the same charges. Torrington contends that the Department addressed this argument in the prior review, citing (AFBs10 and accompanying Decision Memorandum at Comment 45), and that NTN does not allege or demonstrate a different fact pattern. Therefore, Torrington contends, it is proper for the Department to make a circumstance-of-sale adjustment to for these charges. Department's Position: The bank charges NTN incurred on its export-price sales are direct selling expenses. NTN incurred these charges as a result of making export-price sales and the amounts of the charges are tied directly to each of NTN's export-price sales. See NTN's supplemental response dated January 5, 2001, at page C-3. To the extent that NTN incurs bank charges on home-market sales, it is incumbent on NTN to make a claim for an adjustment for such charges and does not mitigate the necessity for the Department to make a circumstance- of-sale adjustment for direct selling expenses incurred on export-price sales in the absence of such claims. Thus, it was proper to add bank charges incurred on export-price sales to normal value as a circumstance-of-sale adjustment. To the extent that NTN incurred bank charges on its home-market sales, it was NTN's responsibility to claim an adjustment to 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 with respect to bank charges on home- market sales. Other Indirect Selling Expenses Comment 17: NTN argues that the Department was not justified in reallocating its home-market and U.S. selling expenses without regard to level of trade. NTN contends that the Department has recognized differences in levels of trade for NTN in its home-market and U.S. sales. NTN 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 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 did not provide evidence that it incurred its selling expenses in the manner in which it allocated the expenses. Torrington also observes that the Department's practice of reallocating NTN's expenses without regard to level of trade has been affirmed by the CIT in NTN Bearing Corp. v. United States, Slip Op. 01-16 (CIT February 23, 2001) (affirming the Department's remand determination without further explanation). Department's Position: We do not dispute that selling expenses differ between NTN's levels of trade. Indeed, we reallocated NTN's packing expenses for the preliminary results in order to calculate expenses that more accurately reflect NTN's commercial situation. See Preliminary Results Analysis Memorandum for NTN, dated January 30, 2001, at page 8. NTN's methodology for allocating expenses to each of the levels of trade does not bear any relationship, however, to the manner in which NTN incurs the expenses in question at each level of trade. Therefore, we find that NTN's methodology is distortive because it allocates expenses to the levels of trade based on the relative sales values at each level of trade and not based on the level of expenses actually incurred at each level. We have addressed this issue in prior reviews. See, e.g., AFBs 8, 63 FR at 33329, first addressed in AFBs 3, 59 FR at 39750. Further, the CIT has affirmed our practice. See, e.g., NTN Bearing Corp. v. United States, Slip Op. 01-16 (CIT February 23, 2001). NTN has not changed the methodology we rejected in prior reviews; nor has it presented any evidence that it incurred the selling expenses in the same way that it allocated the expenses. Other Comment 18: With respect to Koyo's home-market billing adjustments, Torrington argues that the Department should deny adjustments which decrease normal value. Torrington comments that Koyo reported in its supplemental questionnaire that its billing adjustments were of several different substantive types. However, Torrington claims, Koyo did not correlate the various types of adjustments and reporting methods. For instance, Torrington contends, it is not clear why one type of adjustment (BILADJ1H) is allocated on a model- or transaction-specific basis and another type of adjustment (BILADJ2H) is allocated on a customer-specific basis only. Thus, Torrington claims, in the case of the customer-specific allocations, the numerator and denominator include non-scope products, referring to the supplemental response from Koyo dated November 15, 2000, pages 17- 22. Furthermore, Torrington argues, Koyo has not explained adequately why it is unable to report its billing adjustments more specifically or why the allocations do not result in distortions; nor does the Department's verification report address these issues. Thus, the petitioner asserts, the verification report does not indicate that Koyo segregated subject and non-subject merchandise. Torrington goes on to make other, more specific allegations that involve business proprietary information. Torrington asserts that reporting methods which result in the allocation of direct adjustments to sales in cases where the adjustment did not actually apply are inconsistent with the standards articulated by the CAFC. Torrington cites Torrington Co. v. United States, 82 F.3d 1039, 1050- 1051 (CAFC 1996) (Torrington CAFC), and NSK Ltd. v. United States, 190 F.3d 1321, 1328-1329 (CAFC 1999) (NSK CAFC), to support its argument. Torrington asserts further that this is especially true where the reporting method results in the allocation of adjustments which apply to non-subject merchandise. Citing SKF USA Inc. v. INA Walzlager Schaeffler KG, 180 F.3d 1370, 1375-1377 (CAFC 1999) (SKF CAFC), Torrington states that the Department properly disallowed the respondent's billing adjustments and cash discounts because the claimed adjustments were not limited to merchandise within the scope of the antidumping order. Considering the circumstances in this case, Torrington concludes, the Department should deny Koyo's negative home-market billing adjustment two (BILADJ2H) and accept only the positive adjustments as in INA Walzlager Schaeffler KG v. United States, 957 F. Supp. 251, 267-268 (CIT 1997), where the court affirmed the Department's treatment of a respondent's positive billing adjustments and remanded only to direct the Department to eliminate the negative billing adjustments. Koyo contends that the Department properly accepted Koyo's home-market billing adjustments two as reported. Citing Timken Co. v. United States, 16 F. Supp. 2d 1102, 1108 (CIT 1998) (Timken), Koyo argues that Torrington overlooks the CIT's directive that "Commerce's decision to accept [Koyo's billing adjustments two]...is supported by substantial evidence and is fully in accordance with the post-URAA statutory language and the directions of the SAA." Instead, citing AFBs 6, 62 FR at 2091, Koyo contends, Torrington relies on Torrington CAFC as support for its argument despite the fact that the Department has explained that this case is inapplicable to the issue at hand. More fundamentally, Koyo argues, the Department has explained repeatedly that Koyo's billing adjustments are not properly considered expenses at all, but rather are "direct adjustments [to gross price] necessary to identify the correct starting price," referring to AFBs 7, 62 FR at 54049. However, Koyo claims, neither decision which Torrington cites, NSK CAFC and SKF CAFC, supports the petitioner's position. Koyo states that, in both these cases, as in Torrington CAFC, the CAFC was interpreting the antidumping law as it existed prior to the enactment of the URAA. Thus, Koyo asserts, neither of these cases reflect the Department's post-URAA revised treatment of billing adjustments as adjustments to price. Koyo's response to Torrington's more specific proprietary allegations can not be explicitly addressed in this public memorandum; however, Koyo contends, in general, that the Department has never identified any evidence during verification remotely suggesting that Koyo was manipulating its downward price adjustments in order to lower its dumping margins. Furthermore, in response to Torrington's comments regarding verification in the current review, Koyo cites Bomont Industries v. United States, 733 F. Supp. 1507, 1508 (CIT 1990) (Bomont), as support that verification is a selective examination rather than a testing of an entire universe. Therefore, Koyo asserts, the Department's decision to verify some but not all aspects of Koyo's billing adjustments is not sufficient reason to deny its negative price adjustments. Koyo contends that the Department reviewed Koyo's billing adjustments in the home-market sales verification carefully and found no indication that Koyo's methodology would result in distortive allocations. Koyo concludes that, since Torrington has presented no new facts regarding this issue, the Department should reject its arguments and accept Koyo's billing adjustments as reported. Department's Position: First, Torrington's reliance on Torrington CAFC, NSK CAFC, and SKF CAFC is misplaced because the CAFC was interpreting the antidumping law as it existed prior to the enactment of the URAA and the provisions and Commerce's practice relevant to this issue have changed since then. See e.g. Preamble to Antidumping Duties; Countervailing Duties; Final Rule, 62 FR 27296, 27346-27347 (May 19, 1997), AFBs 7 at 54050-54053, AFBS 8 at 3328, and AFBs 9 at 35602. Second, 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. As we explained in prior reviews, given the large number of sales involved, we have accepted the respondent's claims that 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 did not find any indication that Koyo's methodology would result in distortive allocations during our examination of this adjustment at verification nor have we found evidence on the record to indicate that there is any distortion. We agree with Koyo's comment and reference to Bomont, where the court found that verification is not intended to be an exhaustive examination of every topic that is present. We also observe that both the CAFC and CIT have recognized that the Department has significant flexibility in conducting verifications. See, e.g., Rubberflex SDN.BHD. v. United States, 59 F. Supp. 2d 1338, 1344 (CIT 1999) (stating that "[t]he Federal Circuit and [the CIT] routinely sustain Commerce's administration of verifications as within its discretion"); American Alloys, Inc. v. United States, 30 F.3d 1469, 1475 (CAFC 1994) (noting that "the statute gives Commerce wide latitude in its verification procedures"); and PPG Industries, Inc. v. United States, 978 F.2d 1232, 1238 (CAFC 1992) (recognizing Commerce's "authority to determine the extent of investigation and information it needs"). Thus, Torrington's argument that we do not make certain statements in the verification report regarding Koyo's billing adjustments is not sufficient reason for us to reject those adjustments. Furthermore, since Torrington's arguments present no new or compelling evidence to support the proposition that Koyo's negative billing adjustments are distortive, we have allowed Koyo's billing adjustment two as a direct adjustment to normal value. . Comment 19: Torrington argues that the Department should apply partial facts available with regard to Koyo's home- market rebates by adjusting the amounts reported for billing adjustments. Torrington comments that Koyo states in its section B response that it did not grant any rebates to its home-market customers, but it set forth home- market rebate schedules at Exhibit A-26 of its section A response. Furthermore, Torrington argues, the Department's verification report does not address whether Koyo reported home market rebates as billing adjustments improperly. Based on this information, Torrington concludes, it appears that Koyo has reported rebates as billing adjustments. Therefore, Torrington contends, the Department should apply partial facts available by adjusting the reported billing adjustment amounts appropriately. Koyo responds that it included a rebate schedule in its Section A response as it had in prior reviews since it had not yet completed the preparation and analysis of its home-market sales database by the time its Section A response was due on July 24, 2000. Koyo explains that it was not until Koyo completed the analysis of its home-market sales database that it realized it had not actually granted rebates on any of the transactions reported. In any case, Koyo adds, the Department verified Koyo's billing adjustments in this review and found no evidence to suggest that Koyo tried to bury its rebates among its billing adjustments. Koyo asserts that Torrington's argument should be rejected. Department's Position: We are satisfied that Koyo has reported billing adjustments to the best of its ability. See the Department's response to Comment 18. Torrington does not provide any compelling evidence to lead us to believe that Koyo reported home-market rebates as billing adjustments. Therefore, we have not made any changes for the final results with respect to billing adjustments. Resellers Comment 20: SKF USA Inc. (SKF USA), a domestic producer of subject merchandise, argues that the Department incorrectly made preliminary determinations that certain respondents had no shipments during the POR. These respondents are identified by SKF USA as Bearing Discount, Euro- Latin, RIRSA, Rodamientos Rovi, Rovi-Maracay, and Rovi-Valencia (collectively, six respondents). SKF USA asserts that, although the Department did not state its rationale for this determination in its preliminary results of review, it learned later that the Department's determination was based on unofficial letters submitted by the six respondents and on flawed data maintained by the U.S. Customs Service. SKF USA argues that the letters and flawed data provide insufficient evidence upon which the Department can base its no-shipment determinations; thus, the Department's determination is contrary to law. Citing 19 CFR 351.104 and 351.303(f)(1)(i), SKF USA argues that the Department is required to ensure that all interested parties to a proceeding have access to all factual information the Department considers in a proceeding and that a party filing documents containing such factual information is required to simultaneously serve copies of the documents on all other parties on the service list. Furthermore, SKF USA asserts, the Department is required to maintain and make available to the public a written record of any oral communications and, in accordance with section 777(f)(3) of the Act, any ex parte meetings that take place between the Department and interested parties. SKF USA alleges that none of the required actions took place with respect to letters from the six respondents notifying the Department that they had no shipments during the POR. Therefore, SKF USA argues, since such documents were not served on interested parties simultaneously nor placed on the administrative record at the time they were submitted to the Department, the Department should not use the letters to form the basis for its determination. To support its argument, SKF USA cites Nippon Steel Corp. v. United States, 118 F. Supp. 2d 1366 (CIT 2000), stating that the CIT held that government agencies can not rely on documentation or factual information gathered through ex parte communications with interested parties without prompt and proper disclosure. It also cites PPG Industries v. United States, 708 F. Supp. 1327 (CIT 1989), and states that this case was remanded to the Department because the plaintiff did not have the opportunity to review or comment on certain undisclosed documentation. SKF USA argues further that, whether or not the letters submitted by the six respondents are accepted by the Department, the Department can not use the information in them to form the basis for a determination because the submissions did not contain the certifications as to truth and accuracy required in 19 CFR 351.303(g)(1). Furthermore, SKF USA argues, the information is unsubstantiated and unverified. SKF USA continues that, because the information provided in the letters is contradictory to the information SKF USA supplied, pursuant to 19 CFR 351.307, the Department had good cause to verify, and should have verified, the respondents. Citing Suramerica de Aleaciones Laminadas, C.A. v. United States, 14 C.I.T. 366 (1990), SKF USA observes that the CIT ruled in favor of the plaintiff where another party raised contested information for the first time in its post-hearing brief and the plaintiff was prevented from submitting rebuttal information because the rebuttal period had ended. SKF USA claims that, in Suramerica, the CIT recognized the harm of an agency's reliance on otherwise unverified factual allegations made by an interested party when those allegations are not disclosed to other parties. Therefore, SKF USA argues, the Department can not lawfully rely on such insufficient and unverified information to make its determination. Furthermore, SKF USA argues, the second part of the Department's no- shipment determination is based on flawed data. SKF USA states that, in a disclosure meeting, the Department explained that it used import data derived from information submitted by importers on Customs Service entry forms. SKF USA contends that the Department compared the names of the six respondents at issue against the data in a computer field in the Customs Service database that might include just the exporter-specific data or only the name of the manufacturer of the exported merchandise. SKF USA argues that, in the case where an importer provides only the name of the manufacturer in this data field, there is no way to determine whether the exporter is one of the six respondents which the Department determined had made no shipments during the POR. Therefore, SKF USA concludes, the Department can not rely on this Customs Service data to confirm whether the six respondents in fact had no shipments and the Department can not use this confirmation as evidence to support the it's no-shipment determination. SKF USA claims that the Department has acknowledged this problem and, in an effort to address it, proposed a clarification of its policy concerning this issue, citing Antidumping and Countervailing Duty Proceedings: Assessment of Antidumping Duties, 63 FR 55361 (October 15, 1998). Although the clarification has not yet been adopted, SKF USA points out that the Department stated, that "[in cases where] the evidence in the administrative record shows that the producer did not set the price of [certain] sales, the assessment of duties on merchandise exported by the resellers can not be based on the producer's rate." 63 FR at 55363. Moreover, SKF USA adds, it is inherently unfair to allow a party, such as the six respondents in the present case, which could have no commercial tie to the producer, to benefit from the effort and expenditure the producer made by participating in the Department's administrative review process. SKF USA concludes by requesting that the Department reverse its no-shipment determination with respect to the six respondents and apply facts available for the final results. Department's Position: It is the Department's practice to accept, conditionally, a written statement by a company under review claiming that it had no shipments of subject merchandise or did not export subject merchandise during the POR. It is also the Department's practice to then confirm with the Customs Service that a company making such a claim in fact had no shipments entered into the United States during the relevant period. See AFBs 5, 61 FR at 66520 and AFBs 4, 60 FR at 10958. Therefore, based on the letters submitted by the six respondents identified by SKF USA and based on the Department's review of the Customs Service database records, the Department had sufficient evidence to preliminarily determine that the respondents at issue had no shipments of subject merchandise during the POR. See letters from Bearing Discount dated October 24, 2000, EuroLatin dated June 20, 2000, RIRSA dated July 11, 2000, and Rodamientos Rovi dated July 13, 2000. SKF USA's claim that the Department did not provide all interested parties with access to factual information is incorrect. Thus, its implication that the Department did not make available to the public a written record of oral communications and ex parte meetings that took place is also incorrect. We put the letters submitted by these respondents on the public record prior to the issuance of the preliminary results of review. Within five days following the issuance of the preliminary results of review, we provided copies of the letters to SKF USA officials at their request. Then we held a disclosure meeting on March 1, 2001, to discuss the rationale behind our preliminary no-shipment determination for the six respondents at issue. We explained and placed on the record a description of how we searched the Customs Service database for entries involving the six respondents at issue, using as a search parameter the manufacturer/exporter field. See Memo to The File from case analyst dated March 2, 2001. None of the six respondents at issue here was listed in the Customs Service database as an exporter or manufacturer whose subject merchandise entered the United States during the POR. Thus, the record and our confirmation process support our no-shipment determination. Furthermore, there were no oral communications or ex parte meetings with any of the respondents at issue and, therefore, the administrative record does not lack any factual information in these proceedings with respect to the six respondents at issue. SKF USA's claim that the six respondents at issue did not simultaneously serve copies of their letters on all parties on the service list is correct. In this particular situation, however, because these respondents are small companies, have never participated in a review, and do not, to our knowledge, have interpreters or counsel available to assist them in understanding our request for information, we have decided to accept the letters. Furthermore, SKF USA's reference to Nippon Steel Corp. v. United States and PPG Industries v. United States does not support its argument regarding disclosure of factual information since, in the cited cases, the factual information at issue was not disclosed before a final determination was made by the Department and interested parties did not have an opportunity to review and comment on it during the administrative proceeding. In the instant case, factual information was put on the record prior to the issuance of the preliminary results of review. Further, we have disclosed our rationale to the interested parties, and all parties have had the opportunity to comment on our preliminary findings regarding this matter. With respect to SKF USA's argument that the information provided in the respondents' letters should have been verified, in accordance with 19 CFR 351.307(a) and (b), we are not required, but have the option, to verify relevant factual information prior to completing the final results of review. Specifically, our regulations state that "... the Secretary will verify factual information upon which the Secretary relies [upon] in ... [t]he final results of administrative review...if the Secretary decides that good cause for verification exists; and ... [a] domestic interested party, not later than 100 days after the date of publication of the notice of initiation of review, submits a written request for verification...." In this case, we did not find that good cause exists, nor did we receive a written request that we verify the respondents' information from a domestic interested party within the 100-day period after the notice of initiation. The Department normally does not verify no-shipment claims because of resource limitations, particularly in the absence of record evidence that might call into question such claims. Therefore, based upon the record and our methodology of reviewing Customs Service information, we have determined that the respondents at issue had no shipments during the POR and, as such, we have not established margins for use as future cash-deposit rates. Finally, we recognize that the method we use to search for entries by query of the manufacturer/exporter variable field in the Customs Service database may not be completely fail-safe. As aptly stated by SKF USA, the manufacturer/exporter field in that database which may include exporter- specific data (such as the names of the six respondents in this case) could include only the name of a manufacturer of the exported bearing and not the name of an exporter. Therefore, with respect to the six respondents at issue, notwithstanding their letters reporting that they had made no shipments, there are difficulties in verifying the accuracy of their statements based exclusively on our normal Customs query. We will therefore instruct the Customs Service to review at the time of liquidation all documentation for suspended entries of subject merchandise. If the Customs Service finds that any of the six no-shipment respondents in fact had shipments of subject merchandise during the POR, we will instruct the Customs Service to apply a facts-available rate to such respondent based on the adverse facts-available rate we have determined for the applicable country of origin (France, Germany, Italy, or Sweden) and subject merchandise. See Preliminary Results, 66 FR at 8933, for a description of our determination of these rates. Comment 21: PacAmOr Kubar Bearings, Inc. (PacAmOr), a domestic producer of BBs, alleges that Sapporo intentionally avoided and/or circumvented higher antidumping duties and ad valorem cash-deposit rates applicable to BBs made in and exported from Japan by manufacturers or producers for which individual dumping margins and cash-deposit rates have not been designated. PacAmOr further urges the Department to determine whether Sapporo accomplished its circumvention through Koyo. PacAmOr alleges that Sapporo manufactured and delivered bearings to Koyo, which Koyo repackaged and exported to the United States under the Koyo brand. Thus, PacAmOr claims, Sapporo bearings entered the United States subject to a lower antidumping rate applicable to Koyo products. PacAmOr argues further that Koyo's unsubstantiated claim that its suppliers were unaware of the merchandise's final destination (though made without confirmation from the individual manufacturers involved (including Sapporo)) is evidence that the presence of a business relationship between Koyo and Sapporo exists and, thus, supports PacAmOr's allegation that circumvention has occurred. Furthermore, PacAmOr argues, because Sapporo has not responded to the Department's questionnaire with respect to its sales to the United States, further investigation of the business relationship between Sapporo and Koyo is warranted. Specifically, PacAmOr continues, the Department should ask Koyo to: 1) identify the specific models of bearings it purchased from Sapporo; 2) state whether bearing types it manufactured are similar to those it purchased from Sapporo; 3) indicate other companies from which it purchases bearings and in what quantities; 4) list the countries to which it sells the specific bearing models supplied by Sapporo; and 5) state whether it purchases the subject bearings from Sapporo pursuant to a standardized schedule or on an order-by-order basis. PacAmOr concludes that the Department's decision not to address and resolve this matter will only provide Sapporo with the means to continue its possible circumvention of antidumping duties. Koyo responds that there is no evidence on the record to support PacAmOr's claim or the application by the Department of a higher, facts- available, antidumping duty margin to Sapporo-produced, Koyo-brand bearings. Koyo argues that, as it explained in its response to the Department's request for clarification, none of its unaffiliated suppliers shipped merchandise directly to customers in the home market, the United States, or any third-country market during the POR. Furthermore, Koyo contends, none of the subject merchandise produced by these suppliers contained special codes, characteristics, features, or markings to identify the market of final destination nor did any of the suppliers have the right to review Koyo's sales records. Koyo asserts that PacAmOr has pointed to no concrete evidence to contradict the fact that Koyo's unaffiliated suppliers have no way of knowing the ultimate destination of the bearings they supply to Koyo. Furthermore, Koyo asserts that the additional information that PacAmOr suggests that the Department gather from Koyo is already on the record or is irrelevant to the Department's margin calculation. Koyo concludes that the Department should continue to treat all bearings exported by Koyo and imported into the United States by it U.S. affiliate as Koyo-brand bearings subject to Koyo's deposit and assessment rates. Department's Position: Since Sapporo did not respond to either our original questionnaire dated August 11, 2000, or our additional request for information dated October 25, 2000, we determined preliminarily that an adverse facts-available rate should be applied to Sapporo's direct sales to the United States. However, we were unable to determine prior to our preliminary results how to treat Sapporo's sales through Koyo since there was not enough information on the record to indicate whether Sapporo had knowledge of the final destination of the bearings it supplied to Koyo. We treated U.S. sales of Sapporo-made bearings as having been sales made by Koyo for purposes of the preliminary results although we requested that Koyo provide information regarding Sapporo's knowledge of the final destination of its bearings for our final results. On January 30, 2001, we issued a supplemental letter to Koyo requesting additional information regarding subject merchandise made by other producers which Koyo sold to the United States. Based on Koyo's response, we found that Koyo's suppliers had no knowledge, or reason to know, at the time they sold their merchandise to Koyo that the merchandise was destined for the United States. Since there is no evidence on the record that Sapporo had knowledge of the final destination of its merchandise, we have determined that the sale from Koyo is the first sale to the United States. Since Koyo is the first price discriminator, we have used Koyo's pricing structure to measure dumping activity. This is consistent with section 772(b) of the Act which defines CEP sales as "the price at which the subject merchandise is first sold (or agreed to be sold) in the United States before or after the date of importation by or for the account of the producer or exporter of such merchandise...." Therefore, for the reasons set out in the preliminary results, we have continued to apply adverse facts available to Sapporo's direct sales to the United States; however, we have not made any changes through Koyo with respect to sales of Sapporo-produced bearings. 5. Level of Trade Comment 22: NTN argues that the Department should not deduct expenses and profit from CEP sales prior to identifying the level of trade. NTN 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, 1241, (CIT 1998) (Borden), and Micron Technology, Inc. v. United States, Slip. Op. 99-29 (CIT March 25, 1999) (Micron), NTN 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 CIT has affirmed the Department's methodology in FAG Kugelfisher Georg Schafer AG et al., v. United States, Slip Op. 01- 13 at 27-34 (CIT February 2, 2001). Torrington concludes that the Department should not modify its position. Department's Position: The Department's position is 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. The CIT has held that the Department's practice of determining levels of trade for CEP transactions after CEP deductions is an impermissible interpretation of section 772(d) of the Act. See Borden and Micron. The CAFC has reversed the CIT's holdings in these cases, however, on the level- of-trade issue. Specifically, the CAFC held that the statute unambiguously requires the Department to deduct the selling expenses set forth in section 772(d) of the Act from the CEP starting price prior to performing its level-of-trade analysis. See Micron Technology, Inc. v. United States, 243 F.3d. 1301, 1314-1315 (CAFC 2001) (Micron CAFC); see also Borden, Inc. v. United States, Court Nos. 99-1575,-1576 (CAFC 2001) (unpublished opinion). Consequently, the Department has continued to adjust the CEP, pursuant to section 772(d) of the Act, prior to performing the level-of- trade analysis, specified in 19 CFR 351.412. Comment 23: NTN 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 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 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 the CIT has affirmed the Department's methodology in NTN Bearing. Moreover, Torrington argues, NTN 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: As we said in our position in response to Comment 22, 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. 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 section 772(d) of the Act before determining the level of trade. That is not the case. See Micron CAFC 243 F.3d at 1314-1315. Because no home-market levels of trade reported by NTN are 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 those sales. We did make a CEP-offset adjustment to normal value, in accordance with 773(a)(7)(B) of the Act. Comment 24: Torrington argues that the Department should eliminate some level-of-trade adjustments it made with regard to certain EP sales made by NTN. Torrington alleges that, for one of the classes or kinds of merchandise NTN sold, there is no pattern of consistent price differences between two of the home-market levels of trade. Therefore, Torrington asserts, no level-of-trade adjustment for comparisons based on the differences for these levels of trade for this class or kind of merchandise is warranted. NTN argues that there is a pattern of consistent price differences between the two levels of trade for the class or kind of merchandise in question. NTN observes that Torrington bases its argument solely with respect to specific models sold to both levels of trade but ignores the quantities of these models sold to both levels of trade. NTN contends that, in order to determine whether a price difference represents a consistent pattern, the Department must consider those price differences reflected by the total quantity of these models sold to the relevant levels of trade. Department's Position: As NTN observes, we found there was a pattern of consistent price differences between the levels of trade in question for the class or kind in question on the basis of the quantities of the sales. If we made our determination only on the models sold without regard to the quantities of the models sold, it would give undue weight to individual models where there are relatively few sales of small quantities and dilute the effect of those models sold frequently and in large quantities. Thus, focusing only on the models sold without regard to the quantity of the models sold could effectively mask a very real pattern of consistent price differences. Because we found a pattern of consistent price differences between the levels of trade in question for the class or kind of merchandise in question, for these final results we have made a price- based level-of-trade adjustment where appropriate as we did in the preliminary results. 6. Arm's-Length Test Comment 25: NTN and Koyo argue that the Department's arm's-length test is flawed because the Department excludes all affiliated-party sales when the average price of sales to an affiliated party is lower than the average price of sales to unaffiliated customers. NTN argues that the Department's arm's-length test does not consider factors other than price such as the percentage of quantity or price of the affiliated-party sales in terms of total sales in determining what constitutes an arm's-length sale. NTN contends that 19 CFR 351.403(c) mandates that affiliated-party transactions will be used if the prices to affiliated parties are comparable to, not greater than or the same as, prices to unaffiliated parties. NTN cites NEC Home Electronics Ltd. v. United States, 3 F. Supp. 2d 1451 (CIT 1998) (NEC), wherein, according to NTN, the CIT found that, although the respondent had no unaffiliated-party sales in the home market, the respondent demonstrated that its sales to affiliated parties were made at arm's-length prices because they complied with the requirements for levying the Japanese Commodity Tax. According to NTN, the CIT ruled, recognizing that the Department did not consider NEC's compliance with the Japanese Commodity Tax as an indicator of an arm's-length transaction, that NEC had established that its sales to affiliated parties which met the commodity-tax requirement were equivalent to prices charged at arm's length and therefore must be applied for purposes of determining . Citing United States - Antidumping Measures on Certain Hot-Rolled Steel Products from Japan, WT/DS184/R (February 28, 2001) (Hot-Rolled Steel Products), Koyo and NTN argue that the Department should have examined factors other than price before discarding affiliated-party sales from the home-market database. Both NTN and Koyo argue further that the Department's use of its flawed arm's-length test resulted in an unjustifiable exclusion from the calculation of normal value of sales to affiliated parties. In addition, according to these respondents, the Department did not evaluate whether sales to affiliates in which the respondents held a majority interest were made at arm's length before requiring them to report downstream sales and applying facts available. According to Koyo and NTN, the Hot Rolled Steel Products WTO panel found that the arm's-length test, as applied by the United States, is not in accord with the WTO Antidumping Agreement because the test only determines whether prices to affiliated customers are lower, on average, than prices to unaffiliated customers. The respondents contend that the WTO panel stated that there is no reason to suppose that affiliation only results in sales that are outside the ordinary course of trade because they are priced lower on average than sales to unaffiliated customers. Finally, NTN contends, the Department has a strong presumption that any affiliated-party transaction automatically does not qualify as an arm's- length transaction. Specifically, NTN argues that in this review the Department assumed that sales made to certain affiliates were not made at arm's-length prices. NTN asserts that this is evident by the fact that the Department did not even apply the arm's-length test to sales it made to certain affiliated resellers but, rather, used the sales by these affiliated resellers to unaffiliated customers (i.e., downstream sales). NTN requests that the Department include all sales to affiliated parties that were excluded due to the results of the arm's-length test or, in the alternative, to examine factors, other than price alone, to determine whether an affiliated-party transaction was made at an arm's-length price. Torrington argues that the Department should reject NTN's and Koyo's assertions that the decision of the WTO panel is not consistent with the Department's practice to exclude home- market sales to affiliated parties if prices are less than 99.5 percent of the weighted-average prices to unaffiliated customers. According to Torrington, the WTO panel made a narrow determination only, expressly declining to rule whether the arm's- length test was consistent with the Antidumping Agreement. Torrington states that the WTO panel expressly limited its decision to the facts in that case. Torrington contends further that, even if the WTO panel decision were final, the URAA does not permit the Department to modify its practice until a number of specific conditions regarding Congressional consultation and the consideration of public and private sector views have been fulfilled. Therefore, Torrington concludes, the Department relied correctly on price comparisons to determine whether home-market sales to affiliated parties were made at arm's-length prices. Torrington also contends that the CIT decision in NEC did not overturn the Department's arm's-length test. According to Torrington, in that case the CIT held that the Department's application of the test was not reasonable where it presented the respondent, which had no sales to unaffiliated parties, with an impossibility. Torrington also asserts that the alternative test endorsed by the court in NEC also focused on prices. Thus, Torrington concludes, nothing in NEC supports NTN's position that the statute does not support an arm's-length test. Finally, citing NTN Bearing Corp. of America, 905 F. Supp.1083, 1098-1100 (CIT 1995), NSK Ltd. v. United States, 969 F. Supp. 34, 35 (CIT 1997), and NTN Bearing, Torrington contends that the CIT has upheld the Department on this issue. Department's Position: The regulations at 19 CFR 351.403(c) state that, "[i]f an exporter or producer sold the foreign like product to an affiliated party, the Secretary may calculate normal value based on that sale only if satisfied that the price is comparable to the price at which the exporter or producer sold the foreign like product to a person who is not affiliated with the seller." The preamble to our regulations states that "[t]he Department's current policy is to treat prices to an affiliated purchaser as 'arm's length' prices if the prices to affiliated purchasers are on average at least 99.5 percent of the prices charged to unaffiliated purchasers." See Antidumping Duties; Countervailing Duties; Final Rule, 62 FR 27296, 27355 (May 10, 1997). Thus, pursuant to our regulations and our practice as explained in the regulations, the Department interprets the term "comparable" to mean that the prices to affiliated purchasers are on average at least 99.5 percent of the prices charged to unaffiliated purchasers. With respect to NEC, the court's decision is not relevant to the issue raised in the instant review. In NEC, the court held that our application of the arm's-length test was not reasonable where it created an impossibility. In that case, the respondent had no home-market sales to unaffiliated customers. Thus, there was no basis for making a price-based comparison. By contrast, in the instant review, NTN made sales to unaffiliated customers and we used them for comparisons pursuant to our regulations. Next, with respect to the respondents' point that we should consider other, non-price factors, the record shows that NTN and Koyo did not indicate which factors, if any, we should examine in lieu of, or in addition to, prices. NTN and Koyo also made no suggestions as to how we can improve our arm's-length test. NTN and Koyo appear to ignore the fact that we must have some basis for comparison or means of determining whether the sales made to affiliated parties are at arm's-length prices. The NEC decision cited by NTN and Koyo endorsed an alternative methodology based on the specific facts of that case. NTN has not shown how the facts of that case either demonstrate that our normal practice is inappropriate or how an alternative methodology for determining whether sales made to affiliated parties were at arm's-length prices would be any better than our normal practice. Further, the respondents ignore the fact that the Department's arm's- length test takes into account factors other than price, such as the specific terms of sale. For example, because we deduct credit expenses from the prices we use in the arm's-length test, the test properly accounts for differences in payment terms. Also, we account for differences in prices between levels of trade by making price comparisons in the arm's-length test at the same level of trade. Further, NTN complains that we do not consider the quantity or price of the affiliated-party sales in terms of total sales which NTN made in the home market. NTN does not explain, however, how this has any bearing on whether the sales made to affiliated parties were made at arm's-length prices. With regard to NTN's contention that we did not test sales to certain affiliated resellers to determine whether they were made at arm's-length prices, we have reconsidered our treatment of the affiliated resellers in question. For these final results, we tested sales by these affiliates to see whether they were made at arm's-length prices. See our response to Comment 2, above, for a description of our arm's-length test. We found that sales to some affiliated customers were made at arm's-length prices but that sales to other affiliated customers were not at arm's-length prices. Where affiliated-customer sales passed the arm's-length test, we have included the prices to affiliated parties in our calculation of normal value for these final results of review. Where sales to affiliated parties were not made at arm's-length prices in accordance with our test, we excluded such sales from our calculation of normal value. Koyo and NTN also cite to a recent WTO dispute resolution panel finding in Hot-Rolled Steel Products and claim that the Department should change its practice in the current review in accordance with that finding. However, these antidumping duty administrative reviews are being conducted under the domestic laws of the United States, which are fully consistent with international obligations. The arm's-length test addressed by the respondents has been upheld repeatedly as consistent with domestic law by reviewing domestic courts. See NTN Bearing Corp. of America, 905 F. Supp. 1083 (CIT 1995); NSK Ltd. v. United States, 969 F. Supp. 34 (CIT 1997); NTN Bearing Corp. v. United States, Slip Op. 00-64 (CIT June 5, 2000). Consequently, the report by the WTO dispute-resolution panel to which the respondents refer is not relevant to this determination. We also note in passing that the report in question is currently on appeal to the Appellate Body of the Dispute Settlement Body. 7. Prototypes and Sales Outside the Ordinary Course of Trade Comment 26: NTN 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, 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 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 (CAFC 1998) (CEMEX), NTN asserts 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 claims that the evidence it provided in its response for these AFB reviews demonstrates that home-market 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 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. Torrington also observes that the Department's practice of not excluding sales with allegedly abnormally high profits has been affirmed by the CIT in NTN Bearing. 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, e.g., AFBs 9, 64 FR at 35620- 21. In order to determine that a sale is outside the ordinary course of trade due to abnormally high profits, it must be evaluated based on all the circumstances particular to the sale in question and have characteristics that are extraordinary for the home market. See 19 CFR 351.102 (definition of "ordinary course of trade"). Other than the allegedly high profits, NTN has not provided any evidence suggesting that these sales have any characteristics that would make them extraordinary for the home market. NTN's citation to CEMEX is inapposite to this situation. In CEMEX the profitability of the sales in question was merely one of the factors we considered in our determination that those sales were made outside the ordinary course of trade. In addition to profits, we found the sales in question were sales of "specialty products that were sold to a niche market," that these "sales represent[ed] a minuscule percentage of [the respondent's] total sales of cement," that "the shipping arrangements for home market sales of Types II and V cements were not ordinary," and that the record "indicated that the home market sales of Types II and V cements were of a promotional nature." See CEMEX, 133 F.3d at 901. Thus, it was the totality of circumstances, rather than the relative profitability alone, which, in CEMEX, led us to conclude that the sales were made outside the ordinary course of trade. In this case, the level of profitability is the only indicator that the sales might have been made outside the ordinary course of trade. Furthermore, NTN has not shown that these profit levels are in any way abnormally high. Finally, the CIT has affirmed our treatment of this issue in NTN Bearing, and the circumstances in this review are similar. In NTN Bearing, the CIT sustained the Department's rejection of NTN's claim that the verification of certain high-profit sales should have resulted in the exclusion of those sales from the calculation of normal value. See NTN Bearing, 104 F. Supp. 2d at 147. The CIT held that the Department's decision to require additional evidence demonstrating that sales with higher profits were outside the ordinary course of trade before excluding such sales from normal value was a reasonable exercise of discretion. Id. In this case, we have similarly exercised our discretion because of the lack of evidence demonstrating that certain high-profit sales are outside the ordinary course of trade. Consequently, we have not excluded NTN's so-called "high- profit" sales from our calculation of normal value. Comment 27: NSK argues that the Department should have excluded NSK's home-market prototype transactions from the calculation of normal value. NSK contends that it has placed detailed evidence on the record that its prototype sales in the home market have characteristics that are extraordinary for the market in question and therefore constitute transactions outside the ordinary course of trade. NSK also asserts that in prior reviews the Department has repeatedly excluded NSK's home-market prototype sales from the calculation of normal value based on similar information. NSK requests that the Department do so again for the final results of this review. Torrington argues that the burden to support a request to exclude prototype sales is on the respondent and that NSK has not met this burden. Torrington argues that, while NSK provided generic information regarding prototype sales, none of the information is actually linked to the claimed exclusions. Therefore, Torrington concludes that no modifications to the preliminary results are required. Department's Position: Our practice is to exclude home-market sales transactions from the margin calculation if we determine such transactions to be outside the ordinary course of trade, based on consideration of all the circumstances particular to the sales in question. See Murata Mfg. Co. v. United States, 820 F. Supp. 603, 607 (CIT 1993). This practice has been codified in 19 CFR 351.102, which states: [t]he Secretary may consider sales or transactions to be outside the ordinary course of trade if the Secretary determines, based on an evaluation of all of the circumstances particular to the sales in question, that such sales or transactions have characteristics that are extraordinary for the market in question. Examples of sales that the Secretary might consider as being outside the ordinary course of trade are sales or transactions involving off-quality merchandise or merchandise produced according to unusual product specifications, merchandise sold at aberrational prices or with abnormally high profits, merchandise sold pursuant to unusual terms of sale, or merchandise sold to an affiliated party at a non-arm's-length price. We have reconsidered this matter since our preliminary results and find that NSK has adequately demonstrated in its response that the sales in question were made outside the ordinary course of trade. NSK provided ample narrative explanation and documentation allowing us to examine all aspects of the sales it reported as not in the ordinary course of trade. For example, NSK indicated that orders for prototype sales involve direct consultation between and among NSK's sales branches, the engineering department of NSK, and the customer because the customer must communicate specifications and characteristics for the bearing design and use. Conversely, normal sales in the home market do not involve such consultation. See NSK's response to Item 4.b. in its Section A questionnaire response. NSK also states that prototype sales are identified in its system by a special prefix to the full product code which appears on all relevant sales and shipping documents. See NSK's Section B response at page B-33. In its Section B response, Exhibit 24, NSK demonstrated that reports are prepared after NSK's consultations with its customer concerning the technical requirements of prototype bearings. This exhibit also includes the receipt from the customer acknowledging that the customer received a prototype delivery. NSK states that, prior to the current administrative review period, the customer did not purchase models, at commercial prices or quantities, of the prototype sales of the models in question. See NSK's Section B response at page B-34. In addition, NSK has demonstrated that the prices for prototype sales are substantially higher than commercial prices in the home market. See NSK's Section B response, Exhibit 25. After evaluating all of the circumstances of NSK's sales of prototypes in the home market, we determine that the record supports the conclusion that these sales involve prototype merchandise produced according to unusual product specifications requiring special consultations between NSK and its customer, merchandise sold at aberrational prices above normal commercial prices, and merchandise sold pursuant to unusual terms of sale because no models were purchased at commercial prices or quantities prior to the prototype sales purchased by customers in this review period. Accordingly, we determine that NSK's prototype sales in the home market were outside the ordinary course of trade. Therefore, for these final results, we have excluded NSK's home-market prototype transactions from the calculation of normal value. Further Manufacturing Comment 28: Torrington argues that, with regard to imported parts which were further- manufactured in the United States prior to their sale to unaffiliated U.S. customers, the Department should require NTN and Koyo to provide information for a further-manufacturing analysis. Torrington contends that, with respect to further-manufactured merchandise, the Department has considerable discretion in applying the "special rule" which allows the Department to excuse respondents from reporting sales of further-manufactured merchandise when the value added in the United States exceeds a specific threshold. Torrington observes that the Department did not apply the special rule to Koyo in 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, 2561 (January 15, 1998) (TRBs 1995-1996), in spite of the fact that Koyo met the regulatory 65-percent guideline for merchandise with value added in the United States. In TRBs 1995-1996, Torrington asserts, the Department determined that the company did not sell a sufficient quantity of non-further-manufactured merchandise to provide a reasonable basis for calculating margins for sales of further-manufactured merchandise. Torrington argues that the facts surrounding NTN and Koyo in the current review are similar to the facts surrounding Koyo in TRBs 1995-1996. More specifically, with respect to NTN, Torrington argues that not all of NTN's facilities further-processing subject merchandise in the United States have average value-added percentages greater than 65 percent and, therefore, they do not meet the regulatory threshold for application of the special rule. Furthermore, with respect to Koyo, Torrington cites 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, 65 FR 11767, 11769 (March 6, 2000) (TRBs 1997- 1998), and argues that in proceedings regarding other bearing products produced by Koyo, such as the 1997-1998 review of the orders on tapered roller bearings and parts thereof from Japan (TRBs), the Department found that margins on non-further-manufactured merchandise were not representative of margins on further-manufactured merchandise. Torrington also makes a claim with respect to Koyo regarding the components of value added in the United States which concerns proprietary data. Finally, Torrington argues that, because the Department exempted Koyo in prior reviews from full reporting, it should no longer assume that an analysis of other non-further-manufactured merchandise produces representative margins. NTN and Koyo argue that the record demonstrates that the Department acted properly in not requiring them to submit sales and value-added information on their U.S. sales of further- manufactured merchandise. Koyo argues that, in the TRB cases cited by Torrington, its sales of further- manufactured TRBs accounted for a much higher share of Koyo's total sales of TRBs than its total U.S. sales of non-further-manufactured TRBs. Koyo argues that, unlike its sales of TRBs, there is a sufficient quantity of non-further-manufactured sales of AFBs in this review to apply the special rule. Finally, Koyo argues, Torrington points to no evidence on the record in this case to support its claim that applying the special rule would produce inaccurate results. NTN argues that the 65-percent threshold, which it asserts it meets, is based on average prices and value-added figures for the subject merchandise and that Torrington skews the data by attempting to demonstrate that NTN should not be exempted from reporting requirements based on the experience at certain plants. NTN contends that the facilities to which Torrington refers as having value-added percentages below the regulatory threshold account for only a minute fraction of the value of NTN's further-manufactured merchandise. Department's Position: While Torrington is correct in asserting that the Department has considerable discretion in applying the special rule for sales of further-manufactured merchandise, we find it is unnecessary to require NTN or Koyo to report sales of further-manufactured merchandise or the costs associated with such processing in the United States. As we stated in the preliminary analysis memorandum for NTN dated January 30, 2001, at pages 5-6, and for Koyo dated January 30, 2001, at page 4, we found that these respondents sold a sufficient quantity of non-further- manufactured merchandise to provide a reasonable basis for calculating margins for sales of further-manufactured merchandise. This information is based on the entered value of merchandise to which the respondents added value and on the entered value of the merchandise to which the respondents did not add value. We use entered value for purposes of determining whether there is a sufficient quantity of non-further-manufactured merchandise in the AFB cases because of the nature of the product under review. For example, because eight (or more) rolling elements can be imported to produce one bearing in the United States and complete bearings can be imported into the United States, we find that a simple quantity comparison based on number of pieces would not result in a meaningful comparison of quantity entered and quantity sold. Further, components of bearings, as well as complete bearings, can be further-processed into merchandise that is many times more valuable than the imported merchandise. Indeed, the higher the value added in the United States, the higher the final sales value will be relative to the value of imported merchandise. As a result, we find that a comparison based on the final sales value also would not result in a meaningful comparison. Therefore, we base our determination in these proceedings on the basis of entered value because we find that it is the only appropriate basis for making such a determination, given the nature of the merchandise and the nature of the further-processing of such merchandise. With respect to the cases Torrington cites to support its argument, in TRBs 1995-1996, 63 FR at 2561 and TRBs 1997-1998 and accompanying Decision Memo at Comment 1, one reason the Department did not apply the special rule for Koyo was that its further-manufactured U.S. sales represented a large portion of its total U.S. sales during the review period. This is not the case for NTN and Koyo in the current reviews of the AFB orders. Furthermore, in response to Torrington's comment regarding one of NTN's facilities, we calculate the average value added in the United States on the basis of the respondent as a whole, not on individual facilities or models. In response to Torrington's comment concerning a proprietary issue, we have addressed this in the analysis memo dated July 2, 2001. For these reasons, we continue to find that the value added in the United States by affiliates of Koyo and NTN for such sales is likely to exceed substantially the value of the imported subject merchandise, that there is a sufficient quantity of non-further-manufactured merchandise to provide a reasonable basis for calculating margins for sales of further-manufactured merchandise, and that the use of the special rule is appropriate with respect to these two companies. Cost of Production and Constructed Value Profit for CV Comment 29: SNR and NSK/RHP argue that the Department's CV-profit methodology is contrary to law and ignores the statutory definition of foreign like product. SNR states that the statute and legislative history demonstrate that section 773(e)(2)(A) of the Act does not authorize an aggregated CV-profit calculation such as the Department used in the preliminary results and that the only provisions that do authorize such an aggregated CV-profit calculation also require the inclusion of sales made outside the ordinary course of trade in the calculation. SNR states that the Department calculated one CV-profit ratio for each level of trade within each class or kind of product SNR sold in the ordinary course of trade but that, contrary to what is stated in the preliminary results, section 773(e)(2)(A) of the Act does not provide for calculating CV profit in this manner. According to SNR, the Act instructs the Department to calculate CV profit using one of four prescribed methods, the preferred one being that the profit calculation be equal to the sum of profits "in connection with the production and sale of a foreign like product." SNR contends that the Department, in its preliminary results, adopted a different meaning for "foreign like product" than that prescribed in the Act. SNR contends that the Department's CV-profit calculation may be performed pursuant to section 773(e)(2)(B)(i) of the Act and that this methodology is most similar to that which the Department used in its preliminary results. SNR comments that this methodology does not limit the CV-profit calculation to the use of sales within the ordinary course of trade. Finally, SNR acknowledges that it is aware that the CIT has affirmed the Department's CV-profit methodology in two cases (FAG Kugelfischer Georg Schafer AG v. United States, 2000 Court of International Trade LEXIS 78 CIT Slip Op. 00-80, (July 7, 2000) and SKF USA v. United States, Court of International Trade LEXIS 61 Slip Op.00-59, (June 1, 2000), but it states that these decisions have been appealed to the CAFC. SNR also states that the recent CIT decision in Mitsubishi Heavy Industries, Ltd. v. United States, 97 F. Supp. 2d 1203, 1207 (CIT April 26, 2000) (Mitsubishi), comments that allowing the Department to change its determination of what merchandise may reasonably be compared "could lead to awkward result of allowing [the Department] to determine that a 'foreign like product' exists for the purpose of one part of the antidumping statute but not for another within the same investigation." SNR states that the court in Mitsubishi also commented that words used twice in the same statute such as the term "foreign like product" are expected to have the same meaning. SNR also asserts that the Department stated that it had calculated CV profit on "home-market sales of a model in the ordinary course of trade" in the Department's January 30, 2001, analysis memorandum but that, in the calculations themselves, the Department did not calculate model-specific CV-profit amounts. SNR asserts that the Department's written narrative does not accurately reflect how it actually calculated the CV-profit amounts and that the Department actually calculated aggregate CV-profit ratios for each level of trade within each class or kind of product SNR sold in the home market. NSK/RHP contends that the Act instructs the Department to calculate profit for CV on amounts for the foreign like product of the merchandise. According to NSK/RHP, 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; further, NSK/RHP asserts, 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. Although NSK/RHP acknowledges that the CIT upheld the Department's CV-profit methodology in RHP Bearings Ltd. v. United States, 83 F. Supp. 2d 1322 (CIT 1999) (RHP), NSK/RHP states that it has appealed the CIT's decision and argues that the Department should change its practice for the reasons given above. Furthermore, NSK/RHP provides an analysis of its data that, according to the respondent, rebuts the Department's prior position that NSK/RHP's construction of the statute makes the statutorily preferred CV-profit methodology inapplicable to most cases involving CV. NSK/RHP 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/RHP claims, this evidence demonstrates that one of the bases by which the CIT sustained the Department in RHP does not exist in this review. Torrington disagrees with SNR and NSK/RHP. Torrington states that the Department has rejected SNR's arguments and similar arguments by other respondents in previous reviews; citing several decisions, such as AFBs 9, 64 FR at 35610-11, and AFBs10, and accompanying Decision Memorandum at Comment 8A. Torrington also states that the CIT has affirmed the Department's methodology for calculating CV profit in numerous cases, citing several such as RHP Bearings Inc. v. United States, 83 F. Supp. 2d 1322 (CIT 1999), and SKF USA Inc. v. United States, Slip Op. 00-59, 2000 WL 726941 (CIT June 1, 2000). 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." Thus, the preferred method for computing CV profit is to use actual profit and selling, general, and administrative expenses incurred on sales of the foreign like product. We calculated CV profit in this case first by calculating the total revenue and expenses for all home-market sales of the class or kind of merchandise made within the ordinary course of trade on a level-of-trade- specific basis. We then calculated the profit percentage based on the level-of-trade-specific total revenue and total expenses. Finally, we calculated the CV profit by multiplying the applicable profit rate (based on the level of trade) by the per-unit COP. As we stated in previous reviews (see AFBs 10 and accompanying Decision Memo at Comment 57), 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, "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 only upon the basis of merchandise that was used in the price-to-price comparison (i.e., identical or similar merchandise). Moreover, 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 addressed this same issue in a previous administrative review. In RHP, the CIT affirmed the Department's use of profit based upon the class or kind of merchandise which encompassed all foreign like products under consideration for purposes of CV profit because the use of such data matched the criteria of section 771(16)(C) of the Act, i.e., the same general class or kind of merchandise. With regard to NSK/RHP's analysis of data which it says rebuts the Department's previously argued position that a narrow application of the term foreign like product in the CV context causes the preferred CV-profit methodology to be inapplicable in most cases, we disagree with NSK. We did not say it would never be applicable but, rather, that it would rarely be applicable. In NSK's case, because the model or family comparable to certain models sold in the United States was not sold in the home market, 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. In such a case, the Department would be employing an alternative CV-profit methodology when actual profit amounts in connection with the production and sale of the foreign like product are available and should be used according to section 773(e)(2)(A) of the Act. Moreover, NSK's construction would create an additional layer of complexity and uncertainty in antidumping proceedings without necessarily generating more accurate results. With respect to SNR's arguments concerning the CIT's ruling in Mitsubishi, SNR raises the CIT case to support its claim that the Department's CV-profit calculation is contrary to law. The Mitsubishi case does not support such a conclusion. In Mitsubishi, the respondent challenged the Department's determination on grounds that the Department improperly determined that home-market newspaper printing presses constituted foreign like products for purposes of determining CV profit. The CIT had requested clarification on how the Department determined presses in the two markets were comparable since the Department had determined not to use such sales in its price-to-price comparison because a particular market situation existed in that case. The Department provided the clarification on remand, and the CIT affirmed the Department's methodology for calculating CV profit based on the profit from sales of the foreign like product. See Mitsubishi Heavy Industries, Ltd. v. United States, 97 F. Supp. 2d 1203, 1207 (CIT 2000). For these reasons, we have not changed our calculation of CV profit from the method we used in the preliminary results. Affiliated-Party Inputs Comment 30: Torrington argues that the Department should require NSK/RHP to amend its response so that it values affiliated-party inputs at the higher of the transfer price, market price, or the affiliate's COP for major inputs and at the higher of transfer price or market price for minor inputs. Torrington also argues that the Department should not accept at face value NSK/RHP's statement that it can not provide market prices for any of the inputs and should request more information from NSK/RHP. NSK/RHP argues that the statute does not require the Department to substitute a respondent's input cost data with an affiliated supplier's input cost data. NSK/RHP observes that the statute directs that the Department "may" make such an adjustment, but that the Department has the discretion not to make such an adjustment. NSK/RHP also contends that the Department's decision makes sense from a factual standpoint because the adjustment is sufficiently small that it could be considered an insignificant adjustment pursuant to 19 CFR 351.413. NSK/RHP also argues that the Department has the data available with which to make an adjustment should it find such an adjustment necessary. NSK/RHP contends further that its statements to the Department concerning the unavailability of market prices for affiliated-party inputs were accurate and subject to verification. Department's Position: Although the Department has discretion to use the transfer price when it is lower than the market price or the cost of production, we found no basis to exercise that discretion in this case. In keeping with our normal practice, we have adjusted NSK/RHP's costs to value affiliated-party inputs at the higher of the transfer price, market price, or the affiliate's COP for major inputs and at the higher of transfer price or market price for minor inputs for these final results. We inadvertently did not do this for the preliminary results. We used the information in exhibit D-4 of NSK/RHP's September 11, 2000, section D response to make this adjustment. Furthermore, the data in this exhibit is sufficient for us to make the adjustment pursuant to our normal practice. With regard to the market prices of NSK/RHP's affiliated-party inputs, NSK/RHP's response made it sufficiently clear that no such prices exist. NSK/RHP stated that "the affiliated parties do not sell the same parts to unaffiliated parties" and that it "did not purchase from unaffiliated parties any of the parts purchased from affiliated suppliers during the review period" on page D-9 of its September 11, 2000, section D response. Therefore, we find no deficiency in NSK/RHP's response regarding affiliated-party inputs. Comment 31: NTN argues that the Department should not adjust the COP 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 claims that the Department offered no explanation or statutory support for not using the transfer prices it reported. NTN 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 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 considers whether the amount represented as the value of the major input is less than its COP. The Department promulgated 19 CFR 351.407 to implement the statute. That section 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, and AFBs 9, 64 FR at 35612. Moreover, the CIT has upheld our application of the regulation. See Mannesmann v. United States, 77 F. Supp. 2d 1302 (CIT 1999), and American Silicon Technology v. United States, 110 F. Supp. 2d 992 (CIT 2000). Moreover, we asked NTN to provide the market prices but it did not do so. Therefore, we used the supplier's COP as a surrogate for market value. NTN 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), and in NTN Bearing. Therefore, we have continued our practice of using cost data available on the record in determining how to value major inputs. When to Use CV Comment 32: NTN 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 alleges that the Department's current practice is based on an erroneous reading of the determination in CEMEX. NTN 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, 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 has not demonstrated why the CEMEX decision should not be followed. Department's Position: The CAFC 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, 133 F.3d at 904. Section 771(15) of the Act expressly defines below-cost sales in substantial quantities and within an extended period of time as "outside the ordinary course of trade." Therefore, to be consistent with the CAFC's ruling in 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 were no home-market sales of similar merchandise in the ordinary course of trade did we resort to the use of CV for normal value. 10. Packing and Movement Expenses Comment 33: Torrington argues that, using facts available, the Department should increase Koyo's reported international freight expenses to account fully for the cost of shipping goods by air to the United States. Torrington continues that, at the least, Koyo should be required to provide separate amounts for the aggregated costs of ocean freight and air freight it reported. Specifically, Torrington claims that Koyo stated in its questionnaire response that air freight was used only to replenish inventories when stocks of certain bearings dipped below desired levels but that Koyo conceded elsewhere in its response that ocean shipment was the normal process of movement for bearings and that air freight was occasionally used for emergency situations. Torrington asserts that, because Koyo must have separate totals for air- and ocean-freight amounts in order to derive the aggregated amount, Koyo should be required to report these separate amounts to the Department. Torrington concludes that, if Koyo can not allocate the air-freight expenses to specific sales, then the Department should use the separate air- and ocean-freight totals to make allocations and adjustments as it deems appropriate. Koyo argues that Torrington has challenged the calculation of its international freight expenses in prior reviews and that the Department has rejected those challenges. In support of its argument, Koyo cites NSK Ltd. v. United States, 995 F. Supp. 123 (CIT 1998), in which the CIT affirmed the Department's acceptance of Koyo's air-freight expense calculation in the 1992-1993 administrative review. Koyo states further that it has explained its air-freight situation to the Department in prior reviews and why it is unable to tie air shipments to specific invoices. Koyo concludes that, because Torrington has pointed to no new evidence that would necessitate the Department's revisiting this issue, the Department should confirm its acceptance of Koyo's reporting methodology and reject Torrington's arguments. Department's Position: We do not find any of Torrington's arguments to be new or compelling, and we find that there is 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 so that it would now be able to link a sale to an air shipment. We discussed this issue extensively in previous reviews. See AFBs 8, 63 FR at 33340, AFBs 6, 62 FR at 2121, AFBs 5, 61 FR at 66510, and AFBs 4, 60 FR at 10942. We have found that, generally, it is not often feasible for respondents to report air and ocean freight on a transaction-specific basis in these proceedings. Therefore, as we have done in the prior reviews, we have accepted Koyo's reporting of these movement expenses for the final results of these reviews. Comment 34: Torrington argues that the Department should recalculate NTN's reported home-market and U.S. freight expenses on the basis of adverse facts available. Torrington observes that NTN did not provide weight data for bearings sold in the United States and the home market, as requested by the Department in its original and supplemental questionnaires, and that NTN allocated freight expenses in both markets on the basis of value instead of weight. Torrington contends that the Department used partial facts available in the prior administrative review for a similar failure by NTN to provide information on freight. NTN contends that it submitted evidence in its questionnaire responses that demonstrated its calculation of freight expenses; it asserts that the evidence included invoices that showed that NTN's freight expense was incurred by using factors other than weight. NTN argues that it could not allocate these expenses based on any single factor because multiple variables, which were not uniform per sale, were used to determine the expense. NTN states that the Department has accepted its methodology in all administrative reviews except for the prior one. Finally, NTN asserts that the use of facts available is not warranted because it responded fully to the Department's requests for information to the best of its ability. NTN concludes that, therefore, the Department acted properly in accepting NTN's allocation of freight expenses on a value basis. Department's Position: The regulations at 19 CFR 351.401(g)(1) 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. The regulations also provide that "[a]ny 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, 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 do not constitute subject merchandise, the preamble to our regulations states 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". 62 FR at 27348. Although it is true that weight is not the only factor on which NTN incurs freight expenses, there is no evidence on the record that the sales price of merchandise is a factor on which NTN incurs freight expenses. Thus, although it might be true that an allocation based on weight alone may have some degree of distortion, such an allocation is less distortive than an allocation calculated on a basis on which no freight expenses are incurred. Therefore, we find NTN's allocation for freight expenses to be unacceptable. While we have accepted NTN's allocation methodology in prior reviews (see AFBs 8, 63 FR at 33340, and AFBs 7, 62 FR at 54084), in those reviews we did not ask NTN to demonstrate that its allocation methodology was not distortive or to provide weight information in a supplemental questionnaire. Thus, we found in those reviews that it would be improper to deny the adjustment without allowing NTN the opportunity to show that its methodology was not distortive or to provide the requested weight information. For this review we afforded NTN the opportunity to remedy our concerns and it did not do so. Therefore, we find the use of facts available to be appropriate. If NTN had provided the weight information for which we asked in both our original questionnaire and our supplemental questionnaire, then we would have been able to attempt to reallocate NTN's reported expenses. If NTN had, at the very least, provided the information we requested, an adverse inference would not be warranted. Because the data is not on the record, however, and NTN did not provide the information following two specific requests, we find that it did not act to the best of its ability and, therefore, the use of adverse facts available is warranted. We have determined that disallowing the entirety of the respondent's claimed home-market freight expenses would be inappropriate because this is a movement expense, the record shows that the company actually incurred these freight costs for its sales of the foreign like product during the POR, and we are able to calculate an appropriate expense based on record information. See NTN's questionnaire response, dated September 11, 2000, at page B-24. In establishing what to use as facts available, we calculated the freight rates for each of NTN's cost centers based on the worksheets that NTN submitted in its questionnaire response, dated September 11, 2000, 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's reported data in the context of this review. Therefore, for the final results, we calculated NTN's home-market freight by applying the rate we selected to NTN's sales. See the NTN final results analysis memorandum, dated July 2, 2001, for our calculation of this rate. With respect to applying facts available for freight incurred on U.S. sales, we assumed that all freight expenses which NTN incurred during the POR were incurred for 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 final results analysis memorandum, dated July 2, 2001, for our calculation of the freight expense we applied to NTN. 11. Miscellaneous A. Clerical Errors SNR, NTN, Koyo, SNFA U.K., and the petitioner have alleged that we made certain clerical errors in our calculations for the preliminary results. Where we and all parties agree that a clerical error occurred, we have made the necessary correction and addressed the comment only in the final results analysis memoranda, dated July 2, 2001. The comments included in this decision memo 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 35: NTN argues that the Department erred in using sample months to calculate the margins for SPBs because NTN's home-market database contained fewer than 2,000 sales transactions of such merchandise for the POR. NTN requests that the Department calculate margins using all of the home-market transactions that occurred during the POR. Torrington argues that the Department should only make adjustments to its calculations that it determines are appropriate based on the record. Department's Position: Our examination of NTN's sales database revealed that there were more than 2,000 home-market transactions of SPBs. Therefore, it was appropriate to use sample months to calculate normal value. Comment 36: NTN asserts that the home-market packing-expense factor which the Department used in its calculations seems to be in error because the rates applied to sales of BBs and CRBs appear to lead to divergent results from those of a prior review. Torrington takes no position on the issue but comments that NTN's argument, which does not identify a particular calculation error, is speculative. Department's Position: We have made no change to the packing-expense factor for the final results because NTN has not identified a specific error and we have not been able to find any error in our reallocation of packing expenses, as described in the preliminary analysis memorandum dated January 30, 2001. B. Scope Comment 37: Torrington argues that the Department should require INA to report sales and cost data regarding certain products that INA excluded from its U.S. database. According to Torrington, INA did not report certain U.S. sales of CRBs which were not specifically excluded by the Department from the scope of the antidumping duty order. Torrington asserts that, contrary to the instructions contained in section C of the Department's questionnaire, INA did not describe the products it excluded but, rather, referred to them as "steering column supports." Torrington asserts that the description and drawing of the CRBs that INA submitted were not sufficient to identify the bearings and claims that INA has not offered any explanation why it considers the products in question to be outside the scope of the order. Torrington asserts further that INA offered no explanation for its omission of the sales from its database other than reference to a ruling by the U.S. Customs Service in which an automotive column support, which employed a BB in a manner similar to the manner in which the omitted CRBs are used, was found to be outside the scope of the orders on AFBs. Torrington states that INA appears to have made no effort to obtain from the Department a binding scope ruling pertaining to either the column support that was the subject of the Customs ruling or the column support that contains the omitted CRBs. Furthermore, Torrington argues that it is the Department and not the Customs Service that determines the scope of the order. Torrington cites at length to the Department's "Final Scope Ruling: Scope Clarification of the Antidumping Duty Order on Antifriction Bearings (Other than Tapered Roller Bearings) and Parts Thereof from Japan; NTN Bearing Corporation of America" (August 25, 2000), regarding NTN's balls used in scroll compressors for automotive air-conditioning systems, in support of its proposition that a company must submit an adequate explanation to support its omission of sales. Torrington argues that in that scope ruling the Department did not rule in favor of NTN, partly because NTN had not submitted sufficient information to support an exclusion of these balls. Torrington argues that, absent a full description of the excluded CRBs and an adequate explanation of the omission of the excluded sales, the Department should determine the CRBs to be within the scope of the antidumping duty order on CRBs and should require INA to report the omitted sales. In closing, Torrington argues that requiring INA to report all of its sales and cost data regarding the omitted sales comports with the statute and regulation on the use of facts available as well as the cited scope ruling. It comments further that, by not requiring INA to report the omitted sales, the Department removes any incentive for INA to seek a scope ruling regarding the product containing the excluded CRBs and effectively rewards INA for not yet having sought a scope ruling on the product. INA rebuts that its omission of the specific CRB sales from its databases was proper because the products sold were outside the scope of the CRB order and because Torrington's contention that the products are presumptively within the scope of the order until ruled outside the scope by the Department is incorrect. INA cites a Customs Service ruling and two prior scope rulings in support of its claim that the products excluded are outside the scope of the order on CRBs. INA explains that each of the rulings concerns a product that was found to be an automotive part (classified under heading 8708 of the Harmonized Tariff Schedule of the United States (HTSUS)) other than a wheel hub unit and, on the basis that such parts had been found to be outside the scope of the proceeding during the LTFV investigation, such products were determined to be outside the scope of the applicable order on AFBs. INA argues that the classification analysis contained in the April 21, 1997, Customs Service ruling is directly applicable to the omitted products. INA asserts that, since the product subject to that analysis, a steering column support containing a BB, was excluded as an automotive part, a steering column support containing a CRB would similarly be excluded from the order on CRBs. INA dismisses Torrington's cited scope ruling as inapplicable in instances, such as this one, where scope definition depends on Customs classification rather than a full-blown examination of the characteristics of the article in question. INA argues that Torrington is incorrect in its assertion that products are presumptively within the scope of an antidumping duty order unless or until excluded by a scope ruling. It asserts that such a view does not recognize the role of the Customs Service in the administration of antidumping duty orders, which includes making its own scope rulings. INA asserts that, as contemplated by the Department in its regulation on scope rulings and its comments to that regulation, an importer is entitled to rely upon a Customs Service ruling unless the Department rules otherwise in a scope ruling of its own. Finally, INA asserts that Torrington's request for additional sales information at this stage of the proceeding - following verification and the release of preliminary results - is untimely, imposes an unwarranted burden on the administrative review process, and, due to its untimeliness, should be rejected by the Department. Department's Position: We have reviewed INA's response to section C of the questionnaire and its response to our supplemental questionnaire. We find that INA responded adequately to our request for information regarding the omission of certain CRB sales from its database. INA provided a description of the omitted products and an explanation for the omission of sales that satisfied us that the sales should not be included in the database. Based on the description and drawings of the CRBs and the April 21, 1997, Customs Service ruling, all of which were submitted by INA, we concluded that a steering column support unit incorporating the CRBs is, by definition, outside the scope of the order on CRBs. The Customs Service ruling is instructive in that it strongly implies that, as such, parts would be classified under item number 8708 of the HTSUS, an item number specifically excluded from the scope of the order on CRBs. See "Final Scope Ruling - Antidumping Duty Orders on Ball Bearings, Cylindrical Roller Bearings and Spherical Plain Bearings and Parts Thereof from Japan and Singapore - Request by Rockwell International Corporation, for a Ruling on Cushion Suspension Unit Assemblies" (February 10, 1997); "Final Scope Ruling -- Antidumping Duty Order on Ball Bearings, Cylindrical Roller Bearings Spherical Plain Bearings and Parts Thereof from Japan - Request by Dana Corporation, for a ruling on a Center Bracket Assembly" (June 26, 1996); Antidumping Duty Orders: Ball Bearings, Cylindrical Roller Bearings, and Spherical Plain Bearings and Parts Thereof From the Federal Republic of Germany, 54 FR 20900, 20901-02 (May 15, 1989). Therefore, since the CRBs are outside the scope of the order, we have concluded that it was not necessary for INA to report the omitted sales. With respect to the August 25, 2000, scope ruling which Torrington cites extensively, we determined that the balls of the EM Coupling were not outside the scope of the order on BBs because the balls could not be distinguished from balls used in a BB. Because we concluded that INA responded adequately to our request for information regarding the omission of sales from its database and because we are satisfied that the CRBs in question were properly excluded by INA from its response as automotive parts, we did not ask INA to report the sales and cost information for these sales for our final results. C. Other Assigning of Zero-Percent Margins Comment 38: NSK, NSK/RHP, and NTN assert that the Department's practice of assigning a zero-percent dumping margin for sales to the United States made at or above the prices charged in the home market violates Article 2.4.2 of the Agreement on Implementation of Article VI of the General Agreement on Tariffs and Trade (1994) (Antidumping Agreement). In support of their assertion, the companies cite European Communities - Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India, WT/DS141/R, par. 6.116 (October 30, 2000) (Bed Linen Panel Decision). In this report, they state, a WTO panel found that "the European Communities" acted inconsistently with Article 2.4.2 of the Antidumping Agreement in establishing dumping margins by using a methodology that included zeroing negative price differences for some models of bed linen. NSK and NSK/RHP comment that the panel's decision was affirmed by the WTO appellate body. NSK and NSK/RHP argue that the Department's practice of zeroing negative margins to calculate the overall antidumping duty margin likewise violates the provision of the Antidumping Agreement and, thus, violates U.S. antidumping law. NSK and NSK/RHP assert that the practice specifically violates the URAA, which was enacted to amend U.S. antidumping law so as to implement Article 2.4.2. Finally, NSK and NSK/RHP assert that the WTO panel decided that the requirements of Article 2.4.2 applied to the entire process of calculating margins; in their view, the panel decision applies to all dumping methodologies, including the comparison of weighted-average normal values to transaction-specific export prices. In conclusion, NSK and NSK/RHP ask the Department to offset positive margins with negative margins in its calculations for the final results of review. NTN asserts that the reasoning in the WTO panel's findings, which deals with fair comparisons in the investigation phase of a proceeding, is equally applicable to administrative reviews. NTN comments further, citing Bowe Passat Reinigungs-und Washereitechnik GmbH v. United States, 926 F. Supp. 1138, 1150 (CIT 1996), that the CIT has found that the practice of zeroing negative margins creates inflated margins and is incapable of producing a fair result. NTN concludes that the Department should discontinue the practice of zeroing in order to conform with U.S. antidumping law and with U.S. obligations under GATT. Koyo also argues that the Department should abandon the practice of zeroing negative margins in light of the decision of the WTO appellate body, European Communities - Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India, WT/DS141/AB/R (March 1, 2001) (Bed Linen Appellate Decision), in which the panel found such a practice to be inconsistent with Article 2.4.2 of the Antidumping Agreement. Koyo asserts that the appellate body found that Article 2.4.2 required a comparison of normal value with prices of all comparable export transactions and that the European Communities failed in not considering all transactions. Koyo comments further that the WTO appellate body found the zeroing practice to be in violation of Articles 2.4 and 2.4.2 because a comparison between export price and normal value that did not take into account all export prices was not a "fair comparison". Koyo also asserts that the reasoning in the appellate body's findings is equally applicable to administrative reviews. Koyo states that section 752 of the Act reinforces the notion that the Department should determine a dumping margin for each entry, be that margin positive or negative. Finally, Koyo asserts that, to ensure that U.S. practice is consistent with its obligations under the Antidumping Agreement, the Department should eliminate its practice of zeroing and include negative margins in the weighted-average margin calculation for the final results. Torrington rebuts that the decision of the WTO appellate body is a dispute-settlement decision and, as such, creates obligations only for the WTO members involved in the dispute. It contends that the decisions do not create binding precedent even if finalized and, in support of this proposition, it cites to various decisions, including India - Patent Protection for Pharmaceutical and Agricultural Chemical Products, WT/DS79/R, par. 7.30 (August 24, 1998). Torrington argues that, even assuming that the decision creates an obligation for the United States, Article 2.4.2 addresses how investigating authorities may calculate dumping margins during the investigation phase of a proceeding and, thus, the decision would only create an obligation for the Department with respect to investigations. Torrington asserts that, therefore, no inconsistency arises between the appellate body's interpretation of Article 2.4.2 and the methodology applied by the Department in the current reviews. Torrington adds that the distinction between the investigation phase of a proceeding and administrative reviews has been recognized in the past in a WTO panel decision and in the SAA. Department's Position: The Bed Linens Panel and Appellate Decisions concerned a dispute between the European Union and India. We have no WTO obligations to act. Therefore, we have continued the practice of using zero where the normal value does not exceed the export price or CEP in our calculations of overall margins for the final results. Quantity Credit-Note Matching Comment 39: NSK argues that the Department should "net" positive sales and negative returns that occur in the same month for transactions involving the same customer code, product code, and unit price in instances when the positive and negative quantities do not match. NSK argues that the Department currently nets negative returns to positive sales that occur in preceding months when these transactions involve the same customer code, product code, and unit price but not necessarily the same quantity. NSK asserts that the Department should modify the database to reflect the deletion of home-market sales based on returns occurring in the same month in instances when the quantities do not match exactly. Torrington asserts that the Department should reject NSK's request because the suggested programming modifications amount to an invitation to correct or supplement the company's response. It asserts, moreover, that the modifications are inadequately supported since NSK has not demonstrated any correlation between the transactions with different quantities. Department's Position: For these final results, we have modified our calculations to match the returns by sale date rather than by month in order to avoid inadvertently matching a return to a sale that occurred later in the same month. Furthermore, we have refined our return-matching methodology for NSK to use the following hierarchy: 1) we first attempted to match identical quantities to sales made prior to the return (NSK's proposal reflects, evidently, only an attempt to match identical quantities within the same month); 2) we then attempted to match returns to a sale that had a larger quantity than the return and that was made prior to the return; and 3) finally, we "netted out" the remaining returns against the remaining quantities of sales made prior to the return. In instances where the quantity of the return exceeded the reported quantities of sales made prior to the return, we deleted the remaining return. Bearings Produced By Other Manufacturers Comment 40: Torrington argues that NSK did not report the bearings produced by other manufacturers properly. NSK argues that it reported sales in accordance with the instructions provided by the Department in section B of the questionnaire and that to give Torrington's allegation any credibility at this stage in the proceeding would be a gross abuse of the administrative process. Department's Position: We have reviewed the record and are satisfied that NSK has reported the sales of bearings produced by other manufacturers properly. 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 _________ ____________________ Faryar Shirzad Assistant Secretary for Import Administration ____________________ Date