(65 FR 13359, March 13, 2000) A-580-815 A-580-816 AR 8/1/97-7/31/98 Public Document DAS III/9/BB/JHC/BH/MH MEMORANDUM TO: Robert S. LaRussa Assistant Secretary for Import Administration FROM: Joseph A. Spetrini Deputy Assistant Secretary for Import Administration SUBJECT: Issues and Decision Memorandum for the Administrative Reviews of Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products from Korea: August 1, 1997 through July 31, 1998. SUMMARY: We have analyzed the comments and rebuttal comments of interested parties in the fifth administrative reviews of the antidumping duty orders covering cold-rolled and corrosion-resistant carbon steel flat products from Korea. As a result of our analyses, 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 Issues and Decision Memorandum. Below is the complete list of the issues in these administrative reviews for which we received comments and rebuttal comments by parties: GENERAL COMMENTS 1. Normal Value A. Currency Conversions COMPANY-SPECIFIC COMMENTS Dongbu Steel Co., Ltd. ("Dongbu") 1. EP/CEP A. Calculation and Allocation of U.S. Indirect Selling Expense 2. COP/CV A. Use of Fourth Administrative Reviews Databases 3. Other AD Issues A. U.S. Date of Sale Pohang Iron and Steel Co., Ltd. ("POSCO"), Pohang Coated Steel Co., Ltd. ("POCOS"), and Pohang Steel Industries Co., Ltd. ("PSI") (collectively, "POSCO Group") 1. Facts Available A. Adjustment to Costs for Differences in Coating Weight B. Model Match Criteria/Yield Strength C. Weighted-Average Costs D. Cost Reconciliation 2. COP/CV A. Start-up Costs B. Major Input/Transactions Disregarded Rule C. Conversion Costs 3. EP/CEP A. CEP v. EP Classification 4. Other AD Issues A. Level-of-Trade B. Home Market Credit C. Home Market Downstream Sales Credit D. Product Characteristics for Downstream Sales E. Overrun Sales Union Steel Manufacturing Co., Ltd. ("Union") 1. Other AD Issues A. Home Market Credit Days B. Prime v. Non-Prime BACKGROUND: On September 8, 1999, the Department of Commerce ("Department") published the preliminary results of administrative reviews of the antidumping duty orders on cold-rolled and corrosion-resistant steel flat products from Korea. See Certain Cold-Rolled and Corrosion- Resistant Carbon Steel Flat Products from Korea, 64 FR 48767 (September 8, 1999) (preliminary results of antidumping duty administrative reviews) ("Preliminary Results"). The merchandise covered by these orders are cold-rolled and corrosion-resistant carbon steel flat products as described in the "Scope of the Reviews" section of the Federal Register notice. The period of review ("POR") is August 1, 1997 through July 31, 1998. We invited parties to comment on our preliminary results. DISCUSSION OF THE ISSUES: GENERAL COMMENTS 1. Normal Value Comment 1: Currency Conversions AK Steel Corporation; Bethlehem Steel Corporation; Inland Steel Industries, Inc.; LTV Steel Company; National Steel Corporation; and U.S. Steel Group - a unit of USX Corporation (collectively, "Petitioners") argue that the Tariff Act of 1930, as amended ("the Act"), instructs the Department to ignore fluctuations in exchange rates when converting foreign currencies into U.S. dollars and that, in its Preliminary Results, the Department "has chosen a definition of exchange rate fluctuation that selectively ignores some fluctuations and not others, in a manner which unreasonably disadvantages U.S. producers." See Petitioners’ Case Brief regarding the POSCO Group, at 15 (October 8, 1999) ("Petitioners’ POSCO Group Brief"). Therefore, according to Petitioners, the Department should apply an eight-week moving average exchange rate as the benchmark rate throughout the POR. The Pohang Iron and Steel Co., Ltd. ("POSCO"), Pohang Coated Steel Co., Ltd. ("POCOS"), and Pohang Steel Industries Co., Ltd. ("PSI") (collectively, "the POSCO Group") argues that the Department properly used actual daily rates for currency conversion purposes for sales occurring between November 1, 1997 and December 31, 1997, and properly used a modified benchmark rate for sales occurring after December 31, 1997 but before March 1, 1998. According to the POSCO Group, the modified benchmark applied by the Department is consistent with the Department’s Policy Bulletin and with prior Department cases, including Stainless Steel Sheet and Strip in Coils from Korea, 64 FR 15444 (March 31, 1999) (final determination of sales at less than fair value) ("SSSS from Korea") and Stainless Steel Plate in Coils from Korea, 64 FR 30664 (June 8, 1999) (final determination of sales at less than fair value) ("SSPC from Korea"). Union Steel Manufacturing Co., Ltd. ("Union") contends that the Department incorrectly executed the exchange rate methodology during the preliminary results. Union stated that it recreated the exchange rate methodology described in the preliminary results and then compared the results with another exchange rate program sent by the Department to Union. A comparison of the programs revealed that the two programs did not match. Union requests that the Department determine the basis for this error in the exchange rates when calculating Union’s dumping margins for the final results, and apply the appropriate exchange rates as described in the preliminary results for the final results. Department's Position: We agree in part, with the POSCO Group and Union. Section 773A(a) the Act directs the Department to use a daily exchange rate to convert foreign currencies into U.S. dollars, unless the daily rate involves a fluctuation. The Department considers a "fluctuation" to exist when the daily exchange rate differs from the benchmark rate by 2.25 percent or more. The benchmark rate is defined as the moving average of rates for the past forty business days. When we determine a fluctuation to exist, we generally substitute the benchmark rate for the daily rate, in accordance with established practice (an exception to this rule is described, infra). See Policy Bulletin 96- 1: Currency Conversions, 61 FR 9434 (March 8, 1996) (methodology explanation). Further, we also agree with Union, in that the Department inadvertently applied the improper exchange rate database in the preliminary results. For the final results, we will be using the exchange rate database as described below. Our analysis of dollar-Korean-won exchange rates show that the Korean won declined rapidly in November 1997 and December 1997. Specifically, the won declined more than forty percent over this two month period. The decline was many times more severe than any change in the dollar-won exchange rate during recent years, in both speed and magnitude, and it did not rebound significantly in a short time. As such, we determine that the decline in the won during this period was of such magnitude that the dollar-won exchange rate cannot reasonably be viewed as having simply fluctuated at that time, i.e., as having experienced only a momentary drop in value relative to the normal benchmark. Accordingly, the Department used actual daily exchange rates exclusively in November 1997 and December 1997. See SSSS from Korea. We note, however, that we have refined our methodology somewhat from that applied in the Preliminary Results and SSSS from Korea. We recognize that, following a large and precipitous decline in the value of a currency, a period may exist wherein it is unclear whether further declines are a continuation of the large and precipitous decline, or are merely fluctuations. Under the circumstances of this case, such uncertainty may have existed following the large, precipitous drop in November 1997 and December 1997. Thus, we have devised a methodology for identifying the point following a precipitous drop at which it is reasonable to presume that rates were merely fluctuating. Following the precipitous drop in November 1997 and December 1997, we continued to use only actual daily rates until the daily rates were not more than 2.25 percent below the average of the twenty previous daily rates for five consecutive days. At that point, we determined that the pattern of daily rates no longer reasonably precluded the possibility that they were merely "fluctuating." Using a twenty-day average for this purpose provides a reasonable indication that it is no longer necessary to refrain from using the normal methodology, while avoiding the use of daily rates exclusively for an excessive period of time. Accordingly, from the first of these five days, we resumed classifying daily rates as "fluctuating" or "normal" in accordance with our standard practice, except that we began with a twenty-day benchmark and on each succeeding day added a daily rate to the average until the normal forty-day average was restored as the benchmark. See Certain Welded Carbon Steel Pipes and Tubes from Thailand, 64 FR 56759, 56763 (October 21, 1999) (final results of antidumping duty administrative review). Applying this methodology in the instant case, we used daily rates from November 3, 1997 through January 13, 1998. We then resumed the use of our normal methodology, starting with a benchmark based on the average of the 20 reported daily rates from January 14, 1998. We used the normal 40-day benchmark from February 12, 1998 to the close of the review period. DONGBU 1. EP/CEP Comment 2: Calculation and Allocation of U.S. Indirect Selling Expense Dongbu Steel Co., Ltd. ("Dongbu") disagrees with the Department’s revision of the calculation of U.S. indirect selling expenses for the preliminary results regarding the allocation of a certain category of expenses to subject merchandise. Dongbu argues that the common "other" expenses should be allocated to both subject and non-subject merchandise on the basis of direct expenses. The Department had found in the preliminary results that Dongbu did not clarify or provide information requested on common "other" expenses. See Preliminary Results, 64 FR at 48771. Dongbu argues that, contrary to the Department’s determination, such information regarding common "other" expenses was on the record prior to the preliminary results. Dongbu argues that its August 24, 1999 letter to the Department provided clarification of expenses included in the common "other" expense category. Dongbu states that the expense categories are identifiable by comparing its reconciliation worksheet (for Dongbu USA’s SG&A expenses and financial statements, for fiscal year 1997 and the POR, submitted April 22, 1999 at Exhibit C-31) to the supplemental schedule of SG&A expenses in Dongbu USA’s financial statements (for fiscal year 1997, submitted October 23, 1998 at Exhibit A-17). To derive the common "other" expenses for Dongbu USA, Dongbu subtracted salary, payroll, medical and unemployment insurance, and rent, from the total general and administrative expenses. Dongbu then argues for a direct expenses allocation methodology. Dongbu argues that allocation of the common "other" expenses is in line with the Department’s general practice to allow companies to allocate expenses, provided the allocation methodology used does not cause inaccuracies or distortions. See Statement of Administrative Action ("SAA"), H.R. 5110, H.R. Doc. No. 103-316, at 823-24 (1994). Dongbu asks that the Department add the common "other" expenses to the other direct expenses, then allocate them over flat-rolled sales to calculate an indirect selling expense ratio. Dongbu believes that a sales value methodology is inappropriate, because expenses for flat- rolled sales are lower than for other products (Dongbu USA handles only the importation of flat-rolled products, where much of the work is performed by the customs broker. Other products take up more of Dongbu USA’s time because they involve more work, such as arrangement of ocean freight, and commission related work), as flat-rolled products have the highest sales value and require the least amount of work. However, in the alternative, Dongbu argues there is sufficient information on the record to allocate the common "other" expenses using a sales value methodology. Under a sales value methodology, common "other" expenses would be allocated over the total sales of all products, instead of only flat-rolled products. Petitioners claim that the Department correctly revised the calculation of Dongbu’s U.S. indirect selling ratio. Petitioners argue that, contrary to Dongbu’s claim, information on common "other" expenses is not on the record, as Dongbu failed to adequately respond to the Department’s original and two supplemental questionnaires requesting information on Dongbu’s indirect selling expenses. Petitioners argue that Dongbu’s response, that expenses "allocated to the different divisions within Dongbu U.S.A. are limited to salaries, salary-related expenses (e.g., payroll tax, medical insurance, unemployment insurance), and rent" is factually incorrect and that, in actuality, a majority of the expenses are related to common "other" expenses. Petitioners argue that Dongbu’s August 24, 1999 letter attempting to reconcile the 1997 "other" expenses to the 1997 financial statement fails to explain, item-by-item, the types of expenses incurred during the POR. Because the amount claimed by Dongbu as its "other" U.S. indirect selling expenses cannot be broken down into specific categories, Petitioners argue that these expenses could relate to any number of cost categories. Furthermore, Petitioners argue that the 1997 information provided by Dongbu has no bearing on the expenses incurred in 1998, which could differ from the expenses incurred in 1997. Accordingly, Petitioners argue that without itemization of the expenses and their values, the Department has no way of knowing how to allocate the expenses. As a result, Petitioners argue that the Department has no basis for applying either the direct expenses or total Dongbu USA sales allocation methodologies suggested by Dongbu, and argue that to adopt another approach from that used in the Preliminary Results could understate the costs. Petitioners suggest that the Department use an adverse facts available approach and assign the costs at issue to U.S. sales that have the greatest impact on the calculated margins. Department’s Position: The Department disagrees with Respondent. The Department requested information on and clarification of Dongbu’s U.S. indirect selling expenses in each of its questionnaires. However, none of Dongbu’s responses provided clarification of the various "other" expenses. Dongbu made a final attempt to sort out its U.S. indirect selling expenses a week before the preliminary results. In its letter to the Department, Dongbu asserted that the record contains information to define the expenses included in the common expense category. However, Dongbu’s brief explanation provided very little information or detail. Dongbu’s attempt at clarifying its previous submissions consisted of a short explanation stating that after deducting items specifically listed on its worksheet for fiscal year 1997, from its Supplemental Schedule of General and Administrative Expenses for fiscal year 1997, the categories which remained are associated with common expenses. See Dongbu’s letter to the Department, at 3-4 (August 24, 1999). Dongbu reiterates this clarification in its brief in a similar, albeit longer, argument. See Dongbu’s October 8, 1999 Case Brief ("Dongbu’s Brief"), at 2-6. However, while Dongbu’s argument is more extensive, it does not serve to make the information Dongbu previously provided any more detailed. No matter how Dongbu chooses to phrase its explanatory text, the information provided by Dongbu on its common expenses remains lacking in specificity. Furthermore, while Dongbu attempted to clarify the general categories for its 1997 "other" U.S. indirect selling expenses, the Department has no record evidence that these categories remain the same for 1998. Thus, the Department is unable to ascertain how much, if any, of these 1997 "other" U.S. indirect selling expenses were accrued during the POR. Moreover, Dongbu provides no breakdown of the expenses for 1998; thus, the Department cannot determine the values for the "other" indirect selling expenses accrued during the 1998 portion of the POR. Further, without the requested information, the Department is unable to address Dongbu’s request for a direct expenses allocation methodology or its suggested alternative sales value methodology. However, the Department disagrees with Petitioners’ suggestion to use adverse facts available because Dongbu provided detailed information regarding all facts of the indirect selling expense calculation except this one. For that reason, we find that an adverse inference, beyond that contained in the preliminary results, is not warranted. Therefore, for these final results, the Department is maintaining its calculation from the preliminary results, allocating common "other" expenses to subject merchandise. 2. COP/CV Comment 3: Use of Fourth Administrative Reviews Databases Dongbu argues that the Department should use the COP and CV databases from the fourth administrative reviews when performing the cost test on sales that occurred in April through June 1997, the three-month contemporaneity period reported in these fifth administrative reviews. Dongbu argues that because the severe financial crisis in Korea, starting in November 1997, resulted in the rapid devaluation of the Korean won, it is distortive to test the home market prices for sales made before the POR, in the second quarter of 1997, with costs incurred in the second half of 1997 and the first half of 1998, or to determine normal value for those sales based on constructed value using costs incurred in the July 1997 through June 1998 period. Dongbu argues that using the COP and CV data from the fourth administrative reviews to match with the April through June, 1997 sales would be proper in performing the cost test on these sales. In support, Dongbu argues that the COP and CV data has already been placed on the record, and was used in the cost test for the same sales in the fourth administrative reviews final results. Dongbu also argues that the Department is obligated to ensure that calculations are based on costs that reasonably reflect and accurately capture all of the actual costs incurred. See SAA at 835. Finally, Dongbu asserts that calculating a single average cost for the POR is not required, and that the Department has departed from this practice in the past, where cost and price averages calculated over the entire period did not permit an appropriate comparison. See, e.g., Dynamic Random Access Memory Semiconductors of One Megabit and Above from the Republic of Korea, 58 FR 15467, 15476 (Mar. 23, 1993) (final determination of sales at less than fair value) ("DRAMs from Korea"); Erasable Programmable Read Only Memories (EPROMs) from Japan, 51 FR 39680, 39682 (Oct. 30, 1986) (final determination of sales at less than fair value) ("EPROMs from Japan"). Petitioners argue that the Department should not use data from the fourth administrative reviews. Petitioners argue that most instances in which the Department has departed from using a 90/60 day window have involved situations where significant expenses relating to production are incurred over a prolonged period of time. See Christian Marsh and John J. Miller, Use and Measurement of Production Costs Under U.S. Antidumping Law, U.S. Dept. of Commerce, at 33-34 (Sept. 19, 1995). Additionally, Petitioners contend that the cases cited by Dongbu offer no support for departing from this rule. In DRAMs from Korea, the Department based home market value on weighted- average monthly prices, and did not use COP information from a previous review. Petitioners argue that the circumstances in EPROMs from Japan, that led to matching U.S. sales prices with costs from three months prior to the date of sale, do not exist in the case at hand. The industry in EPROMs from Japan was subject to significant increases/decreases in the cost of production within a short period of time due to technological advancements; Petitioners argue that technological advancements in the steel industry generally occur over years. Finally, if the Department agrees with Dongbu’s argument that the cost test is distorted for certain sales due to the devaluation of the Korean won, Petitioners suggest an alternative to using data from the fourth administrative reviews. Petitioners argue that the Department should merely disregard those U.S. sales that require the use of home market sales outside the POR in conducting the cost test. Department’s Position: The Department disagrees with Dongbu’s argument that data from the fourth administrative reviews period should be used due to the decline in the Korean won during the POR. It is not the Department’s practice to use data from a previous review when faced with fluctuations in the exchange rate. We agree with Petitioners in noting that the DRAMs from Korea and EPROMs from Japan cases cited by Dongbu in its argument are distinguishable from the case at hand because the Department did not depart from the POR in DRAMs from Korea, and unlike the semiconductor industry in EPROMs from Japan, the steel industry is not characterized by "technological advancements and rapid changes in the production process." See EPROMs from Japan, 51 FR at 39682. However, we disagree with Petitioners’ suggestion to disregard U.S. sales impacted by the decline in the Korean won in the Department’s cost test. The Department notes that Petitioners did not argue that similar U.S. sales for the POSCO Group or Union should be disregarded. As discussed, supra, the Department is using a benchmark rate in lieu of the daily rate for part of the POR, consistent with our established practice. Accordingly, the Department is making adjustments to the methodology used in the Preliminary Results, and modifying its exchange rate dataset. See Comment 1: Currency Conversion, supra. 3. Other AD Issues Comment 4: U.S. Date of Sale Dongbu argues that the Department used the incorrect date of shipment for determining the U.S. date of sale (defined as the earlier of invoice date to the unaffiliated U.S. customer or shipment date from Korea) and the credit expense for U.S. sales. For the preliminary results in these fifth reviews, the Department used the bill of lading date as the date of shipment from Korea. Dongbu argues that, in previous reviews, the Department used the date of shipment from the factory as the date of shipment from Korea. Accordingly, Dongbu asks that the Department replace the bill of lading date (SHIPDT1U) with the date of shipment from the factory (SHIPDT2U) in its margin calculation, to be consistent with prior reviews. Petitioners provided no comment on this issue. Department’s Position: The Department disagrees with Dongbu’s request. In the final results in the fourth administrative reviews, contrary to Dongbu’s argument, the Department used the bill of lading date as the date of shipment. See Analysis Memo from Steve Bezirganian and Juanita Chen to The File, at B.1. and B.2. (March 8, 1999). Dongbu has not presented the Department with any facts relating to the date of shipment, nor any change between the fourth reviews and these reviews, to warrant changing the margin calculation to use date of shipment from the factory. Therefore, the Department is retaining the bill of lading date as the date of shipment in its margin calculation program. THE POSCO GROUP 1. Facts Available Comment 5: Adjustments to Costs for Differences in Coating Weight Petitioners argue that the Department’s revision to the POSCO Group’s submitted costs to account for differences in coating weight "significantly favors POSCO, due to the specific characteristics of the products sold in the United States during the period of review." See Petitioners’ POSCO Group Brief, at 27-28. Petitioners state that identifying the unique costs associated with different coating weights is an appropriate adjustment, but it argues that in these fifth administrative reviews, the POSCO Group is being rewarded for its failure to report its costs on a CONNUM-specific basis. Therefore, according to Petitioners, the Department should remove this adjustment. The POSCO Group argues that since the Department has deemed this adjustment necessary in prior reviews, it would be "arbitrary and irrational" to remove this adjustment simply because it favors POSCO. Therefore, the POSCO Group states that the Department should reject Petitioners’ claim. Department's Position: We agree with the POSCO Group. The adjustment made by the Department for differences in coating weight reflects a methodology comparable to that used in the final results of the second, third, and fourth administrative reviews. The Department continues to find it appropriate to account for significant cost differences associated with differences in physical characteristics, regardless of its effect on the margin. Comment 6: Model Match Criteria/Yield Strength Petitioners argue that the Department should apply total facts available because the POSCO Group failed to report its yield strength accurately. Petitioners contend that, according to the Department’s verification report, the Department was not able to confirm the POSCO Group’s yield strength product characteristic information for one of the five home market sales examined for product characteristic reporting. Petitioners note that yield strength is ranked fifth out of the Department’s fourteen matching criteria for cold-rolled products and sixth out of twelve matching criteria for corrosion- resistant products. Specifically, petitioners point to home market sales observation 12, for which the POSCO Group reported a yield strength of "A" (less than 35,000 psi). Petitioners state that when the Department asked why it reported an "A" for this sale of full- hard coils, the POSCO Group explained that the specification in question, the internal POSCO Group specification guidelines, and the mill test certificates did not specify a minimum yield strength, and thus it assigned a yield strength of the CONNUM most similar to that CONNUM. When requested to identify all CONNUMs for which a surrogate yield strength was utilized, Petitioners note that "the information POSCO provided in response was found to have errors." See Home Market Sales Verification Report from Steve Bezirganian, Becky Hagen, and Marlene Hewitt through James C. Doyle to the File, at 15-16 (August 10, 1999) ("HM Sales Verification Report"). Petitioners argue that the Department has no basis for accepting the POSCO Group’s yield strength information for any of its sales, and without accurate yield strength data, the Department is prohibited from accurately matching U.S. and home market sales and ensuring proper CONNUM reporting. Therefore, Petitioners argue that the Department should apply total facts available. Respondent argues that the Department correctly utilized the reported yield strength data in the Preliminary Results. The POSCO Group states that the Department verified the reported product characteristics, including yield strength, for numerous observations and the characteristics verified completely. Respondent notes that, during verification, the Department found one incorrect reported yield strength but the POSCO Group explained that was due to a clerical input error. Additionally, the POSCO Group states that the Department also verified the first three steps of the POSCO Group’s cascading methodology, which the POSCO Group used "for the vast majority of observations." However, because the POSCO Group was not able to identify the minimum yield strength for a few observations using the first three steps of its cascading methodology, the last step of the POSCO Group’s methodology for reporting minimum yield strength was used. In this final step, the POSCO Group reported the minimum yield strength of the CONNUM most similar to the CONNUM for which no minimum yield strength could be reported under the first three steps of its cascading methodology. The POSCO Group states that its yield strength reporting methodology is reasonable and appropriate. Respondent asserts that Petitioners’ claim, that "the information in Sales Verification Exhibit 47 ‘was found to have errors’ and therefore the Department has no way of knowing POSCO’s reported yield strengths for all observations" is not justified because the one mistake found using this surrogate methodology was due to a clerical error. Moreover, Respondent argues that there is no basis for applying total facts available because it has cooperated fully, has not refused to provide information, has not failed to cooperate to the best of its ability, and has no pattern of discrepancies that would render its data unusable. Therefore, the Department should reject Petitioners’ argument for using total facts available. Department’s Position: We agree with Petitioners, in part. The first time the Department received an explanation of the POSCO Group’s cascading methodology in its entirety was during the home market sales verification, while the Department was in the process of verifying the product characteristics of five home market sales observations. See Home Market Sales Verification Report at 15-16. During the Department’s review of home market sales observation number 12, the Department determined that this observation had a reported minimum yield strength which was inconsistent with the supporting sales documentation provided by the POSCO Group. This sale consisted of full-hard coils, for which the appropriate minimum yield strength is "C," not "A," the yield strength that the POSCO Group had reported for this observation. When the Department asked the POSCO Group for an explanation of this apparent discrepancy, the POSCO Group explained its cascading yield strength methodology in its entirety. Additionally, the fact that the use of this surrogate specification would have still resulted in an incorrect minimum yield strength for this particular observation (because the surrogate specification assigned to this particular specification had a minimum yield strength of "B," not "C," as required by the Department’s questionnaire for all full-hard products), reflects that the POSCO Group’s surrogate methodology is faulty, at least for this particular CONNUM. Since the POSCO Group stated in their December 4, 1998 Section B Response at page 21 that "full-hard is coded as ‘B,’ not annealed," the Department was able to further analyze the home market sales data for the POSCO Group’s full-hard products by reviewing those observations with a reported annealing characteristic of "B" and an incorrect reported minimum yield strength characteristic of either "A" or "B." The Department was not able to identify any additional observations for which the reported yield strength characteristic is inconsistent with the reported annealing characteristic. Therefore, we are applying adverse facts available because the POSCO Group’s explanation for assigning the missing yield strength characteristic was inconsistent with their yield strength methodology presented at verification. Thus, the Department has assigned the affected CONNUM the highest adjusted total cost of manufacturing, because the incorrectly reported yield strength for this particular CONNUM may have distorted product concordance by matching this full-hard product to a much weaker product, when it should have matched up to a full-hard product if the yield strength characteristic had been properly coded. See the Final Results Analysis Memorandum for the POSCO Group (March 6, 2000). Comment 7: Weighted-Average Costs Respondent argues that the Department mistakenly increased POSCO’s reported costs, and the Department should eliminate the adjustment in its Final Results. POSCO contends that the Department increased costs because it believed that POSCO did not account for a specific Representative Product Group ("RPG") in two of its cost centers that produced the RPG. However, Respondent argues that its reported costs accurately included the cost for both cost centers in which the RPG was produced. Respondent states the Department adjusted POSCO’s reported costs by comparing the weighted-average cost of manufacturing RPG in the two aforementioned cost centers to cost in another single cost center. POSCO noted that this adjustment is in error because the Department believed that the reported costs for this specific RPG did not reflect costs incurred in the two cost centers. However, Respondent argues that an examination of the verification exhibits demonstrates that POSCO’s submitted costs for this RPG included costs from both cost centers. Respondent states that the RPG in question was manufactured at the Pohang works in the first half of 1998 and was part of a larger group. Additionally, Respondent states that Cost Verification Exhibit 24 demonstrates its calculation of the adjusted cost used in the calculation for this RPG. See Cost Verification Report from Steve Bezirganian, Becky Hagen, and Marlene Hewitt through James C. Doyle to the Edward Yang, at Exhibit 24 (August 5, 1999) ("Cost Verification Report"). POSCO notes that the ratio used to adjust the standard RPG to the fully absorbed costs in the financial statements is at page 74 of the exhibit. Further, Respondent states that using the different column totals (e.g., "HB-total") from a variety of exhibits and then applying the adjustment factor from page 73 of Cost Verification Report Exhibit 24 to the unpacked adjusted standard RPG cost will result in an actual unpacked RPG unit value. Furthermore, Respondent argues that the Department verified the actual packed RPG unit value, and the only difference between the Department’s verified packed RPG unit value and the weighted-average adjusted standard RPG value (from the Department’s Preliminary Results Analysis Memorandum (August 31, 1999)) was a packing value, and if the packing value was subtracted from the standard RPG value, the figures would match. Thus, Respondent concludes that its reported costs for this RPG was correct. Additionally, Respondent states that the Department’s verification outline instructed POSCO to prepare sample cost buildups, in order to demonstrate how its process-based costs are accumulated through the various processes to derive a finished product cost, including the product at issue, for the first half of 1998. Respondent prepared sample traces based on standard costs for each mill, and believed that these traces were responsive to the Department’s instruction. POSCO noted that it has used sample buildups in this format for submission and verification purposes in prior administrative reviews, and was never found to be non-responsive or deficient. Further, Respondent states that the Department verified that accumulated costs are transferred to subsequent cost centers, and an additional cost buildup would have been identical to the one already presented, with one additional cost center showing the transfer of product from one cost center to another cost center. Thus, POSCO argues that the majority of costs incurred in both cost centers were accounted for in the prepared cost buildups. Additionally, Respondent contends that the cost buildups mentioned in the verification report were not directly used by POSCO in the calculation of costs presented to the Department, but were prepared for informational purposes only, to show how costs were accumulated in the system. Finally, Respondent asserts that the Department verified that POSCO’s reported costs are based on actual costs and not standard costs. Therefore, Respondent argues that all costs incurred at POSCO in manufacturing the subject merchandise have been included in the costs reported to the Department and no adjustments are necessary. Petitioners argue that POSCO’s reported costs should be adjusted because the Department found errors in POSCO’s cost reporting system at verification. Thus, Petitioners claim that the Department should modify POSCO’s submitted costs to reflect its findings at verification. Department’s Position: We disagree with Respondent. The Department stated, in its cost verification outline, that "complete supporting documentation must be prepared in advance" for various CONNUMs and RPGs, including the RPG in question. See Cost Verification Report, at 2-3. The unadjusted RPG buildups for the first half of 1998, for this RPG, provided by Respondent at verification and appearing as Cost Verification Exhibits 33 - 35, do not include the quantity of the RPG produced at one cost center, as noted in the Cost Verification Report. "No reference to output of this RPG from the . . . line is included in Cost Verification Exhibits 33 through 35 in the Pohang cost buildups for the unadjusted RPG for the first half of 1998." See Cost Verification Report, at 42. While such unadjusted RPG buildups "were not directly used in the calculation of costs for the Department," according to Respondent, these buildups "are . . . the basis for the adjusted RPG cost calculations" Id. at 43. Furthermore, as noted in the Cost Verification Report, the POSCO Group stated at verification "that the cost centers used for the adjusted RPG calculations that are the basis for the data reported to the Department (after adjusting for packing and non-process specific variances . . .) are the same cost centers that are identified in individual unadjusted RPG cost buildups (e.g., see Cost Verification Exhibit 15, at 1)." Id. at 44. Thus, for the RPG in question, the POSCO Group failed in its cost calculations to identify a cost center at which a small quantity of the output of the RPG in question was produced. Id. at 42. Moreover, for another unadjusted RPG cost buildup, more than one final output cost center is listed. See Cost Verification Exhibit 15. Additionally, we note that the quantity associated with the missing cost center was extremely small, relative to the quantity for the cost center used in the cost buildups. In the Preliminary Results, the Department calculated a new adjusted first half 1998 cost for the RPG in question by weight- averaging the adjusted per ton cost for the RPG for the cost center identified in Cost Verification Exhibits 33 - 35 (see Cost Verification Exhibit 53, at 3) with the adjusted per ton cost for the RPG for the cost center not included in the buildups in Cost Verification Exhibits 33 - 35 (see Cost Verification Exhibit 53, at 4) to determine the effect of the POSCO Group’s exclusion of the latter cost center. Therefore, for these final results, we have continued to apply facts available, and have adjusted costs for all CONNUMs (i.e., corrosion-resistant steel and cold-rolled steel). This treatment should not be construed as adverse facts available, as we are simply applying a correction factor calculated for one of the primary RPGs focused on at verification to the entire database. Comment 8: Cost Reconciliation Petitioners argue that a sampling of the POSCO Group’s reported cost data reveals an "unusually large proportion of CONNUMs with inaccurately reported cost data" and, therefore, the Department should apply an adverse facts available adjustment to the POSCO Group’s submitted costs. According to Petitioners, the Department randomly selected two CONNUMs from the POSCO Group’s April 2, 1999 submission (which was the POSCO Group’s first response using the revised reporting methodology of calculating weighted-average costs on a CONNUM-specific basis) and in the Department’s April 21, 1999 Supplemental Questionnaire, requested that the POSCO Group provide worksheets which demonstrated the calculation of costs for the two CONNUMs under both the old and new methodologies. Petitioners argue that the POSCO Group, in its May 25, 1999 Response to the Department’s April 21, 1999 Supplemental Questionnaire, stated that it had identified certain errors and subsequently revised the submitted costs for these two CONNUMS dramatically upwards. According to Petitioners, the POSCO Group stated that it had corrected the errors in its CONNUM-specific reporting methodology subsequent to the Department’s April 21, 1999 Supplemental Questionnaire, but the Department’s random selection of another four CONNUMs for sampling at the cost verification demonstrates that the POSCO Group had, in fact, failed to correct the errors in its reporting methodology. Petitioners further argue that these findings are not isolated errors, but rather a systemic problem due to the POSCO Group’s cost reporting methodology. Petitioners allege that each time the Department requested additional information on specific CONNUMs, the POSCO Group responded by revising the submitted costs, which they did on three occasions (two times prior to the Department’s cost verification, and once after the cost verification). They state that it is the Department’s normal practice to "adjust the untested portion of the data in line with the verified findings based on facts available." Petitioners state that the Department should thus adjust all of the POSCO Group’s reported costs by the simple-average difference in the pre-verification and post-verification cost datasets for the four sampled CONNUMs. The POSCO Group argues that its submitted costs are accurate and have been thoroughly verified, both in these reviews and in other recent cases, including SSPC from Korea and SSSS from Korea. The POSCO Group states that the Department’s Cost Verification Report indicates no methodological errors or distortions in the POSCO Group’s submitted costs, and "Petitioners themselves do not even attempt to point to any." See the POSCO Group’s Rebuttal Brief, at 20 (October 15, 1999) ("the POSCO Group’s Rebuttal"). According to the POSCO Group, of the four CONNUMs sampled at the cost verification, the POSCO Group only revised one CONNUM in its post- verification cost dataset. The POSCO Group argues that the fact that only one CONNUM was revised demonstrates that the error was not systemic, and furthermore, this change in costs for that one particular CONNUM was solely due to one of the changes they presented at the beginning of the cost verification concerning the misallocation of the zinc scrap credit. Also, the POSCO Group states that Department verified total reported costs, by first reconciling the adjusted RPG values used to calculate the CONNUM-specific costs to the corresponding inventory product groups, which the Department then reconciled to the audited financial statements. The POSCO Group states that just because it revised the submitted costs for a few CONNUMs, there is no reason to believe that all their reported costs are inherently unreliable. Since the Department verified the corrections to the particular CONNUMs cited by Petitioners, and did not note any systemic flaws in the cost data, the POSCO Group argues that no adjustment to the submitted data is justified. Department's Position: We agree with Petitioners, in part. Regarding the POSCO Group’s revision to its submitted costs for one out of the four CONNUMs cited by Petitioners, the POSCO Group adequately explained at the cost verification how the misallocation of zinc scrap revenue to that specific CONNUM could result in such a distorted cost. The POSCO Group’s correction of this error was fully verified and the corrections were accepted for the record of these cases. However, while reconciling the POSCO Group’s home-based product group ("HBPG") costs, which reflect the value of the POSCO Group’s products at a broader level as they enter finished goods inventory, to the adjusted representative product group ("RPG") costs at the cost verification, the Department identified a number of RPGs with unusually low RPG costs in Cost Verification Exhibit 24 (which is a printout indicating the relationship between adjusted (not packed) RPGs to HPBG actual costs). We chose one RPG with an unusually low cost to review and the POSCO Group provided the adjusted Product Cost Sheet for that particular RPG, which is Cost Verification Exhibit 52. The adjusted Product Cost Sheet indicated that the unusually low adjusted RPG cost of this product resulted from almost all of the output being generated as severely defective merchandise (the lowest grade of non-prime). The POSCO Group explained that they determined the scrap offset by the following formula: estimated prime grade cost times defective grade selling price divided by prime grade selling price. For this RPG, the scrap offset was greater than that of the unit cost of the incoming coils, and therefore, resulted in an unusually low RPG cost. See Cost Verification Report at 28-31. Because many RPG costs were found to be unusually low at the cost verification, which the POSCO Group explained was a result of their scrap offset calculation, the resulting total cost of manufacturing ("TOTCOM") for the CONNUMs associated with these low cost RPGs are also unusually low. The Department has further analyzed the POSCO Group’s scrap offset calculation for these Final Results, and has determined that the POSCO Group’s scrap offset calculation leads to underestimated TOTCOMs. This is because the scrap offset can be greater than the unit cost of the incoming coils, and therefore the total RPG cost of these coils, which are then allocated to the various CONNUMs, leads to unusually low TOTCOMs as well. In the instances where the reported TOTCOMs were too low, the reported cost components in the corrosion-resistant database are also understated. This is due to the fact that the POSCO Group subtracted average unit costs for the first three stages of the production process (slab, hot-rolling, and cold-rolling) from the understated TOTCOMs to calculate the costs at the coating stage. This understatement of the TOTCOMs may have been caused by the scrap offset, which the POSCO Group explained for the first time at the cost verification, and which the Department has determined to be inappropriate. The Department, therefore, has further analyzed the reported cost of production data for these Final Results, and has determined that there are a large number of CONNUMs with reported cost components at the coating stage which appear to be distorted. In the Department’s Final Results Adjustment Program (which precedes the Arm’s Length, Model Match, and Margin Calculation Programs), the Department inserted programming to determine how many CONNUMs in the cost of production database had negative material, variable overhead, and/or fixed overhead components, as well as positive scrap values. Many of the incorrect cost components occur at the coating stage, which the Department has determined may be the result of the POSCO Group’s incorrect scrap offset methodology mentioned above. Because material, variable overhead, and fixed overhead should all be positive values due to the fact that the POSCO Group incurs costs for these various inputs and expenses, and since the POSCO Group has not addressed why it is appropriate for these cost components to be negative values, the Department has applied the weighted-average TOTCOM of those CONNUMs that did not have incorrect cost components to those CONNUMs in the cost of production database that have negative material, variable overhead, and fixed overhead components. In addition, the Department analyzed how many CONNUMs had a positive scrap revenue component in the cost of production database. Since there is scrap generated at each of the four processing stages which the POSCO Group then sells and receives revenue, there should be a negative value for each scrap component. However, in many instances, the scrap component is a positive value. Since the POSCO Group did not explain why many of their scrap values are positive, the Department has applied, as facts available, the weighted-average TOTCOM of those CONNUMs that did not have incorrect cost components to those CONNUMs with positive scrap values in the cost of production database. See Section 776(a)(1) of the Act. The POSCO Group did explain that in some cases, the value for yield may be either positive or negative, because "in some circumstances, less than one ton of base material is needed to produce one ton of CORE product." Therefore, the Department is not applying facts available to those CONNUMs with positive yield values. See the POSCO Group’s December 4, 1998 Section D Response at 41. Section 776(b) allows the Department to use an inference that is adverse to the respondent, if it finds that the "interested party has failed to cooperate by not acting to the best of its ability to comply with a request for information." However, because we were able to verify the POSCO Group’s overall costs at the cost verification and the POSCO Group has fully cooperated with all the Department’s requests in this review, the Department has chosen not to apply adverse facts available as requested by Petitioners. Nevertheless, since the POSCO Group’s scrap offset methodology appears to understate TOTCOM which in turn leads to nonsensical cost components, as non-adverse facts available, we have applied the weighted-average TOTCOM of those CONNUMs which did not have incorrect cost components. In addition, we have determined that the home market variable cost of manufacturing ("VCOMH") and US variable cost of manufacturing ("VCOMU") may be similarly distorted, as they are comprised some of the same components which make up TOTCOM and which we have determined to be incorrect as discussed above. Therefore, as facts available, we have set the difference of merchandise adjustments ("DIFMER") to zero for the home market models that are comprised of the CONNUMs with incorrect cost components. For a more detailed description of the proprietary information contained in our analysis, please see the Final Analysis Memorandum dated March 6, 2000. 2. COP/CV Comment 9: Startup costs The POSCO Group claims that the Department erroneously failed to grant its requested startup adjustment because the Department claimed the new line was one of many necessary production steps required to produce corrosion-resistant products. The POSCO Group argues that record evidence supports its assertion that it has satisfied the requirements for a startup adjustment, as defined in section 773(f)(1)(C)(ii) of the Act. Specifically, the POSCO Group claims that abnormally high production costs were incurred at a new facility during the review period due to startup operations, and that these higher costs resulted from production levels that were limited by technical factors associated with the initial phase of commercial production. The POSCO Group claims that establishing its new production line required substantial effort and investment, and that the new line was still in the startup phase from July 1997 through December 1997. During this initial phase, the POSCO Group argues that it was necessary to carefully monitor and analyze the output from its new line to ensure that quality standards were met, before increasing to commercial production levels in December 1997. The POSCO Group notes that it limited production to well below the line’s normal capacity during the startup period to permit such monitoring and to allow calibration of the new equipment to avoid output variations. According to the POSCO Group, this limiting of production levels reflects the impact of technical factors as defined in the Act. The POSCO Group claims that record evidence also demonstrates that the new line’s production levels were not limited by any factors other than startup, since demand for its products was strong, no chronic production difficulties were experienced, and operating performance on the new line improved steadily throughout and subsequent to the startup period. Finally, the POSCO Group argues that the Department has granted start-up adjustments without regard to whether new products were manufactured in the new facility. See Notice of Preliminary Results of Antidumping Duty Administrative Review and Intent to Revoke Order; Brass Sheet and Strip from the Netherlands, 64 FR 48760 (September 8, 1999). Petitioners argue that the Department properly rejected the POSCO Group’s startup adjustment claim in its Preliminary Results. Petitioners assert that the startup adjustment claim should continue to be rejected because "the facts on the record easily demonstrate that POSCO failed to satisfy both prongs." See Petitioners’ Rebuttal Brief regarding the POSCO Group, at 23 (October 15, 1999) ("Petitioners’ POSCO Group Rebuttal"). According to Petitioners, the first prong of the SAA test regarding start-up adjustments requires the POSCO Group to demonstrate that its start-up operations produce a "new product" that requires substantial investment, or constitute "new production facilities." Since the new line produces corrosion- resistant products, which are not new products, Petitioners argue that in order to satisfy the first prong, the start-up operations must constitute new production facilities. According to Petitioners, the setting-up of the POSCO Group’s new line did not involve the replacement of nearly all production machinery, or the equivalent rebuilding of existing machinery, which they argue is necessary under the first prong of the Statute. Petitioners further argue the POSCO Group has failed to meet the second prong of the test as well, because the POSCO Group has not provided sufficient information for the Department to determine that technical factors were responsible for limiting production. Finally, Petitioners argue that, in Brass Sheet and Strip from the Netherlands, the Department determined that the "respondent’s new casting strip-casting line represented an ‘investment in new technology,’ and that it ‘entirely replaced’ its existing casting mill," facts which are not present in the instant case. Department's Position: In its Preliminary Results, the Department found that the statute’s requirements for granting a startup adjustment were not met by the POSCO Group, and we, therefore, did not apply the startup adjustment to calculate the POSCO Group’s cost of production ("COP") and constructed value ("CV"). For these Final Results, the POSCO Group has claimed that the installation of a new production line at one of its two works constitutes a new facility at which new products are manufactured, and that the claimed startup adjustment should be applied to products manufactured on this new line. The POSCO Group also claimed a startup adjustment for this same production line in the previous review. See, e.g., Final Results of Antidumping Administrative Reviews; Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products from Korea, 64 FR 12927 at 12949-12950 (March 16, 1999) ("Korean 4th Reviews Final"). As in that review, we find that this new line does not produce a "new product," and does not constitute a "new production facility," as required by the startup adjustment provision. See section 773(f)(1)(C)(ii)(I) of the Act. As the Department stated in the Preliminary Results and in the fourth reviews, the SAA sets a high standard for startup adjustment claims when it states that, "{n}ew production facilities includes the substantially complete retooling of an existing plant. Substantially complete retooling involves the replacement of nearly all production machinery or the equivalent rebuilding of existing machinery." See SAA at 836 (emphasis added). The implication of this language are that the startup adjustment should only be applied when substantial modifications have been made to an entire production plant. When determining whether substantial modifications have been made, the Department must consider, along with other factors, the extent to which the improvements relate to the total production process. In the instant case, the new line is but one of many processing steps necessary to produce corrosion-resistant products performed by the POSCO Group. We also note that, although the equipment in question is large and expensive, its relative size to the other production equipment involved in the production of corrosion-resistant products at the POSCO Group is small. Therefore, we do not believe that the installation of this equipment constitutes the substantial retooling of one of the POSCO Group’s facilities and, therefore, does not meet the standard established in the statute of a "new production facilities." See section 773(F)(1)(C)(ii)(I) of the Act. Because section 773(f)(1)(C) of the Act establishes that both prongs of the test must be met before a start-up adjustment is warranted, this finding that the new line does not constitute a substantial retooling of an existing facility is sufficient to deny the POSCO Group’s claim. Accordingly, because the first prong of the test has not been met, we need not address the POSCO Group’s and Petitioners’ arguments concerning the technical factors that limit commercial production levels. Comment 10: Major Input/Transactions Disregarded Rule Petitioners argue that the Department inappropriately failed to apply the major input rule (section 772(f)(3) of the Act) to transactions between POSCO and its affiliated parties. Petitioners state that the Department, in its antidumping duty questionnaire, asked POSCO to provide information on transfer price, COP, and fair value for major inputs transferred between affiliated parties. Petitioners note that two of POSCO’s affiliates, POCOS and PSI, purchased major inputs from POSCO during the review period, and thus the major input rule must be applied to these transfers. Petitioners also argue that, since the POSCO Group failed to provide the requested information, the Department should use facts available to value transfers between POSCO and its affiliates. The POSCO Group argues that, in these and the three prior reviews, the Department has already rejected Petitioners’ arguments to apply the major input rule. See Korean 4th Reviews Final at 12948. The POSCO Group states that the Department has "collapsed" POSCO, POCOS, and PSI into a single entity for the third successive review, indicating that it is now well-settled Department practice to not apply the major input rule to transactions within a single collapsed entity. The POSCO Group argues that, in such circumstances, it is consistent with the statute to not apply the major input rule because the statute requires application of the rule only to transactions between persons. Department's Position: We agree with the POSCO Group. It is now well-settled Department practice not to apply section 773(f)(2) and (3) of the Act to transfers within a collapsed entity. Rather, because we are treating POSCO and its affiliated producers as a single producer for purposes of the antidumping analysis, we find it appropriate to value the substrate inputs at issue according to POSCO Group-wide weighted- average costs, just as we attribute all POSCO Group home market and U.S. sales to the entity as a whole. As the Department stated in the third reviews, "the decision to treat affiliated parties as a single entity necessitates that transactions among the parties also be valued based on the group as a whole and, as such, among collapsed entities the fair-value and major-input provisions are not controlling." See Final Results of Antidumping Administrative Reviews; Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products from Korea, 63 FR 13185 (March 18, 1998) ("Korean 3rd Reviews Final"). See also Final Determination of Sales at Less Than Fair Value; Stainless Steel Wire Rod from Korea, 63 FR 40408 (July 29, 1998). The POSCO Group did not provide the data related to the major input issue in response to the Department’s original questionnaire, and the Department did not request that information in its supplemental questionnaires, consistent with the Department’s determination in the Korean 3rd Reviews Final that such information was not needed. The Court of International Trade ("CIT") recently affirmed this practice, holding that "Commerce reasonably determined that it should act consistently with its collapsing determination and not apply inconsistent solitary provisions, thereby arbitrarily increasing respondents’ liability." AK Steel Corp. et al. v. United States, 34 F.Supp.2d 756, 765 (Ct. Int’l Trade 1998), aff’d slip op. 99-1296 (Fed. Cir. 2000). Comment 11: Conversion Costs Petitioners argue that the Department failed to include the POSCO Group’s reported foreign exchange costs in the calculated cost of production. The POSCO Group did not comment on this issue. Department’s Position: The Department agrees with Petitioners, and has therefore revised its calculation of the cost of production to include foreign exchange costs. 3. EP/CEP Comment 12: CEP v. EP Classification The POSCO Group argues that the Department was correct in categorizing U.S. sales by Pohang Steel Industries Co., Ltd. ("PSI") as export price ("EP") sales, but it erred in categorizing POCOS’ and POSCO’s U.S. sales as constructed export price ("CEP") sales. The POSCO Group argues that such classification is unsupported by the record evidence and contrary to consistent Department practice. The POSCO Group indicates that the functions of the U.S. affiliates were limited to those of processors of sales-related documentation and communications links with the unaffiliated U.S. buyers, and that their role was only ancillary in the sales process. The POSCO Group argues that POSCO and POCOS either approved or disapproved primary terms of sale for U.S. customers, and that the U.S. affiliates did not set prices with those unaffiliated U.S. customers. The POSCO Group argues that the Department’s findings in the second reviews of these orders, in which the Department treated its U.S. sales as EP sales, are consistent with this description of the limited role of the U.S. affiliates. The POSCO Group states that its U.S. sales meet the three-part criteria considered by the Department in classifying transactions as EP sales: (1) the merchandise was shipped directly from the manufacturer to the unaffiliated U.S. customer; (2) this was the customary commercial channel of trade between the parties involved; and (3) the U.S. sales affiliates were mere processors of sales related documentation and communications links with unaffiliated customers in the United States. The POSCO Group argues that since they have met the Department’s three-part test, there is a long line of Department and court precedent classifying these transactions as EP sales and the Department should reclassify POSCO’s and POCOS’ U.S. sales as EP sales. The POSCO Group cites Independent Radionic Workers of America v. United States, 19 CIT 375 (1995); Zenith Electronics Corp v. United States, 18 CIT 870, 873-75 (1994); Industrial Phosphoric Acid from Belgium, 63 FR 25830, 25831 (May 11, 1998); and Circular Welded Non-Alloy Steel Pipe from Korea, 62 FR 55574, 55579 (October 27, 1997). While the POSCO Group acknowledges that the Department found the U.S. sales of its affiliates to be CEP sales in the final results of the third and fourth reviews of these orders, the POSCO Group contends that finding was incorrect. The POSCO Group argues that there is no evidence in these fifth reviews indicating that the U.S. affiliates had any input on the price charged to U.S. customers, which the Department incorrectly asserted did exist in the record of the third reviews. See Korean 3rd Reviews Final, 63 FR at 13183. The POSCO Group also notes that, unlike in the third reviews, in these fifth reviews it provided all of the direct and indirect selling expenses incurred by the U.S. sales affiliates, and the POSCO Group argues that these expenses clearly show that the levels of selling, general and administrative expenses ("SG&A") attributable to sales of subject merchandise through those affiliates are an insignificant portion of total SG&A of those affiliates. The POSCO Group notes that the Department treated as EP transactions sales made by Pohang Steel America Corp. ("POSAM") in Stainless Steel Wire Rod from Korea. In that case, the Department found that "POSAM had no substantial involvement in the sales process, such as sales negotiation, providing technical support, or handling warranty claims, with respect to subject merchandise;" that "POSAM does not negotiate sales terms with U.S. customers, but rather relays pricing information" between the Korean producer and the U.S. customer; that for each sale examined at verification the Korean producer "ultimately accepted or rejected the sales price;" and evidence of indirect contact between the Korean producer and the U.S. customer. See Stainless Steel Wire Rod from Korea, 63 FR at 40419. The Department found that the functions performed by POSAM, "document processing and other ancillary activities related to the sales of subject merchandise to the U.S. customer (e.g., clearing customs, arranging for U.S. transportation, issuing invoices, and collecting payment)," were consistent with a classification of sales as EP sales, and that "POSAM had no substantial involvement in the sales process." Id. The POSCO Group argues that this analysis is consistent with that for the U.S. affiliates in these reviews, given that they did not negotiate prices but merely relayed pricing information, and that they did not provide technical support, handle warranty claims, or accept or reject the price for all sales of subject merchandise. The POSCO Group also notes that the Department treated as EP transactions sales made under similar circumstances in Certain Welded Stainless Steel Pipe from Taiwan, 63 FR 38382 (July 16, 1998) (final results of administrative review) ("Pipe from Taiwan"). In that case, the POSCO Group notes, the U.S. affiliate did not play a major role in the sales negotiation process or the selling activities, was not responsible for setting the prices of U.S. sales, and did not control the markup it earned on the resale of the goods purchased in back-to-back deals for the foreign parent. Id. at 38385. The Department found that the functions performed by the U.S. affiliate in that case, "issuing invoices, collecting payment, paying antidumping duty deposits, and taking title to the subject merchandise after entry into the United States" (id. at 38386) were consistent with the classification of the sales as EP sales, and these functions were the same as those performed by the U.S. affiliates in these fifth reviews. Finally, the POSCO Group cites several cases in which it claims that the U.S. affiliates were more involved in the sales process than the affiliates in these fifth reviews, but for which the U.S. sales were classified as EP sales. Some of the additional functions performed by the affiliates in those cases included limited advertising, processing certain warranty claims, warehousing, and provision of technical services. See e.g., Final Determination of Sales at Less Than Fair Value, Stainless Steel Wire Rod from the Republic of Korea: 63 FR 40404 (July 29, 1998) and Preliminary Determination of Sales at Less Than Fair Value, Certain Cut-to-Length Carbon-Quality Steel Plate Products from the Republic of Korea: 64 FR 41224 (July 29, 1999). Petitioners note that the Department found the POSCO Group’s U.S. sales in the third and fourth reviews to have been CEP sales, and that the POSCO Group had stated affirmatively on the record of the fifth reviews that its U.S. affiliates performed the same functions during the fifth reviews period as they did during the third and fourth reviews periods. Furthermore, the Department had conceded as erroneous its determination in the second reviews that the POSCO Group’s U.S. sales were EP sales rather than CEP sales in the appeal process before the CIT. Petitioners argue that the POSCO Group misstates the Department’s CEP test when it implies that its U.S. sales are EP because its affiliates do not independently negotiate or approve sales to unaffiliated customers. Rather, the Department analyzes whether or not the role played by the U.S. affiliates was "ancillary" in the sales process. As long as the U.S. affiliate plays an active role in bringing about the U.S. sale, that sale will be classified as a CEP transaction even if the foreign parent does play some role in the sales process. Petitioners note that even if POSCO had some contact with a U.S. customer in a particular sale, which they assert is not established by the record, it does not demonstrate that POSAM was not involved in negotiating price terms for the underlying sale, or that POSAM was not involved in the sales process in more than an "ancillary" way more generally for U.S. sales. Petitioners note that the POSCO Group does not even argue that POCOS had direct contact with any of its customers. Petitioners distinguish between the sales processes in these reviews and those in the other cases cited by the POSCO Group. Petitioners note that in Wire Rod from Korea, 63 FR at 40418, the Department indicated that the Korean exporter had direct and substantive contact with U.S. customers (export strategy meetings wherein substantive terms of sale, payment and delivery terms were discussed and from which the exporter established its quarterly price lists). Petitioners note that in Pipe from Taiwan, 63 FR at 38385-6, the Department indicated that the respondent’s customers had frequent direct contact with the producer in Taiwan, and that the producer itself responded directly to customers’ price inquiries. Petitioners state that these aspects of the aforementioned two cases were not present in the reviews at issue here. Petitioners argue that the level of general expenses incurred for subject vs. non-subject merchandise is not relevant for determining the importance of the U.S. affiliates in the sales process for subject merchandise and, given that the POSCO Group reported indirect selling expenses by dividing all such expenses by total sales, the percentage of indirect selling expenses borne by all of the affiliates’ U.S. sales, subject or non-subject, will be the same. Furthermore, the functions performed by the U.S. sales affiliates for non-subject merchandise sales are not relevant in this context, either. Rather, the more appropriate comparison would be of indirect selling expenses incurred by POSCO and POCOS vs. those incurred by the U.S. affiliates for the same sales. Finally, regarding the Department’s classification of PSI’s U.S. sales as EP transactions in the Preliminary Results, Petitioners argue that the Department was incorrect and that it should reclassify PSI’s U.S. sales as CEP transactions, because POSAM was substantially involved in the sales negotiation process. Petitioners argue that the Department "reversed its preliminary findings and rejected POSCO’s claim that its PSI sales were EP transactions, based on what it found at the exhaustive sales verifications it conducted during that review. In the fourth reviews, the Department confirmed this finding that PSI’s sales were CEP transactions." The POSCO Group argues that the Department properly characterized PSI’s U.S. sales as EP transactions in the Preliminary Results. The POSCO Group cites its Supplemental Section C Response to support its claim that, because PSI resumed selling the subject merchandise to the United States during the POR (and did not have any U.S. sales of the subject merchandise during the third and fourth reviews), it therefore needed to create new relationships with U.S. customers, leaving POSAM out of the negotiation process for U.S. sales. In that submission, the POSCO Group stated that "PSI directly contacted all customers for all sales and all entries to all customers during the period of review" and that PSI "negotiated with them directly to determine price, sales terms, and payment terms." See the POSCO Group Supplemental Section C Response, at 7 (April 2, 1999). The POSCO Group argues that, since POSAM did not participate in the sales negotiations and POSAM’s role in PSI’s U.S. sales was limited to relaying the invoice from PSI to the customer (after the sales negotiation had taken place and the purchase order was finalized), the Department properly classified PSI’s U.S. sales as EP transactions. Department's Position: We agree with Petitioners. As noted by Petitioners and the POSCO Group, the essential facts surrounding the activities of Respondents’ U.S. sales affiliates did not change from the third and fourth to the fifth administrative reviews periods. In the third reviews, based in part upon extensive verifications of the U.S. sales affiliates of the POSCO Group, the Department determined that U.S. sales by the POSCO Group were properly classified as CEP, rather than EP, transactions. Specifically, the Department concluded: In these cases, the record clearly establishes that the respondents’ affiliates in the United States were in most instances the parties first contacted by unaffiliated U.S. customers desiring to purchase the subject merchandise and also that the sales affiliates in question signed the sales contracts and engaged in other sales support functions. These facts indicate that the subject merchandise is first sold in the United States by or for the account of the producer or exporter, or by the affiliated seller, and that the sales in question are therefore CEP transactions. Korean 3rd Reviews Final, 63 FR at 13172. In continuing to find that CEP classification is appropriate, U.S. sales affiliates need not be determined to have independently set U.S. prices and other terms of sale for those affiliates’ involvement to be deemed more than ancillary. Rather, the Department must examine the totality of the circumstances surrounding the U.S. sales process, and assess whether the reviewed sales were made "in the United States" for purposes of Section 772(b) of the Act. Accordingly, in this analysis, neither the magnitude of indirect selling expenses incurred, nor the performance of a specific sales function – such as actual negotiation – is controlling. Turning to the evidence in this case, the U.S. affiliates: (1) take title to the subject merchandise; (2) invoice and receive payment from their unaffiliated U.S. customers; (3) arrange for other aspects of the transactions, including Customs clearance, brokerage, and freight; and (4) serve as a source of information about the producer/exporters’ products. Therefore, in keeping with the United States Court of Appeals for the Federal Circuit ("CAFC") recent decision in AK Steel Corp., et. al. v. United States, et.al. No. 99- 1296 ( Fed. Cir. February 23, 2000), the Department has determined that, because the POSCO Group’s U.S. affiliates invoice and receive payment from their unaffiliated U.S. customers, these sales were made in the United States and, thus, are properly classified as CEP transactions. 4. Other AD Issues Comment 13: Level-of-Trade The POSCO Group argues that downstream sales made by the POSCO Group’s affiliated service centers are at a different level-of-trade ("LOT") than POSCO’s home market and U.S. sales. They argue that the Department did not look at the actual facts on the record of these reviews, but rather it relied on the LOT analysis performed in the third reviews. According to the POSCO Group, the service centers provide same-day delivery or "just in time" services to their customers, hold inventory for their customers, sell in small lot sizes, and perform slitting services, while POSCO does not perform any of these services. Furthermore, according to the POSCO Group, the service centers do not provide warranty services, they simply forward warranty claims back to POSCO, which has an intricate warranty claim process. The service centers, according to the POSCO Group, also do not provide technical services and do not have personnel to handle market research, research and development, and quality control, in contrast to POSCO, which has separate personnel to provide these functions. Finally, the POSCO Group argues that "even if some selling activities performed by the service centers and POSCO were the same, this should not preclude a finding that the service centers’ sales were at a different level-of-trade from POSCO’s other sales." See the POSCO Group’s Brief, at 13 (October 8, 1999). Petitioners argue that the Department’s determination that the POSCO Group’s affiliated service centers’ sales are at the same LOT as POSCO’s other sales "is based on the ‘actual facts’ in this review." Petitioners also state that the Department’s determination is consistent with both Department practice and other determinations involving POSCO. They cite Final Determination of Sales at Less Than Fair Value; Stainless Steel Plate in Coils from the Republic of Korea, 64 FR 15444 (March 31, 1999) ("Stainless Steel Plate in Coils"), as a recent case where the Department determined that sales made by POSCO’s affiliated service centers are at the same LOT as POSCO’s home market and US sales. Petitioners argue that Stainless Steel Plate in Coils dealt with POSCO, its affiliated service centers and a similar flat steel product as in the instant case. Petitioners state that the POSCO Group’s claim that all service centers provide same-day delivery and just-in-time services, while POSCO does not, and therefore their sales are at different LOTs, is unpersuasive. According to Petitioners, in none of the cases cited by the POSCO Group was same-day deliveries or just-in-time services at one channel of distribution the sole basis for the Department’s determination that sufficient differences in selling functions existed between channels of distribution to warrant different LOTs. Furthermore, according to Petitioners, the POSCO Group’s argument that its affiliated service centers hold inventory for customers, while POSCO does not, is unsupported by evidence on the records of these reviews. Petitioners state that both POSCO and its affiliated service centers carried inventory of the subject merchandise during the POR, and, therefore, there is no difference in this selling function between the service centers and POSCO. Petitioners also state that with regard to the POSCO Group’s argument that the service centers and POSCO sell to different customer categories, the record evidence also proves the POSCO Group’s assertion is incorrect, as substantial portions of home market sales were made to end-users by both POSCO and the service centers. Petitioners argue that the POSCO Group’s attempt to differentiate between POSCO’s home market sales and that of the service centers based on delivery terms is insufficient, because, according to Petitioners, the same sales terms were available to both POSCO and the service center’s customers. The POSCO Group’s argument concerning warranty claims is also unpersuasive according to Petitioners, because "warranties were paid through both home market channels of distribution." Finally, Petitioners state that the slitting and shearing services that the service centers provide does not constitute a "substantive difference in selling activity under the Department’s LOT analysis." Department's Position: The Department agrees with Petitioners. As stated in the Preliminary Results of these reviews, after examining selling functions along the chain of distribution for both the POSCO Group’s and their affiliated service centers’ sales, we found that "the cumulative functions of the POSCO Group and the service centers for sales made by the service centers are essentially the same as the cumulative functions of the POSCO Group for sales made by the POSCO Group." Preliminary Results, 64 FR at 48773. This determination was based on both our analysis of the POSCO Group’s questionnaire responses and our review of sales functions at the service center and POSCO Group sales verifications. Therefore, we continue to conclude that sales within or between each market are not made at different LOTs, and an adjustment pursuant to section 773(a)(7) is unwarranted. Comment 14: Home Market Credit Petitioners argue that the Department should revise the POSCO Group’s reported credit period for local sales made in the home market because the POSCO Group has the ability to report the actual date of payment for these sales. According to Petitioners, documents obtained during a sales trace conducted at the home market sales verification demonstrate that the POSCO Group was able to link the payment to this particular local sale, but continued to use the average credit days calculation for this particular customer. Therefore, according to Petitioners, the Department should apply facts available for determining the date of payment for home market local sales. The POSCO Group argues that record demonstrates that the POSCO Group is not able to determine the actual date of payment for home market sales, including local sales. According to the POSCO Group, since the POSCO Group companies maintain an open account system, when a home market customer provide notes in order to draw down its account receivable, it is not possible to link these notes to specific transactions. The POSCO Group states that it appropriately calculated customer-specific credit expenses on the average aging of accounts receivable, a methodology which the Department’s questionnaire specifically authorizes. Furthermore, the POSCO Group argues that "{t}he Department’s Verification Report notes that even if in some cases a note might be tied directly to an individual local sale transaction, POSCO could not rely on the date if maturity of the note as the date of payment because, ‘(1) sometimes payment for local sales could be covered by the same notes as payments for domestic sales and/or freight revenue; and (2) many customers purchase both domestic and local sales and also pay the company for freight, and the resulting accounts receivable and notes receivable associated with these customers are combined.’ " Department's Position: The Department agrees with the POSCO Group. The POSCO Group calculated customer-specific credit expenses on the average aging of accounts receivable, a methodology which the Department’s questionnaire specifically authorizes. Furthermore, the POSCO Group explained at the home market sales verification why, even if in some cases a note might be tied directly to an individual local sale transaction, the POSCO Group would not be able to rely on the date of maturity of the note as the date of payment because, "(1) sometimes payment for local sales could be covered by the same notes as payments for domestic sales and/or freight revenue; and (2) many customers purchase both domestic and local sales and also pay the company for freight, and the resulting accounts receivable and notes receivable associated with these customers are combined." See HM Sales Verification Report at 9. The Department has verified this methodology in this and several previous reviews and we continue to find it reasonable and consistent with the Department’s questionnaire instructions. Therefore, we reject Petitioners’ claim that the Department should apply facts available for determining the date of payment and credit expenses for home market local sales. Comment 15: Home Market Downstream Sales Credit Petitioners argue that because the first invoice for home market downstream sales by the POSCO Group’s affiliated service centers is entered into the service center’s accounting system at the time the first invoice is issued and prior to completion of the merchandise, the calculation of customer credit days for downstream sales is distortive. Petitioners state that, as facts available, the Department should deny the POSCO Group’s credit expense adjustment for all downstream sales. The POSCO Group states that, in a few cases, the affiliated service centers invoiced the customer separately for the coil and the further processing. After invoicing the customer for the coil, the service centers also transferred the title to the coil to the customer and, according to the POSCO Group, the service center would then invoice the customer for the processing after it was performed. The POSCO Group argues that because credit days are based on the relationship of the average accounts receivable to sales, there is no impact on the credit calculation since, for these sales, invoices are generated in two stages and payment is likely received in two stages as well. Department's Position: The Department agrees with the POSCO Group. Since, in some instances, the service centers invoice their customer for the purchase of the coil itself, as well as transfer title to the coil to the customer, it is reasonable for them to also enter this invoice into their accounting system, in order to track outstanding invoices and accounts receivable. The Department’s questionnaire specifically allows calculation of credit days based on the average age of accounts receivable for each customer. Therefore, the Department finds Petitioners’ claim that the Department should reject the POSCO Group’s credit expense adjustment for all downstream sales unwarranted. Comment 16: Product Characteristics for Downstream Sales Petitioners argue that POSCO has misreported product characteristics for a considerable portion of their downstream sales. Petitioners contend that the verification report stated that POSCO misreported thickness (CTHICKH) for the majority of the downstream sales examined at verification. Additionally, Petitioners note that the verification report stated that POSCO attempted to correct these errors with a detailed methodological explanation at verification, but POSCO’s presentation appeared to contain new information, and the Department chose not to accept it. Further, Petitioners state that due to the errors in the coding of the thickness product characteristic, the Department is not able to rely on POSCO’s reported information for model matching. Thus, Petitioners contend that the Department should apply facts available to all U.S. sales that could possibly match to this particular company’s downstream sales, the highest rates from the petition, 54.61 percent for cold- rolled steel and 64.50 percent for corrosion-resistant steel. Respondent argues that the Department should reject Petitioners’ request to use facts available for a certain company’s downstream sales. Respondent explained at verification why a coding error in the product thickness reported by the company occurred because the Department flipped the order of the width and thickness product characteristics in the questionnaire. Further, in its Preliminary Results the Department had made an adjustment to the product characteristic to account for this error, therefore, no adjustment is necessary. Respondent mentions that in the first proceeding POSCO was required to report downstream sales made by its affiliated service centers and that it cooperated with the Department’s request and reported the sales. POSCO contends that at verification the Department verified its reporting methodology used to report product characteristics for the service center sales and the Department found no discrepancies. At this particular company’s verification, Respondent asserts that it explained the coding error in the reporting of the thickness characteristic, that this error resulted from the fact that when it tied the sales by the service center to the incoming coil from POSCO, it had relied in part upon data from the fourth administrative reviews period for sales by POSCO to the service centers. Also, POSCO stated that the Department’s fourth reviews questionnaire had flipped the order of the width and thickness characteristics from previous review periods. The error was simply a clerical error. Further, POSCO asserts that this error did not affect all sales, and that the Department verified observations for product thickness for other sales and those sales product thickness verified, thus, the universe of sales is limited by the nature of the issue. Furthermore, POSCO stated that Petitioners did not challenge the Department’s adjustment to the product thickness characteristics from the Preliminary Results. Therefore, Respondent argues that the Department’s adjustment and using the nominal thickness from the downstream sales database is appropriate and reasonable, and Petitioners’ claim of applying facts available to any U.S. sales that matches to a sale made by this particular company should be rejected. Department’s Position: We disagree with Petitioners. At verification, POSCO provided a reasonable explanation why a coding error in the product thickness reported by POSCO had occurred. At verification POSCO stated that in "the fourth review period, the ordering of the corrosion resistant product characteristics in the Department’s questionnaire differed from that in the fifth reviews questionnaire (width and thickness variables were reversed), and that the company had ordered its characteristics consistent with the order of the questionnaires, so a mismatch of characteristics would occur for affiliated service center sales consisting of POSCO substrate that was purchased from POSCO prior to October 1997." See Verification Report of the POSCO Group with respect to Service Centers, at 9 (August 2, 1999). Additionally, at the preliminary results stage of these administrative reviews, we made an adjustment to the thickness product characteristic in our model match program to account for POSCO’s coding error of the corrosion-resistant product. Accordingly, we have not made any changes to our SAS programming. Comment 17: Overrun Sales Petitioners argue that POSCO’s home market overrun sales do not appear to be at prices that would allow them to be designated overruns. Thus, Petitioners contend that the Department should apply facts available to all home market overrun sales. Respondent argues that Petitioners’ claims concerning POSCO’s overrun sales should be rejected. First, POSCO contends that the Department verified its reporting of overrun and non-overrun sales and found no discrepancies. Second, POSCO states that the Department asked in a supplemental questionnaire why in certain instances overrun sales were not at prices that would allow them to be designated overruns. POSCO stated that the difference in prices "is mainly due to differences in time between the dates of the overrun sales and the dates of the non-overrun sales." See Supplemental Questionnaire (April 2, 1999). Additionally, POSCO stated that in responding to this question, it referenced specific CONNUMs. Further, POSCO noted that the Department issued a second supplemental questionnaire, which asked were there any reasons why the prices would have varied other than differences in time. In responding to this question, POSCO stated that the CONNUMs identified by the Department had price differences due to the time between the dates of the overrun sales and the dates of the non-overrun sales. Furthermore, Respondent asserts that Petitioners have provided no basis or rationale for the Department to make any adjustment to its overrun sales, and have not identified any discrepancy on the record. Therefore, Respondent argues that no adjustments should be made to its reporting of overrun sales. Department’s Position: We disagree with Petitioners. At verification, we examined POSCO’s methodology for reporting its non-prime merchandise and found their reporting to be consistent with their questionnaire responses. We specifically examined at verification the overrun CONNUMs that petitioners had requested the Department to review at verification. Through our home market sales traces, we found that the overrun sales were consistently reported and we found no discrepancies in any of the overruns sales. See HM Sales Verification Report at 35-36, observations HM 16 and HM 17. Therefore, we have determined that there is no basis to apply facts available to POSCO’s overrun sales. UNION 1. Other AD Issues Comment 18: Home Market Credit Days Petitioners argue that Union’s home market credit adjustment should be denied because the system is distortive. Petitioners state that Union claims that customer payments cannot be traced on a transaction- specific basis because customers "generally pay on an ‘open’ (i.e. revolving account) basis;" therefore, because of this system, Union calculated its credit period by taking the ratio of average daily receivables to average daily sales. Petitioners further argue that, according to Union, under its "open account payment system, payments are made with promissory notes on a rolling basis and thus are not directly tied to any specific sales"; its credit period calculation does not segregate payment, sale or receivable information between "flat rolled products" for the purpose of deriving its credit period. Petitioners contend that Union has calculated a "customer-specific credit period," as opposed to a customer-specific and product- specific credit period; accordingly, Union’s claimed credit adjustment includes both in-scope and out-of-scope merchandise. Petitioners argue that this type of price adjustment calculation methodology has recently been explicitly rejected by the U.S. Court of Appeals for the Federal Circuit ("CAFC"). See SKF USA Inc. v. United States, et al., 180 F.3d 1370, 1376 (Fed. Cir. 1999) ("SKF"). Specifically, the CAFC has recently held that price adjustments must relate exclusively to merchandise within the scope of the review. Petitioners state that based on the CAFC’s ruling, Union is not entitled to a downward adjustment to the price of home market sales based on its credit calculation, because Union’s claimed home market credit adjustment is not based solely on in-scope merchandise. Union states that the Department’s practice has been consistent in dismissing Petitioners’ claims, i.e. that Union’s customer-specific credit period methodology was distortive, and in using Union’s reported customer-specific credit periods. In the Korean 4th Reviews Final, the Department addressed Respondents’ use of "open payment" systems in their home market and the purported distortions allegedly created by calculating customer-specific credit periods for their home market sales. Noting "the preference of the Department to calculate imputed credit on as specific a basis as possible," the Department stated that it did not appear that "any systematic credit- reporting distortion exists for respondents in these reviews" and pointed to the fact that "the Department has accepted the inherent reasonableness of respondents’ methodology and the accuracy . . . of the data used as the basis for the calculation of credit days. See Korean 4th Reviews Final, 64 FR at 12930. Union argues that Petitioners used a false assertion (claiming Union has included non-scope merchandise) to cite the recent decision from the CAFC regarding direct price adjustment, their relationship to in-scope merchandise, and the accuracy of the Department’s dumping calculations. Union states that its credit periods are based solely on in-scope merchandise. As stated in its supplemental response, "Union was able to exclude pipe and tube sales from flat-rolled sales for purposes of providing data on customer -specific credit periods because Union has different customer codes for flat-rolled customer and pipe and tube customers and, therefore, payments are made separately for each product group." See Union’s Supplemental Response (June 14, 1999) There can be no distortion with credit periods calculated for customers that purchase both cold-rolled and corrosion-resistant products. For these customers, the reported sales and outstanding receivables which are the basis for calculating each customer’s credit period include all sales and outstanding receivables of both cold-rolled and corrosion-resistant products. Union therefore argues that its credit periods are based solely on in-scope merchandise. Although Petitioners suggest that the standard should be customer-specific and product-specific credits periods, Union argues that the Department’s practice is "to calculate imputed credit on as specific a basis as possible." Union argues that there can be no distortion with credit periods calculated for customers that purchase both cold-rolled and corrosion-resistant product. For these customers, the reported sales and outstanding receivables which are the basis for calculating each customer’s credit period include all sales and outstanding receivables of both cold-rolled and corrosion-resistant product. Department’s Position: The Department does not accept Petitioners’ assertion that Respondent’s claimed credit adjustments should be rejected. According to the Department’s Regulations at 19 CFR 351.401(g), the Department may consider allocated expenses if it is shown by the submitting party that the allocation is calculated on as specific a basis as is feasible and it is explained why the allocation methodology used does not cause inaccuracies or distortions. See also Preamble to 19 CFR 351.401. In the June 14, 1999 Supplemental Questionnaire, the Department requested the Respondent to indicate the greatest level of detail, by product, for which sales and receivables data can be determined for purposes of the calculation of credit periods. The Respondent stated that the greatest product- specific level of detail that it maintained was to distinguish pipe and tube products from the in-scope flat-rolled products. The Respondent indicated that this degree of allocation could be achieved because it had different customer codes for flat-rolled customers and pipe and tube customers; therefore, the payments are made separately for each product group. It would not be feasible for the Respondent to further break down flat-rolled products into corrosion-resistant and cold-rolled sales due to the method of payment, which is by way of promissory notes on a rolling basis and not directly tied to any specific sales. The Department further determines that, as argued by the Respondent, this level of allocation does not result in unreasonable distortions. The Petitioner has presented no evidence of any fundamental difference between the credit granted on corrosion- resistant and cold-rolled sales. It is also noted that the Department has, in prior reviews of these cases, accepted respondent’s methodology for calculating customer specific credit periods using the open payment system. See ("Korean 4th Reviews Final"). Furthermore, the Department does not find, as claimed by Petitioners, that the ruling by the CAFC in SKF is the proper guide for the facts of this review. The fact pattern of these reviews is distinct from that in SKF, where the court focused on in-scope versus out-of-scope products. In contrast, in this case, Petitioners are seeking to differentiate between two in-scope products. Therefore, in accordance with our Regulations, we have continued to utilize Respondent’s credit period adjustment methodology. Comment 19: Prime v. Non-Prime Union argues that it is the Department’s policy to compare prime merchandise with prime merchandise, and non-prime merchandise with non-prime merchandise. Moreover, in the Analysis Memo that accompanied the Preliminary Results, the Department stated that Union only had sales of prime merchandise to the United States. Thus, the Department ostensibly matched comparison market models to U.S. models with the same characteristics, i.e., it matched U.S. sales of prime merchandise only with home market sales of prime merchandise. However, the Department’s SAS computer program for the preliminary results shows no evidence that this policy was implemented. In order to remedy this omission, Union argues that the Department’s SAS program should drop Union’s home market sales of non-prime merchandise. Therefore, home market sales of non-prime merchandise will be excluded from the home market sales database before any model matches are performed and any possibility of matching prime with non- prime merchandise will be eliminated. Department’s Position: We agree with Respondents that we did not prevent possible matches of prime to non-prime. Therefore, for the final results, we have adjusted our model matching program to reflect the matching of U.S. sales of prime merchandise only with home market sales of prime merchandise. See Final Analysis Memorandum dated March 6, 2000. 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 review and the final weighted- average dumping margins for all reviewed firms in the Federal Register. AGREE____DISAGREE____ __________________________________________ Robert S. LaRussa Assistant Secretary for Import Administration __________________________________________ Date