Antidumping: Fresh Atlantic salmon from— Chile,

[Federal Register: June 9, 1998 (Volume 63, Number 110)]

[Notices]

[Page 31411-31437]

From the Federal Register Online via GPO Access [wais.access.gpo.gov]

[DOCID:fr09jn98-38]

DEPARTMENT OF COMMERCE

International Trade Administration

[A-337-803]

Notice of Final Determination of Sales at Less Than Fair Value: Fresh Atlantic Salmon From Chile

AGENCY: Import Administration, International Trade Administration, Department of Commerce.

EFFECTIVE DATE: June 9, 1998.

[[Page 31412]]

FOR FURTHER INFORMATION CONTACT: Gabriel Adler or Kris Campbell, Office of AD/CVD Enforcement 2, Import Administration, International Trade Administration, U.S. Department of Commerce, 14th Street and Constitution Avenue, N.W., Washington, D.C. 20230; telephone: (202) 482-1442 or (202) 482-3813, respectively.

The Applicable Statute and Regulations

Unless otherwise indicated, all citations to the statute are references to the provisions effective January 1, 1995, the effective date of the amendments made to the Tariff Act of 1930 (the Act) by the Uruguay Round Agreements Act (URAA). In addition, unless otherwise indicated, all citations to Department of Commerce (the Department) regulations refer to the regulations last codified at 19 CFR part 353 (April 1, 1997).

Final Determination

We determine that fresh Atlantic salmon from Chile is being sold, or is likely to be sold, in the United States at less than fair value (LTFV), as provided in section 735 of the Act. The estimated margins are shown in the Continuation of Suspension of Liquidation section of this notice.

Case History

The preliminary determination in this investigation was issued on January 8, 1998. See Notice of Preliminary Determination of Sales at Less Than Fair Value and Postponement of Final Determination: Fresh Atlantic Salmon from Chile, 63 FR 2664 (January 16, 1998) (Preliminary Determination). Since the preliminary determination, the following events have occurred.

In February and March 1998, we conducted on-site verifications of the questionnaire responses submitted by Aguas Claras S.A. (Aguas Claras), Cia. Pesquera Camanchaca S.A. (Camanchaca), Pesquera Eicosal Ltda. (Eicosal), Pesquera Mares Australes Ltda. (Mares Australes), and Marine Harvest Chile (Marine Harvest)(collectively, ``the respondents'').

On April 17, 1998, we received case briefs from the Coalition for Fair Atlantic Salmon Trade (the petitioners) and, on behalf of the respondents, the Association of Chilean Salmon and Trout Producers (the Association). On April 23, 1998, we received rebuttal briefs from the same parties. We held a public hearing on April 28, 1998.

Scope of Investigation

The scope of this investigation covers fresh, farmed Atlantic salmon, whether imported ``dressed'' or cut. Atlantic salmon is the species Salmo salar, in the genus Salmo of the family salmoninae. ``Dressed'' Atlantic salmon refers to salmon that has been bled, gutted, and cleaned. Dressed Atlantic salmon may be imported with the head on or off; with the tail on or off; and with the gills in or out. All cuts of fresh Atlantic salmon are included in the scope of the investigation. Examples of cuts include, but are not limited to: crosswise cuts (steaks), lengthwise cuts (fillets), lengthwise cuts attached by skin (butterfly cuts), combinations of crosswise and lengthwise cuts (combination packages), and Atlantic salmon that is minced, shredded, or ground. Cuts may be subjected to various degrees of trimming, and imported with the skin on or off and with the ``pin bones'' in or out.

Excluded from the scope are (1) fresh Atlantic salmon that is ``not farmed'' (i.e., wild Atlantic salmon); (2) live Atlantic salmon; and (3) Atlantic salmon that has been subject to further processing, such as frozen, canned, dried, and smoked Atlantic salmon, or processed into forms such as sausages, hot dogs, and burgers.

The merchandise subject to this investigation is classifiable as item numbers 0302.12.0003 and 0304.10.4093 of the Harmonized Tariff Schedule of the United States (HTSUS). Although the HTSUS statistical reporting numbers are provided for convenience and customs purposes, the written description of the merchandise is dispositive.

Period of Investigation

For all companies, the period of investigation (POI) corresponds to each respondent's four most recent fiscal quarters prior to the month of the filing of the petition (June 1996). For four of the five respondents, the POI is April 1, 1996, through March 31, 1997. The remaining respondent, Marine Harvest, has a different fiscal period. The POI for this company is March 24, 1996, through March 22, 1997.

Fair Value Comparisons

To determine whether sales of fresh Atlantic salmon from Chile to the United States were made at less than fair value, we compared the export price (EP) or constructed export price (CEP), as appropriate, to the normal value. Our calculations followed the methodologies described in the preliminary determination, except as noted below and in company- specific analysis memoranda dated June 1, 1998, which have been placed in the file.

Export Price and Constructed Export Price

For the price to the United States, we used EP or CEP as defined in section 772 of the Act. We calculated EP and CEP based on the same methodology used in the preliminary determination, with the following exceptions:

Mares Australes

We excluded sales to Canada from the U.S. sales database. See Comment 17.

Marine Harvest

We made an adjustment for accrued rebate expenses to the CEP calculated for one customer. See Comment 19.

Normal Value

We used the same methodology to calculate normal value as that described in the preliminary determination, with the following exceptions. For Eicosal, Mares Australes, and Marine Harvest, we determined that the differences between premium and super-premium salmon are so minor as to not warrant separate classification in an antidumping analysis, and considered all such sales to be of premium salmon. See Comment 1. With respect to specific respondents' data, we made the following changes:

Aguas Claras

We did not rely on Canadian sales of salmon fillets to calculate normal value for comparison to U.S. sales of fillets. Instead, we compared U.S. sales of fillets to constructed value (CV). See Comment 7.

Mares Australes

We made an adjustment to normal value for duty drawback.

Cost of Production

In accordance with section 773(b)(3) of the Act, we calculated the weighted-average cost of production (COP), by model, based on the sum of each respondent's cost of materials, fabrication, general expenses, and packing costs. We relied on the submitted COPs except in the following specific instances where the submitted costs were not appropriately quantified or valued.

Marine Harvest

1. We increased the reported cost of eggs and feed purchased from affiliated parties to reflect market prices. See Comment 22.

2. We increased the reported cost of processing performed by an affiliated party to reflect the transfer price. See Comment 22.

   [[Page 31413]]

3. We revised the consolidated financial expense ratio to include exchange losses associated with loans denominated in foreign currencies. See Comment 24.

4. We recalculated the general and administrative expense (G&A) ratio to correct certain errors discovered during verification.

   Mares Australes

1. We increased the cost of manufacturing (COM) to include the price-level adjustments for harvested salmon which were required by Chilean GAAP. See Comment 27.

2. We increased the COM to include bonus expenses. See Comment 31.

3. We revised the consolidated financial expense ratio to remove the claimed offset to financial expense for accounts receivable and inventory. See Comment 24.

4. We recalculated the G&A expense ratio based on total G&A expenses incurred by the producing entities. See Comment 30.

   Aguas Claras

1. We increased the COM to include the price-level adjustments for harvested salmon which were required by Chilean GAAP and were recorded in the company's normal books and records. See Comment 27.

2. We revised the claimed ``feed cost adjustment'' by amortizing the total amount specified in the contract over the life of the contract. We then allocated the amortized adjustment to individual fish groups based on each group's relative biomass. See Comment 36.

3. We excluded from G&A expenses the gains from the sales of common stock investments. Additionally, we included the cost incurred by Sociedad Agricola Rio Rollizo Ltda. (``Rio Rollizo'') which held the marine concession for the Rio Rollizo hatchery. See Comment 38.

4. We revised the financial expense ratio to include exchange losses associated with loans denominated in foreign currencies. Additionally, we removed the claimed offset to financial expenses for accounts receivable and inventory. See Comment 24.

5. We revised the manner in which we calculated indirect selling expenses for CV so as to add an amount proportionate to the cost of each product, rather than a fixed amount. See Comment 40.

   Camanchaca

1. We increased the COM to include the price-level adjustments for harvested salmon that were required by Chilean GAAP and were recorded in the company's normal books and records. See Comment 27.

2. We revised the consolidated financial expense ratio to include exchange losses. Additionally, we removed the claimed offset to financial expenses for accounts receivable and inventory. See Comment 24.

3. We revised the G&A expenses to include the non-operating gains and losses that related to the general operations of the company. Also, we calculated the G&A expense ratio based on total G&A expenses incurred by the company. See Comment 33.

   Eicosal

1. We increased the COM to include the price-level adjustments for harvested salmon which were required by Chilean GAAP and were recorded in the company's normal books and records. See Comment 27.

2. We revised the consolidated financial expense ratio to include exchange losses. Additionally, we removed the claimed offset to financial expenses for holding accounts receivable and inventory. See Comment 24.

3. We revised the G&A expenses to include the non-operating gains and losses that related to the general operations of the company. Also, we calculated the G&A expense ratio based on total G&A expenses incurred by the salmon producing company. See Comment 29.

   Currency Conversions

   As in the preliminary determination, we made currency conversions in accordance with section 773A of the Act. The Department's preferred source for daily exchange rates is the Federal Reserve Bank. The Federal Reserve Bank publishes daily exchange rates for Japanese yen, but not for Chilean pesos. In cases involving comparisons to third- country market sales in Japan, which were necessary for three respondents, we made conversions of values denominated in Japanese yen based on the official exchange rates published by the Federal Reserve. For conversions of values involving Chilean pesos, we relied instead on daily exchange rates published by Dow Jones News/Retrieval on-line system. The parties did not comment on these exchange rate methodologies.

   Verification

   As provided in section 782(i)(1) of the Act, we verified the information submitted by the respondents for use in our final determination. We used standard verification procedures, including examination of relevant accounting and production records, as well as original source documents provided by the respondents. We also met with officials of the Association to discuss its grading standards.

   Interested Party Comments

   Sales Issues--General

   Comment 1: Distinction between ``Premium'' and ``Super-Premium'' Grades.

   The petitioners argue that the Department erred in the preliminary determination by accepting as a bona fide grade distinction the ``super-premium'' designation adopted by the Association with respect to whole salmon sold to Japan. The petitioners contend that most of the Chilean salmon exported to both the United States and Japan was graded as premium until shortly before the POI. According to the petitioners, the Association's adoption of the super-premium grade in 1996 coincided with active preparations for an impending antidumping petition against salmon from Chile, and was designed to avoid comparisons of low-priced sales of premium-grade salmon to the United States to high-priced sales of the same merchandise to Japan.

   The petitioners add that verification revealed that the respondents' classification of premium versus super-premium salmon is based only on very minor differences in the external aspects of the salmon. According to the petitioners, these differences are insignificant, and do not meet the Association's stated criteria for differentiation among premium and super-premium salmon. Further, the petitioners argue that the finding at verification that the super- premium/ premium distinction rests primarily on such minor differences in grading is at odds with the respondents' earlier representations that the color of the salmon meat is the principal distinguishing factor between premium and super-premium salmon. The petitioners contend that verification established that: (1) the respondents' premium and super-premium salmon are of uniformly high color, and (2) the respondents do not evaluate the color of salmon during the grading process.

   As further evidence that the respondents' grading practices are at odds with the Association's standards, the petitioners note that the records maintained by Marine Harvest (one of the three respondents that export the foreign like product to Japan) do not distinguish even nominally between premium and super-premium salmon. According to the petitioners, Marine

   [[Page 31414]]

   Harvest's invoices, ledgers, and other documentation refer to top-grade Chilean salmon invariably as ``superior,'' regardless of whether the salmon is exported to the United States or to Japan. Moreover, the petitioners argue, the same designations are used by Marine Harvest's Scottish affiliate for sales of Scottish salmon to the United States and Japan, noting that the Scottish standard for superior grade is equivalent to the U.S. standard for premium grade.

   The Association responds that the Department confirmed at verification that super-premium and premium salmon are distinct products with different physical characteristics and market values. According to the Association, its super-premium grading criteria were established before the beginning of the POI in order to formalize a long-standing requirement by Japanese customers for salmon with no imperfections. The Association contends that, at verification, the Department observed that the grading criteria were strictly applied and enforced by independent, internationally-recognized quality assurance agencies, and it maintains that the Department confirmed the application of these criteria during the POI.

   The Association further asserts that the discernible differences between premium and super-premium salmon are evidenced by the differences in prices obtained for the two grades in the Japanese market. In this respect, the Association notes that Mares Australes, the only respondent to sell both super-premium and premium grade salmon to Japan, reported higher prices for sales of super-premium grade salmon.

   With respect to Marine Harvest's recording of the grade of merchandise sold to Japan, the Association claims that, although the Marine Harvest processing plant follows its own separate grading standards for the U.S. and Japanese markets, these standards are consistent with the Association's standards. Thus, even though Marine Harvest's salmon are nominally referred to as being of ``superior'' grade on invoices to both markets, there are discernible physical differences between the merchandise shipped to those markets. Further, the Association argues, the Marine Harvest plant also relies on independent quality certification agencies to rate its compliance with Association grading standards, and the plant received perfect scores in those evaluations in reports corresponding to the POI that were examined at verification.

   DOC Position: In the preliminary determination, we tentatively accepted the Association's distinction between premium and super- premium salmon, pending verification and further analysis of this issue. After conducting verification and carefully considering the evidence on the record, we have concluded that any differences between premium and super-premium salmon are so minor as to not warrant separate classification in an antidumping analysis.

   At the outset, we note that we are not persuaded by the petitioners' assertion that the Association's adoption of the super- premium grade in 1996 was designed primarily to avoid comparisons, in the event of an antidumping case, of low-priced sales of premium-grade salmon to the United States to high-priced sales of the same merchandise to Japan. We acknowledge that the Association's grading standards and those of some of the individual respondents did include distinct ``premium'' and ``super-premium'' classifications. During verification, we found that quality control inspections at the respondents' plants were supervised by independent certification agencies, which certified the respondents' compliance with the Association's grading standards, and that these standards specified distinct ``premium'' and ``super-premium'' grades. The reports issued by the independent certification agencies during the POI indicated high scores in the category of adherence to these grading standards. See Memorandum from Case Analysts to Gary Taverman, Regarding Inspection of Eicomar Processing Plant (April 7, 1998) (Eicomar Verification Report) at 3-4 and Exhibit P-2; see also Memorandum from Case Analysts to Gary Taverman, Regarding Verification of Sales by Marine Harvest (April 7, 1998) (Marine Harvest Sales Verification Report), at 8-9 and Exhibit M- 25.

   However, the record also contains evidence that the distinctions between the two grades were, in practice, nominal. At the outset of this proceeding, the Association explained that the single most important factor considered by Japanese customers in purchasing fresh Atlantic salmon is the color of the meat. See letter from the Association to the Department of Commerce (November 3, 1997) (alleging particular market situation in Japan) at 14. Both the Association standards and the respondents' individual standards require higher meat color for super-premium salmon than for premium salmon. See letter from the Association to the Department of Commerce (October 10, 1998) at Attachment 1 (transmitting Association standards); see also letter from Mares Australes to the Department of Commerce (November 3, 1997) (Mares Australes Section A and B Questionnaire Response) at 19-20; and letter from Eicosal to the Department of Commerce (November 3, 1997) (Eicosal Section A and B Questionnaire Response), at 4. Despite these claims regarding the significance of color in distinguishing the two grades, we found at verification that, in practice, the respondents adjust the feed delivered to the salmon pens so as to ensure a uniformly high red color to the salmon meat for all salmon produced. See, e.g., Eicomar Verification Report at 2. Further, verification established that the respondents do not measure the color of the whole salmon during processing, but rather take an occasional sample to ensure that the fish are of sufficiently high color. Id. at 3.\1\ Thus, respondents routinely export to the United States salmon that has the same meat color as the salmon exported to Japan and do not consider the criterion (color) that was initially claimed to be of paramount significance in distinguishing super-premium from premium salmon.

   \1\ Although the Association claims that a shiny blue exterior on a whole salmon is indicative of very red meat color, at verification we found that in practice this was not used as a yardstick to differentiate premium from super-premium salmon: ``According to plant officials, salmon exhibiting a shiny blue exterior will have meat surpassing the Association's standards for color required for premium and super-premium grades.'' Id. at 2.

   The Association argues that, in addition to color, its standards also distinguish among minor external imperfections in the salmon. During the plant tour conducted at verification, Department verifiers observed that there were in fact minor differences between salmon classified as premium and salmon classified as super-premium, such as small scale loss or light lacerations. These minor differences, however, do not establish a different grade of salmon for purposes of our analysis. While the Chilean respondents that sell to both the United States and Japan may sort their harvest based on the premise that Japanese customers are more likely to take notice of a light defect than U.S. customers, such differences are not recognized by the salmon producers of any other nation that exports to Japan. The Norwegian, Scottish, Canadian, and U.S. farmed salmon industries do not recognize any grade higher than ``superior.'' The ``superior'' grade is consistent with the premium grade and permits minor defects.\2\ Because the grading standards

   [[Page 31415]]

   of ``superior'' salmon recognized by the world's largest salmon farming countries provide for a range of quality (e.g., from zero defects to up to three minor defects) we note that, by definition, there will be some merchandise within this grade with no imperfections, as well as some merchandise that will be closer to the lower end of this range. Nonetheless, all salmon in this range are graded equally (i.e., as ``superior''/``premium''), and are comparable products in the market place.\3\

   \2\ We note that one of the respondents in this investigation, Marine Harvest, has an affiliate in Scotland that produces and exports fresh Atlantic salmon to Japan. At verification, we reviewed the grading standards followed by Scottish producers, and found that the highest-quality salmon produced by those producers is graded as ``superior.'' The ``superior'' standard allows for light defects, and is comparable to the Chilean ``premium'' standard. See Marine Harvest Sales Verification Report at 13 and Exhibit M-24. Further, we found that invoices for Marine Harvest's sales of Chilean salmon and invoices for the Scottish affiliate's sales of Scottish salmon refer to salmon sold in Japan as ``superior'' salmon, and do not distinguish the two in any manner.

   \3\ While the Association's ``super-premium'' specification for fresh Atlantic salmon does not tolerate any defects in the fish, the Association has no such standard for other types of salmon, such as coho salmon. Thus, by the Association's own standards, a range of small defects is generally permissible for a variety of different types of fish sold in Japan. The respondents have not demonstrated that fresh Atlantic salmon is so unique to Japanese customers in comparison with other salmon that a heightened quality standard is required for this particular type of salmon.

   Finally, regarding the Association's claim that there are price differences in Japan for salmon sold as ``super-premium'' versus that sold as ``premium,'' we note first that, as shown above and in accordance with our practice, our matching criteria are based on the actual physical characteristics of the merchandise. Moreover, even if we were to consider the Association's analysis, it rests entirely on sales made by the one company that made POI sales of both designations to Japan. The pricing of this company's sales of merchandise labeled ``premium,'' which covered only a few months of the POI and involved relatively small quantities, is an insufficient basis on which to find systematic price differences between the two labels, much less to employ a matching methodology based on such differences.

   The nominal distinctions noted above do not preclude an apples-to- apples comparison of the salmon sold in the two markets. For this final determination, we have considered that salmon reported as super-premium are in fact of premium grade and have matched such sales to premium- grade salmon sold in the United States, where otherwise appropriate.

   Comment 2: Distinction between Vacuum-Packed Fillets and Regular Fillets.

   The petitioners argue that the Department erred in preliminarily accepting the respondents' treatment of vacuum-packed fillets and regular fillets as separate forms of merchandise, thereby precluding comparisons of identical merchandise. The petitioners argue that vacuum-packed salmon fillets sold in Japan are identical to regular fillets sold in the United States in every respect except packing, and claim that their prices can be compared after the appropriate adjustment for differences in packing costs.

   The petitioners further contend that, in responding to the Department's cost of production questionnaire, Marine Harvest and Eicosal erroneously included vacuum-packing costs in the reported cost of manufacturing of fillets that were vacuum-packed. According to the petitioners, vacuum-packing costs should be regarded as costs of packing for shipment (i.e., the cost of containers incidental to placing the foreign like product in a ready condition for shipment), consistent with section 773(b)(3)(C) of the Act.

   In addition, the petitioners argue that the Department incorrectly relied on Washington Red Raspberry Commission v. United States, 859 F.2d 898, 905 (Fed. Cir. 1988)(Red Raspberry Commission) in distinguishing vacuum-packed fillets in the preliminary determination. According to the petitioners, the CAFC ruled in that case that packing can only be considered an integral part of a product if the product could not survive in its natural form without such packing. According to the petitioners, vacuum packing is not necessary to bring salmon fillets to market, as they are regularly wrapped in sheets of plastic, without vacuum packaging. Petitioners argue that, at most, vacuum packing lengthens the shelf-life of a fillet, an advantage that is obviated if the product is quickly consumed.

   Finally, petitioners argue that Department practice supports the treatment of vacuum packing as packing costs, rather than as physical differences, citing, inter alia, Tapered Roller Bearings and Parts Thereof, Finished and Unfinished, From Japan and Tapered Roller Bearings, Four Inches or less in Diameter, and Components Thereof, From Japan; Final Results of Antidumping Duty Administrative Reviews and Revocation in Part of an Antidumping Finding, 61 FR 57629, 57630 (November 7, 1996)(TRBs from Japan). Petitioners claim that TRBs from Japan stands for the proposition that not comparing identical products that differ only by their packaging would constitute ``an additional matching factor which is unwarranted by the statute.'' Id.

   The Association responds that the Department correctly determined that vacuum-packed fillets sold in Japan are physically different from fillets sold in the United States and thus cannot be used for comparison. The Association contends that vacuum packing represents a significant additional processing step, akin to smoking or canning, that enhances the shelf life of the product, rather than merely placing the product in a condition ready for shipment. According to the Association, the proper reading of the CAFC's decision in Red Raspberry Commission is that packaging is an integral part of the product when it is in effect a part of that product. The Association argues that the Department has consistently followed this rule in other cases, and maintains that the cases cited by petitioners are inapposite.

   DOC Position: We agree with the Association. Vacuum packing is not incidental to shipment, but is instead an extra processing step that doubles the shelf life of fresh Atlantic salmon. Such packing is an integral part of the product, and its cost is appropriately included among costs of manufacturing, rather than among costs of packing for shipment.

   At the outset of this investigation, after considering the parties' comments with respect to vacuum packing, we recognized the distinction between regular fillets and vacuum-packed fillets, and instructed the respondents to treat these as separate forms. See Antidumping Questionnaire at B-6 and C-6 (August 26, 1997). The respondents appropriately included the cost of vacuum packing in the costs of manufacturing, and included the cost of Styrofoam boxes and cooling materials as packing materials.

   The cases cited by the petitioners do not require a different result. In those cases, the issue was whether products sold individually could be compared to groupings of products, or to bulk sales. See, e.g., Final Determination of Sales at Less Than Fair Value: Fresh Cut Roses from Ecuador, 60 FR 7019, 7022 (February 6, 1995)(Roses from Ecuador)(noting that roses are not transformed by virtue of being bunched or placed in a bouquet); see also TRBs from Japan, 61 FR 57629, 57630 (November 7, 1996)(noting that bearing cups or cones sold individually could be compared to package sets); and Gray Portland Cement and Clinker from Mexico: Final Results of Antidumping Duty Administrative Review, 63 FR

   [[Page 31416]]

   12764, 12777 (March 16, 1998)(Cement from Mexico)(noting that bagged cement and bulk cement are identical except in packaging, and could be compared). In the instant case, the issue is not whether fillets sold individually should be compared to fillets sold by the box, or to fillets sold in bulk quantities. Rather, it is whether the product is transformed by vacuum packing, such that the packing becomes an integral part of the product.

   In Red Raspberry Commission, the CAFC found that packing of raspberries is an integral part of the product, stating that the cardboard containers are necessary for the very survival of the merchandise. The CAFC held that, because the packing was an integral part of the product, it was properly included in the cost of manufacturing rather than treated as packing for shipment. However, the ruling does not suggest that packing that otherwise transforms the physical properties of a product cannot also be considered an integral part of the product. In significantly extending the shelf life of a fillet, the vacuum packing transforms the product. We also note that the vacuum-packing process extends the shelf life not only by the packaging itself but also by other aspects of the vacuum-packing process, such as the use of ethyl alcohol, which significantly lowers the bacteria count of the salmon relative to salmon that is not vacuum packed. For these reasons, we have continued to regard regular fillets and vacuum-packed fillets as separate forms of fresh Atlantic salmon.

   Comment 3: Averaging of Prices for Comparison to CV.

   The Association contends that the Department erred in the preliminary determination by comparing U.S. prices that were averaged by form, grade, and weight band to CVs that, due to the nature of the product, essentially do not vary except by form. The Association claims that salmon of different grades and weight bands have distinct physical differences resulting from natural variation in salmon populations, rather than from differences in production inputs or techniques. According to the Association, while the cost of production of a particular form of salmon (e.g., salmon fillets) may be the same regardless of differences in grades and weight bands, such differences affect the market value and selling price of salmon. The Association argues that, to make an apples-to-apples comparison, the Department should average all U.S. sales prices by form only and not by grade or weight band, such that a form-specific price is compared to a form- specific CV.

   According to the Association, the Department's practice in cases involving flowers and roses supports such an approach. The Association states that, in the Flowers cases (e.g.,Certain Fresh Cut Flowers from Colombia: Final Results of Antidumping Administrative Review, 55 FR 20491, 20496 (May 17, 1990) (Certain Fresh Cut Flowers from Colombia)(Comment 19)), the respondents were able to provide only an average cost for each type of flower, rather than a unique cost for each unique variety within the particular flower type. Under these facts, the Association contends, the Department found it appropriate to compare an average price for each flower type to the average CV of that flower type. Similarly, in the Roses cases (e.g., Fresh Cut Roses from Colombia, 60 FR 6980, 6990 (February 6, 1995) (Comment 5)), where the Department had the same cost for different rose types, the Department averaged the prices of roses across types prior to comparison to CV. The Association argues that there is no material difference in the fact pattern of the flowers cases compared to the fact pattern of this investigation. According to the Association, failure to conduct price- to-CV comparisons on a form-average basis in this case would violate not only the statutory requirement for a fair comparison, but also violate the fair-comparison requirements imposed by the GATT/WTO. The Association also argues that such a methodology would run counter to the findings of a GATT panel with respect to the LTFV investigation of salmon from Norway.

   The petitioners respond that the antidumping statute directs price- to-CV comparisons to be based on the prices and costs of each unique product, as defined by the physical characteristics of those products. According to the petitioners, the respondents could have reported costs of production specific to different weight bands and grades, but opted not to do so. Specifically, the petitioners argue that the respondents could have attempted to differentiate costs for weight bands based on differences in feed conversion ratios, and for grades based on differences in post-harvest costs. The petitioners argue that it would be inappropriate to correct this deficiency in the respondents' reporting by averaging U.S. prices, since there are price differences corresponding to differences in weight bands and grade.

   DOC Position: We disagree with the Association. For the final determination, we have continued to average U.S. prices by form, grade, and weight band.

   We accept the Association's contention that, with minor exceptions, each company's recorded costs of the subject merchandise do not vary by grade or weight band. Our examination of the voluminous record evidence concerning this issue, including our verification findings, confirms that the costs as reported reasonably reflect the actual costs of producing each matching group (i.e., each combination of form, grade, and weight band), and that the costs of certain of these matching groups are the same. In this respect, we disagree with the petitioners' arguments that the respondents should have been required to report costs based on methodologies that deviate from their normal accounting practices, e.g., through the use of feed conversion ratios, in order to estimate differences in costs.

   With this in mind, when comparing U.S. prices to CV, the Department is charged with determining whether sales are made to the United States at prices below the actual cost of production. The CAFC has ruled definitively on this issue:

   By its terms, the statute expressly covers actual production costs * * *. The broad language of section 1677b(e) [the CV portion of the statute] does not at any point expressly authorize adjustment of these production costs to account for products of a lower grade or less value.

   See IPSCO Inc. v. United States, 965 F. 2d, 1056, 1059-1060 (Fed. Cir. 1992)(IPSCO).

   As in the instant proceeding, IPSCO involved merchandise (steel pipe used for oil and gas wells) that varied in grade (prime and limited service) but not in the cost of producing each grade. As with salmon, the same materials, processes, labor, and overhead went into the production of both grades, and buyers purchased both grades ``for the same purpose--``down hole'' use in oil and gas wells.'' Id. at 1058. Thus, both grades had the same actual costs:

   Because IPSCO expended the same materials, capital, labor, and overhead for both grades of OCTG, the constructed value of one ton of limited-service pipe necessarily matched the constructed value of one ton of prime pipe.

   Id. at 1060.

   As with premium salmon, prime-grade pipe was of a higher quality and, as such, commanded a higher price in the marketplace. Id. at 1058. In the proceeding underlying the IPSCO decision, the Department compared U.S. sales of prime-and limited-service grade pipe to CVs based on the actual costs of each grade, which were identical. There, as here, the respondents objected to this methodology vis-a-vis comparisons involving U.S. sales of the lower grade

   [[Page 31417]]

   of merchandise. The CAFC rejected this claim, ruling that the Department had ``calculated constructed value precisely as the statute directs'' in basing CV on the actual cost of production for each grade. Id. at 1060.

   While making the same complaint as that made by the respondent in IPSCO, the respondents in the instant proceeding have proposed a different solution. Rather than arguing for an adjustment to CV, the respondents suggest that the Department average the reported U.S. prices without respect to two of the three matching characteristics (grade and weight band) for comparisons involving CV.

   We reject the respondents' proposal for the following reasons. First, no change to either side of the antidumping analysis (EP/CEP and normal value) is necessary because, in accordance with IPSCO and with a basic tenet of the dumping law, the Department's methodology in this case properly compares the price of U.S. sales of a given product with the actual costs of that product where normal value is based on CV, without regard to whether that product's actual costs are the same as, or different from, other products under investigation.

   Further, the methodological changes proposed by the respondents are inappropriate under the facts of this case to the extent that they conflict with other requirements imposed by the statute and Department practice. Specifically, the proposal to eliminate two of the three matching criteria from our analysis with respect to CV comparisons would reduce the accuracy of that analysis and, depending on the manner employed, would either eliminate price-based matches entirely, or would result in inconsistent matching groups depending on whether a U.S. sale is matched to comparison market sales or to CV.

   Pursuant to sections 771(16) and 773(a)(1) of the Act, it is our practice first to match U.S. sales with comparison market sales of the most physically comparable merchandise. We require the matching categories to be as precise as possible in order to effect a meaningful comparison:

   In determining the comparability of sales for purposes of inclusion in a particular average, Commerce will consider factors it deems appropriate, such as the physical characteristics of the merchandise, the region of the country in which the merchandise is sold, the time period, and the class of customer involved.

   Statement of Administrative Action accompanying the URAA (SAA) at 842 (emphasis added). Thus, the statute and SAA recognize the importance of developing, under the facts of each case, matching categories that allow for meaningful comparisons, while preventing, to the extent possible, the masking of dumping through overly broad averages. The discretion afforded to the Department by the SAA (to consider such factors as it deems appropriate) reflects the fact that this is arguably the most case-specific aspect of the dumping analysis, depending as it does on the particular characteristics of the product under investigation.

   In light of the importance of determining our matching categories, it is our longstanding practice to consider comments submitted by interested parties regarding the relevant matching characteristics of the product under investigation. Early in this proceeding, both parties agreed that form, weight band, and grade were critical physical characteristics of fresh Atlantic salmon. See letter from the Association to the Department of Commerce, (August 7, 1997); see also letter from the petitioners to the Department of Commerce (August 7, 1997). Having established these matching categories, we averaged U.S. and comparison market sales of these product groups and made price-to- price matches, where possible. Only where we could not make such matches did we resort to CV. We have based CV on the actual costs of each matching category; where the respondents reported differences in actual costs (e.g., Marine Harvest's reporting of different costs by weight band), we have taken this into account.

   Significantly, in arguing that we should eliminate two of the three matching characteristics with respect to CV comparisons, the respondents do not address the fact that, unlike the Flowers line of cases, this investigation involves price-to-price matches that were made using matching characteristics (form, grade, and weight band) that the respondents themselves agreed were the defining features of the subject merchandise in terms of our matching groups. Their argument does not address the inconsistency of maintaining one set of averaging and matching characteristics (form, grade, and weight band) for one set of U.S. sales (those for which we are able to find a price-based match), while averaging and matching other U.S. sales (the remainder) according to form alone. The contingency of whether a given U.S. sale has a priced-based match or a CV-based match would not be an appropriate means of determining the averaging methodology for that sale.

   When the respondents first raised this issue, it appeared that they would have resolved this inconsistency by eliminating price-based matches altogether for any company that would have any CV matches (all of them). See Mares Australes Section A and B Questionnaire Response (November 3, 1997) at 4 (``We suggest that because there are U.S. grades that do not match, the Department reject Japanese sales entirely as the basis for normal value and rely instead upon constructed value.'' (citing Roses from Colombia, Roses from Equador, and Fresh Cut Flowers from Colombia)). ‹SUP›4‹/SUP›Since the respondents have not addressed in the case briefs how to treat U.S. sales that would otherwise have suitable price-based matches, it is not clear whether the respondents continue to advocate this approach. We note for the record that we also disagree with this proposal, as it would undermine the statutory preference for price-to-price matches, as reinforced by the CAFC's decision in Cemex v. United States, WL 3626 (Fed. Cir.).

   \4\ We note that this argument by respondents for rejecting Japanese sales is separate from their argument that we should disregard such sales due to a particular market situation, as addressed in Comment 4, infra.

   Here again, the analogy to the Flowers cases fails, and serves only to illustrate why the SAA explicitly instructs the Department to use its discretion in determining the appropriate matching methodology under the facts of each case. To state the obvious, flowers and salmon are different products that are sold in different markets under different conditions. While we have determined to date in the Flowers line of cases that the merchandise and markets involved do not permit reasonable price-based comparisons (due to, for example, the holiday- driven demand patterns in the U.S. market), that is not the case with the merchandise and markets involved in this investigation. It is not appropriate to force such a case-specific finding involving the physical characteristics of flowers, and the selling practices that relate specifically to flowers, onto the matching methodology for fresh Atlantic salmon, thereby effectively eliminating the valid methodology developed early in this case. We would likewise disagree with the concept of averaging U.S. sales that have price-based matches only with respect to form, as this would undermine the precision of our analysis with respect to such sales.

   Finally, with respect to the relevance of the 1992 GATT panel report in United States: Imposition of Antidumping Duties on Imports of Fresh and Chilled Atlantic Salmon from Norway, we note that the panel's findings were limited, by the panel's

   [[Page 31418]]

   own terms of reference, to the facts of that pre-Uruguay Round proceeding. Moreover, the GATT panel faulted the Department for its lack of an explanation regarding its matching methodology in the Norwegian salmon case:

   While the United States had explained that because of the absence of differences in costs of production between salmon of different weights no separate constructed values for individual weight categories had been calculated, the United States had not put forward any arguments to explain why export prices of individual weight categories had been used in the comparison with the single constructed values. The public notice of the affirmative final determination was also silent on this point.

   Id. at. 470.

   Unlike the Norway case, we have provided a detailed explanation for our methodology in this respect.

   Comment 4: Particular Market Situation in Home Market.

   The Association argues that the Department erred in finding that a particular market situation exists in the home market, and disputes the Department's underlying conclusion that the home market is an incidental market consisting of sales of non-export quality salmon. The Association contends that the home market unquestionably passes the statutorily mandated viability test, and that the merchandise sold in that market is within the scope of the investigation. According to the Association, the Department's finding of a particular market situation is based on an unprecedented and extra-statutory consideration of the amounts and percentages of each grade of merchandise sold in the home market, compared to the merchandise sold in the United States. The Association asserts that any such differences can be adjusted for under the Department's normal calculation methodologies, and do not warrant rejection of the home market.

   The Association argues that, in the alternative, the Department should also find that a particular market situation exists in the Japanese market. According to the Association, the differences between the salmon sold by the respondents in Japan and that sold in the United States are greater than those between the salmon sold in the home market and that sold in the United States.

   The petitioners respond that the Department properly rejected the home market as a comparison market. According to the petitioners, the Department had ample statutory and regulatory authority to make a finding of a particular market situation with respect to the home market, and properly concluded that the Chilean market is incidental to the export-based Chilean salmon industry.

   The petitioners further argue that the Japanese market does not present a particular market situation, since any differences between the salmon sold in Japan and that sold in the United States are minor distinctions within export-quality merchandise. The petitioners urge the Department to continue its reliance on the Japanese market as the basis for normal value for the respondents in question.

   DOC Position: We agree with the petitioners. The Department's reasons for rejecting the use of the home market were set forth in detail in a memorandum addressing this issue. See Memorandum from Case Analysts to Richard Moreland, Regarding Appropriateness of Chilean Market as a Comparison Market (October 17, 1997) (Particular Market Determination Memorandum). As explained in that memorandum, the home market is incidental to the Chilean salmon industry, which is export- oriented. The home market is comprised almost exclusively of salmon graded by the respondents as ``industrial'' or ``reject,'' which the respondents sell locally for drastically reduced prices compared to export merchandise. The perfunctory marketing and distribution of salmon in the home market is consistent with the incidental nature of those sales.

   The Association has not raised substantial new arguments in its case brief, and instead has reiterated arguments advanced prior to the preliminary determination. We therefore refer interested parties to our Particular Market Determination Memorandum and to the Memorandum from Gary Taverman to Richard Moreland, Issues Concerning the Preliminary Determination of Sales at Less Than Fair Value (January 8, 1998) (Preliminary Issues Memorandum) for more detailed discussions of the issue.

   With respect to the Association's claims regarding the home market, we add only that our verification findings refuted one of the Association's arguments regarding this issue. The Association characterizes the difference between the home market and the United States as one of differences in ``product mix,'' suggesting that the same grades of merchandise are sold in both markets, only in different proportions. This contention has been premised to a large extent on a claim that one of the respondents had exported ``industrial'' grade salmon to the United States, albeit in small quantities, and that this merchandise was identical to that sold in the home market. However, as we found at verification, the U.S. sales in question in fact were not of industrial-grade salmon, but rather of premium-grade salmon that was subject to a post-sale quality claim. The Association now recognizes that these sales were reported improperly. See Association rebuttal brief at 54. Thus, the record clearly establishes that the grade of merchandise sold by the respondents in the home market is not exported to the United States or Japan.

   We also continue to find that the Japanese market does not present a particular market situation. As explained in our Preliminary Issues Memorandum, the respondents' Japanese market is far from incidental. Moreover, as explained above in response to Comment 1, the premium- grade salmon sold in the United States and the super-premium salmon sold in Japan are essentially the same merchandise. By contrast, as ascertained at verification, the salmon sold in the home market have severe defects. See Eicomar Verification Report at 3 (noting ``severe scale loss, greenish outer color, and numerous red spots due to early sexual maturation''); see also Marine Harvest Sales Verification Report at 7-8 (noting ``deformed mandibles, greenish-brownish external color, and marked lacerations'').

   Comment 5: All-Others Rate.

   The Association argues that the Department's exclusion of de minimis rates from the calculation of the ``all-others'' rate violates the constitutional due process and equal protection rights of Chilean producers/exporters of subject merchandise and their U.S. importers. According to the Association, exclusion of de minimis rates results in an unrepresentative and skewed all-others rate, because the Department limited its investigation to a minority of producers/exporters, did not accept voluntary participation by other firms, and found that the majority of the investigated firms were not dumping. The Association contends that the Court of International Trade (CIT) expressly stated in Serampore Indus. Pvt. Ltd. v. United States Dep't of Commerce, 696 F. Supp. 665, 668 (Ct. Int'l Trade 1988) (Serampore) that where the Department limits the number of firms to be investigated, there is no basis for excluding de minimis margins in the calculation of the all- others rate.

   The petitioners respond that the Department is bound by the plain language of the antidumping statute to exclude de minimis rates from the calculation of the all-others rate. According to the petitioners, Serampore

   [[Page 31419]]

   is specific to situations where the Department selects a sample of firms for investigation from among a much larger group of potential respondents. The petitioners note that in this case the Department did not select a sample of firms, but chose instead those exporters accounting for the largest volume of exports to the United States during the POI. The petitioners also point out that the Association specifically requested at the outset of this proceeding that the Department limit its investigation to those producers/exporters accounting for 50 percent of the exports during the POI, and note that those companies investigated account for approximately that figure.

   DOC Position: We agree with the petitioners. Section 735(c)(5)(A) of the Act unambiguously directs the Department to exclude ``any zero and de minimis margins'' from the calculation of the estimated all- others rate (emphasis added). There is no indication in the legislative history of this provision that Congress intended for exceptions to this rule. We therefore have no basis to ignore the Act's clear directive to exclude de minimis margins from the calculation of the estimated all- others rate.

   Further, as the petitioners note, the Association itself requested that the Department limit its selection of firms to be investigated to those exporters accounting for 50 percent of exports to the United States, in addition to ``a relatively small number of volunteer respondents.'' See letter from the Association to the Department of Commerce (August 4, 1997), at 4-6. The Department selected a pool of exporters accounting for very close to that volume of exports, and the Association did not voice its concerns about the implications of limiting the number of respondents with respect to the all-others rate until after the preliminary determination was issued.‹SUP›5‹/SUP›

   \5\ In accordance with section 777A(c)(2) of the Act, the Department limited its investigation to the five largest producers/ exporters. However, in limiting its investigation, the Department stated that if a selected respondent failed to cooperate, and companies wishing to be treated separately as voluntary respondents had submitted a response to our antidumping questionnaire, the Department would consider replacing the uncooperative respondent with a voluntary respondent, to be selected based on the order of each company's submission of a written request for investigation as a voluntary respondent. See Memorandum from the Team to Richard Moreland, Regarding Selection of Respondents (August 26, 1997), at 6.

   Comment 6: Industry Support for the Petition.

   The Association argues that the Department should not have initiated this antidumping investigation because the petitioners did not demonstrate sufficient industry support for the petition. The Association claims that the petition identified only U.S. producers of whole salmon, and failed to identify U.S. producers of cuts of fresh Atlantic salmon (``fillet producers''), which were also under the scope of the petition. The Association contends that fillet producers comprise an industry separate from the whole salmon industry.

   The Association argues further that, even if these two segments can be considered one industry, such that production from these two segments could be combined in the industry support ratio, the Department should have polled the fillet producer portion of the industry rather than derive an estimate of such production. The Association asserts the following errors in the Department's estimate of fillet production: (1) the calculation inappropriately estimates the size of the fillet producer industry on the basis of the value added in the processing of whole salmon into salmon cuts, rather than on the basis of the total value of the salmon cuts; (2) it focuses only on the basic processing of whole salmon into fillets, ignoring ``higher value- added products,'' such as portions; and (3) it relies on the cost data derived from a single source, rather than from a variety of sources.

   The petitioners respond that the Department appropriately determined that there was industry support for the petition on the basis of data in the petition as well as data gathered from external sources. According to the petitioners, the Act does not require polling to determine the domestic industry under such circumstances.

   DOC Position: Section 732(c)(4)(E) of the Act provides that, after the administering authority determines that it is appropriate to initiate an investigation, the determination regarding industry support shall not be reconsidered. Therefore, we have not reconsidered our determination regarding industry support. We refer interested parties to our notice of initiation and companion memorandum, which set forth in detail the methodologies followed in establishing industry support. See Initiation of Antidumping Duty Investigation: Fresh Atlantic Salmon From Chile, 62 FR 37027, 27028-29 (July 10, 1997).

   Sales Issues--Aguas Claras

   Comment 7: Use of the Canadian Market as Comparison Market.

   The petitioners contend that the Department should reject Aguas Claras' sales to the Canadian market as the basis for normal value for three reasons: (1) the Canadian market is an unimportant market for Chilean salmon exporters as a whole, such that prices to this market are not ``representative'' within the meaning of section 773(a)(1)(B)(ii)(I) of the Act; (2) the particular market situation in Canada renders that market an improper comparison market; and (3) verification findings indicate that the reporting of Canadian fillet sales is unreliable.

   The petitioners first argue that prices to Canada are not representative because total Chilean exports of fresh Atlantic salmon to Canada constitute a minuscule percentage of Chile's worldwide exports of that merchandise, i.e. Canada is an unimportant market. Citing the preliminary results of the tenth administrative review of Flowers from Colombia, 63 FR 5354, 5357 (February 2, 1998), the petitioners claim that the Department recently rejected the use of Canada and Japan as comparison markets where: (1) the Department did not examine all potential respondents, such that the rate for non- selected companies would be based on an average of the rates found for the respondents; and (2) exports to the Canadian market were a small percentage of total exports. The petitioners claim the same facts apply to the instant proceeding.

   The petitioners' second argument, that a particular market situation in Canada renders that market an improper comparison market, rests on the following claims: (1) the narrow margin of the five- percent viability determination, which was affected by the timing of Aguas Claras' acquisition of its U.S. affiliate, Bowrain Corp., during the POI; (2) the existence of a high degree of integration in the channels of trade for subject merchandise in the United States and Canada, which, petitioners assert, renders Canada an inappropriate comparison market because it is essentially the same market as the U.S. market; and (3) the recent Canada/Chile free trade agreement, which ended each country's right under the GATT to initiate antidumping proceedings against each other and, according to the petitioners, has rendered Canada a secondary dumping ground.

   Finally, the petitioners argue that the Department's verification findings suggest that Aguas Claras' reporting of Canadian market sales of fillets is unreliable and that the Department must resort to CV for such sales.

   Aguas Claras responds that there is no reason for rejection of the Canadian market as the basis for normal value. First, with respect to the allegation that

   [[Page 31420]]

   the Canadian market is unimportant to the Chilean exporters as a whole such that prices to this market are unrepresentative, Aguas Claras contends that the Department's decision in the tenth review of Flowers from Colombia is factually distinguishable because, in the Flowers proceedings, the Department has consistently rejected price-based normal values for all respondents. Thus, the respondents argue, the Department's rejection of Japan and Canada as comparison markets in the tenth Flowers review was consistent with its general practice in the Flowers proceedings. Aguas Claras further argues that the export statistics cited by the petitioners are based on direct exports, and thus mis-classify sales to Canada made through the United States as U.S. sales. According to Aguas Claras, all of its own sales to Canada were made through this route. Therefore, Aguas Claras concludes, there is no basis for a finding that the Canadian market is unimportant.

   Second, with respect to the allegation that there is a particular market situation in Canada, Aguas Claras argues that the Canadian market passes the ``bright line'' (five-percent) test for viability, and maintains that no heightened standards should be applied to that market. Aguas Claras adds that the high degree of integration between the U.S. and Canadian salmon markets actually supports the use of Canada as the basis for normal value, because similarities between the two markets support a finding that there is no particular market situation in Canada that would render prices in that market not comparable to U.S. prices.

   Finally, with respect to the verification findings cited by the petitioners, Aguas Claras argues that there is no evidence of any price distortions in the Canadian market with respect to fillet sales.

   DOC Position: We disagree with the petitioners that the Canadian market is characterized by ``unrepresentative'' prices or by a particular market situation, within the meaning of sections 773(a)(1)(B)(ii)(I) and (II) of the Act. However, we agree with the petitioners that, based on our verification findings, we are unable to match Aguas Claras' POI Canadian sales of fillets, as reported, to its U.S. sales. We have based normal value for such sales on CV.

   To address the petitioners' arguments in turn, we first disagree that the Canadian market is characterized by unrepresentative prices. Contrary to the petitioners' assertions, the recent finding in the preliminary results of the tenth review of Flowers from Colombia does not compel the rejection of an otherwise viable Canadian market in the instant proceeding. As we state in our response to Comment 3, above, the Flowers cases have relied on CV as the sole basis for normal value for each of the past 10 reviews, for a variety of product- and market- specific factors that do not pertain to this investigation (e.g., holiday demand patterns). The unique history of the market-selection determinations made in the Flowers and Roses cases does not lend itself to broad application of those findings to a salmon respondent that, as verification demonstrated, sells to a viable Canadian market in the same manner, and through the same channels of distribution, as it sells to the U.S. market.

   We also disagree with the basis of the petitioners' numerical analysis regarding exports to Canada versus exports to the United States vis-a-vis their ``unrepresentative prices'' argument. As Aguas Claras correctly notes, all of its own sales to Canada were made through its U.S. affiliate in Miami, after entry of the merchandise into the United States. The effect of this distribution pattern is to inflate significantly the apparent volume of exports to the United States, and to deflate the apparent volume of exports to Canada. The size of this distortion of ``direct'' export numbers with respect to the one company whose Canadian sales we are examining is a reasonable indication that the overall export figures provided by the petitioners understate the volume of Chilean fresh Atlantic salmon that is destined for the Canadian market. The Department has not found any statistics establishing the ultimate destination of merchandise exported by the Chilean industry. Therefore, in view of the demonstrated viability of the Canadian market for Aguas Claras, and in the absence of persuasive evidence to the contrary, we have not rejected Canadian sales prices as unrepresentative.

   Regarding the petitioners' particular market situation claim, we agree with Aguas Claras that similarities between the U.S. and Canadian markets are not evidence of a particular market situation. As for the contention that Canada has become a secondary dumping ground due to the terms of the Canada/Chile Free Trade Agreement, we note that such trade agreements are not designed to promote dumping, and their mere existence is not evidence of such. In addition, the below-cost test that we have applied to sales made by Aguas Claras in the Canadian market prevents the inclusion of such sales, when made in substantial quantities, in our analysis.

   However, we agree with the petitioners' argument that our verification findings call into question the reporting of certain data essential to price-to-price comparisons, specifically with respect to fillets.‹SUP›6‹/SUP› Although we do not agree that this is sufficient to disregard the Canadian market in its entirety, we have rejected the use of price-based comparisons for fillets, and have instead compared U.S. fillet sales to CV. For sales of whole fish, which are unaffected by the problem involving fillets, we have made price-to-price comparisons where otherwise appropriate. For a detailed explanation of this methodology, see Aguas Claras Analysis Memorandum.

   \6\ We cannot address the specifics of the verification finding in this public forum, as a meaningful discussion is only possible by means of reference to business proprietary information. We have addressed the petitioners' argument in a separate memorandum to the file, which will be placed on the official record and served upon parties with access to such information under administrative protective order. See Memorandum from the Case Analyst to Gary Taverman, Regarding Analysis of Aguas Claras Data for Final Determination (June 1, 1998)(Aguas Claras Analysis Memorandum).

   Comment 8: Sales by Affiliated Producer/Exporter.

   The petitioners argue that Aguas Claras failed to report U.S. sales made by an affiliate, Pesquera Invertec, that produced and exported subject merchandise during the POI. The petitioners state that the existence of these sales was found only at verification, a situation that warrants the application of the facts available to derive the dumping margins on such sales. Noting that the Department obtained the total volume of Pesquera Invertec's U.S. sales at verification, the petitioners argue further that the inclusion of this figure in Aguas Claras' total U.S. sales causes the Canadian market to drop below the Department's viability threshold. The petitioners state that this constitutes another reason for the Department to reject the use of the Canadian market as a comparison market (in addition to the arguments made in Comment 7, above) and compare U.S. prices to CV.

   Aguas Claras responds that it has never been affiliated with Pesquera Invertec, and was never required to report that exporter's sales. According to Aguas Claras, Pesquera Invertec was affiliated for part of the POI with Aguas Claras' parent company, Antarfish S.A. (Antarfish), by virtue of their joint control of a salmon processing company. However, Aguas Claras argues, there is no transitive principle of affiliation in the statute, such that Antarfish's affiliation with Pesquera

   [[Page 31421]]

   Invertec would extend to Aguas Claras. Aguas Claras contends that it reported all of its own sales, and those of its affiliates, but was never requested to report the sales of its affiliates' affiliates.

   Aguas Claras further argues that even if it were deemed to be affiliated with Pesquera Invertec, there would be no basis for collapsing the two companies and requiring the reporting of the latter's U.S. sales. In this respect, Aguas Claras maintains that the Department collapses affiliated companies only where there is such a high degree of integration between the companies' operations that there is a significant potential for price manipulation. Aguas Claras claims that verification established that, at most, Antarfish was only distantly affiliated with Pesquera Invertec during part of the POI through joint ownership of a processing facility, but that the two companies were not otherwise related. Aguas Claras also states that, prior to the end of the POI, Antarfish fully divested itself of its interests in the processing facility, such that there is no potential for future price manipulation.

   Finally, Aguas Claras argues that it could not have provided Pesquera Invertec sales data even if requested to do so, because Antarfish and Pesquera Invertec are involved in a business dispute, and Pesquera Invertec would not have supplied those data. According to Aguas Claras, the application of adverse facts available is only appropriate where a party has demonstrably failed to act to the best of its ability; therefore, it would be inappropriate to penalize Aguas Claras with respect to information that was not within its control.

   DOC Position: We disagree with the petitioners that Pesquera Invertec's sales should have been included in Aguas Claras' sales database. Even if we were to assume, arguendo, that Aguas Claras was affiliated with Pesquera Invertec for part of the POI, the record does not warrant collapsing these two parties. The Department's practice is to collapse affiliated producers when the companies: (1) have production facilities that are sufficiently similar so that a shift in production would not require substantial retooling; and (2) present a significant potential for the manipulation of price or production. See 19 CFR 351.401(f) of the Department's regulations. See also, Cement From Mexico at 12774. As detailed below, it would be inappropriate to collapse Aguas Claras and Pesquera Invertec because there is not a significant potential for the manipulation of price or production.

   As provided at section 351.401(f)(2) of our regulations, we consider three factors in identifying a significant potential for the manipulation of price or production: (1) the level of common ownership; (2) the extent to which managerial employees or board members of one firm sit on the board of directors of an affiliated firm; and (3) whether operations are intertwined, such as through the sharing of sales information, involvement in pricing and production decisions, etc. In examining these factors as they pertain to a significant potential for manipulation, we consider both actual manipulation in the past and the possibility of future manipulation. See Preamble to Final Regulations, 62 FR 27296, 27346 (May 19, 1997). The preamble underscores the importance of considering the possibility of future manipulation: ``a standard based on the potential for manipulation focuses on what may transpire in the future.'' Id. We have, therefore, examined all three factors in light not only of actual manipulation during the POI but also with respect to the possibility of future manipulation.

   Applying these criteria to this case, Aguas Claras and Pesquera Invertec do not, and did not during the POI, have common stock ownership or common directors on their respective boards, as confirmed at verification. See Memorandum from Case Analysts to Gary Taverman, Regarding Verification of Sales by Aguas Claras (April 7, 1998) (Aguas Claras Sales Verification Report) at 3 and Exhibits A-15 and A-16. Thus, the first two factors suggest no potential manipulation during the POI or in the future. Regarding the third factor, Aguas Claras' parent company, Antarfish, fully divested itself of its participation in the processing facility it jointly owned with Pesquera Invertec, and ceased any processing of salmon at that plant. Moreover, at verification we reviewed extensive documentation involving arbitration proceedings over a significant business dispute between Pesquera Invertec and Antarfish.

   See Aguas Claras Sales Verification Report at 3-4 and exhibit A-15. As for the possibility that Aguas Claras/Antarfish and Pesquera Invertec engaged in price or production manipulation during the POI, we note that only a very small percentage of Aguas Claras/Antarfish's sales of subject merchandise were processed at the facility owned jointly with Pesquera Invertec, and the vast majority of Aguas Claras/ Antarfish salmon was processed at Aguas Claras' own plant. Further, as part of our cost verification testing, we reviewed transactions between affiliates and specifically examined whether the company had transactions with Pesquera Invertec. We did not find any such transactions. See Aguas Claras Cost Verification Report at 6 and exhibit B-2. Thus, we did not find evidence that the two companies' operations were significantly intertwined during the POI, or that they shared sensitive business data.

   Accordingly, because Aguas Claras and Antarfish share no common stock ownership or board members with Pesquera Invertec, and Antarfish terminated its relationship with Pesquera Invertec during the POI, we find no evidence to suggest a significant possibility for the manipulation of price or production, and we have determined that it would not be appropriate to collapse Aguas Claras and Pesquera Invertec.

   Comment 9: CEP Offset.

   The petitioners argue that the Department erred in making a CEP offset adjustment to normal value. According to the petitioners, Aguas Claras' U.S. and Canadian sales are made through the same sales affiliate, which performs exactly the same functions for both kinds of sales. The petitioners contend that, in determining the level of trade of U.S. sales, the Department ignored selling functions associated with the U.S. affiliate's CEP selling expenses, and erroneously concluded that the level of trade of Canadian sales was more advanced. The petitioners argue that such a comparison, and the resulting CEP offset adjustment, ignores commercial reality, and that the CIT has rejected such ``automatic'' CEP offset adjustments, citing Borden et al. v. United States, Slip Op. 98-36 (March 26, 1998).

   Aguas Claras responds that the Act explicitly directs the Department to determine the level of trade of CEP sales based on the price as adjusted, i.e., after deducting CEP selling expenses, and to ignore the selling functions associated with those expenses.

   DOC Position: We agree with Aguas Claras. As discussed in detail in the preliminary determination, the Act requires us to determine the level of trade of CEP sales without consideration of the selling functions associated with economic activities in the United States. See Preliminary Determination at 2670. See also section 351.412(c)(ii) of the Department's new regulations (62 FR 27495 and preamble at 27370- 27371). Based on this analysis, we continue to find that the level of trade of Canadian sales is more advanced than the level of trade of U.S. sales. Therefore, we have made a CEP offset to normal value. With respect to the petitioners' claim that the CIT recently overturned the

   [[Page 31422]]

   Department's practice of comparing the level of trade of comparison market sales to a constructed level of trade for CEP sales in Borden et al. v. United States, we note that the Department is in the process of considering the Court's remand order.

   Comment 10: Adjustment to Cash Deposit Rate for Re-Exports to Canada.

   Aguas Claras argues that its cash deposit rate should be adjusted to account for the fact that it routinely re-exports a portion of its U.S. inventory of salmon to Canada. With respect to such inventory, Aguas Claras states that entries that result in re-exportation are not liable to assessment of antidumping duties, yet U.S. importers must post antidumping cash deposits for all entries into the United States, since there is no way to identify at the time of entry those products that will ultimately be sold to Canada. In view of this, Aguas Claras argues that the Department should lower the cash deposit rate so that the total deposits collected do not exceed the total duties ultimately assessed on sales of subject merchandise. Aguas Claras contends that the Department made such an adjustment in cases involving flowers imported from Colombia, where consignment importers resell a portion of their U.S. inventory to Canada.

   Petitioners argue that, given the small size of the Canadian market, there is no guarantee that Aguas Claras will continue to make sales to Canada, and that it would be improper to lower Aguas Claras' calculated deposit rate to account for some hypothetical volume of U.S. entries that might be re-exported to Canada in the future.

   DOC Position: We agree with the petitioners that it would be inappropriate to adjust Aguas Claras' cash deposit rate. The cash deposit rate applies to all entries entered into the United States for purposes of consumption. The fact that Aguas Claras made sales to Canada during the POI is not an indicator of the likely volume of future sales, nor a guarantee of any future sales, to that market, particularly in light of the small portion of U.S. imports that were re-exported to Canada. Therefore, it would be inappropriate to reduce the cash deposit rate applicable to all entries of subject merchandise into the United States to account for past re-exportation of subject merchandise to Canada.

   The adjustment to cash deposit rates in the Flowers cases was made under a materially different fact pattern. In those cases, the Department found that a portion of entries of flowers into the United States are never sold due to perishability problems, and are instead destroyed. Because those products are inherently perishable, and it is reasonable to expect a percentage of entries of those products to go unsold in any given period, the Department found it appropriate to make a reduction to the cash deposit rate. Although the flowers respondents also re-exported a portion of their flowers to Canada, that was not the rationale for the adjustment to the cash deposit rate. See Certain Fresh Cut Flowers from Colombia at 20494.

   Comment 11: Allegation of Affiliation with Kenbourne International.

   Aguas Claras disputes the petitioners' allegation that Aguas Claras and its wholly-owned U.S. sales affiliate, Bowrain Corp., are affiliated with Kenbourne International, the Miami-based company that administers importer sales activities on behalf of Bowrain Corp.\7\ With respect to the nature of the relationship between these companies, Aguas Claras states there are no stock relationships or common officers between Aguas Claras/Bowrain Corp. and Kenbourne International. According to Aguas Claras, Bowrain Corp., which is incorporated in Florida but whose officials work for Aguas Claras in Chile, retained Kenbourne International to function as a U.S. consignment agent. Aguas Claras states that Bowrain Corp. has always required Kenbourne International to maintain a separate set of books and records for Aguas Claras sales, and shipments of Aguas Claras' merchandise are never recorded in Kenbourne International's own inventory, so that Bowrain Corp. retains significant control over its sales. Therefore, the respondent contends, Kenbourne International cannot be found to control Bowrain Corp., nor Aguas Claras itself.

   \7\ Aguas Claras' brief responds to allegations with respect to Kenbourne International made by the petitioners prior to the Department's preliminary determination. The petitioners did not reiterate these allegations in their case brief, but, as summarized below, did respond to Aguas Claras' comment in their rebuttal brief.

   In rebuttal, the petitioners argue that, consistent with case precedent involving exporter/agent relationships (see Final Results of Antidumping Duty Administrative Review: Furfuryl Alcohol from the Republic of South Africa, 62 FR 61081, 61088 (Nov 14, 1997) (Furfuryl Alcohol from South Africa), Kenbourne International should be deemed affiliated with Aguas Claras through an agency relationship. According to petitioners, Kenbourne International is in operational control of all aspects of U.S. imports of Aguas Claras merchandise, and thus is in a position to exercise direction over Aguas Claras.

   DOC Position: We agree with Aguas Claras, and have continued to regard Kenbourne International as unaffiliated with Aguas Claras and Bowrain Corp.

   Kenbourne International's role in the importation and sale of Aguas Claras' merchandise is that of an unaffiliated consignee. In all significant respects, this role is identical to that played by the consignees of other respondents in this proceeding (e.g., Aquastar, the consignee of Mares Australes). As discussed in detail in the preliminary determination, a consignment relationship alone is not sufficient basis for a finding of affiliation. See Preliminary Issues Memorandum at 4.

   The record of this investigation does not support the conclusion that the exporter (Aguas Claras) controls the consignee (Kenbourne International), or vice-versa. In Furfuryl Alcohol from South Africa, the Department found that the U.S. importer was an agent of the exporter and, therefore, was controlled by the principal/exporter. That is not the case here, as Kenbourne International is a consignee, not an agent (e.g., the two parties do not jointly market subject merchandise to U.S. customers, jointly negotiate prices/sales with U.S. customers, or interact with U.S. customers on product testing and quality control). Therefore, there is no basis on which to conclude that Aguas Claras controls Kenbourne International.

   There is also no basis for finding that Kenbourne International controls Aguas Claras. As noted above, Kenbourne International provides essentially the same services to Aguas Claras that unaffiliated consignees perform for the other respondents, and such services do not establish control of the exporter by the consignee. Other than these basic functions, the fact that Kenbourne International maintains a set of books and records on behalf of Bowrain Corp., and deposits revenues from sales of Aguas Claras merchandise into Bowrain Corp.'s bank accounts (after which Kenbourne International cannot access the revenues) is insufficient for a finding of affiliation based on control.

   Sales Issues--Eicosal

   Comment 12: Affiliation between Eicosal and its Consignee.

   The petitioners argue that Eicosal and its consignee, Stolt Sea Farm Inc. (Stolt Inc.), should be considered affiliated parties because Stolt Inc. is in a position to exercise control over Eicosal through the terms of a ``close supplier'' business arrangement.

   Eicosal argues that the Department should continue to find, as it did in the preliminary determination, that Eicosal and Stolt Inc. are not affiliated parties.

   [[Page 31423]]

   According to Eicosal, the two parties have no direct or indirect stock ownership in each other, nor do they have a close supplier relationship. Eicosal contends that, even if all of its salmon sales to the United States are made through Stolt Inc., its voluminous sales of salmon to other markets (such as Japan and Brazil) do not involve Stolt Inc. at all.

   DOC Position: We agree with the petitioners that Eicosal and Stolt Inc. are affiliated parties, although we base our finding on a different statutory basis from that alleged by the petitioners. Whereas the petitioners allege that the two parties are affiliated by virtue of a close supplier relationship (affiliation via ``control'' as per section 771(33)(G) of the Act), we find that the parties are affiliated by virtue of equity ownership exceeding five percent in accordance with section 771(33)(E) of the Act, and therefore do not reach the issue of affiliation via control.

   Stolt Inc. is a wholly-owned subsidiary of Stolt-Nielsen Holdings B.V. (Stolt-Nielsen). This parent company has another wholly owned subsidiary, Stolt Sea Farm Ltda. (Stolt Ltda.), which owns well over five percent of Eicosal's stock. In the preliminary determination, the Department found that this equity relationship was not sufficient to establish affiliation under section 771(33)(E) of the Act. The underlying presumption for this finding was that Stolt Inc. and Stolt Ltda. were separate (albeit affiliated) corporate entities. See Preliminary Issues Memorandum at 5 and n.3.

   At verification, however, the Department gained a greater understanding of the interrelationship of the Stolt companies, which suggests that Stolt-Nielsen, Stolt Inc., and Stolt Ltda. are effectively a single corporate entity. First, the Department learned that Stolt Ltda. was created for the purpose of allowing Stolt-Nielsen to hold an equity interest in Eicosal. See Memorandum from Case Analysts to Gary Taverman re: Verification of Sales made by Pesquera Eicosal Ltda (April 9, 1998) (Eicosal Sales Verification Report) at 4. Second, the Department found that Stolt-Nielsen's operational control over Stolt Inc. (its wholly-owned subsidiary) extended to Stolt- Nielsen's negotiation of the distribution arrangement with Eicosal. See Memorandum from analysts to Gary Taverman re: Verification of Sales Made by Pesquera Eicosal Ltda through Stolt Sea Farm Inc. (April 9, 1998) (Eicosal CEP Sales Verification Report) at 3. Moreover, the distribution arrangement with Eicosal was signed on the same day that Stolt Ltda. purchased its shares in Eicosal, which further indicates the extent of coordination between these companies with respect to their relations with Eicosal. See Eicosal Sales Verification Report at 4.

   In view of the above, we have determined that the Stolt companies (i.e., Stolt-Nielsen, Stolt Inc. and Stolt Ltda.) effectively constitute a single corporate entity (i.e., a person). For purposes of a dumping analysis, we believe that it is appropriate to view the equity interests of this single corporate entity in other companies in toto. Since this entity (of which Stolt Inc. is a part) owns in excess of five percent of Eicosal's stock, we find that Stolt Inc. is affiliated with Eicosal within the meaning of section 771(33)(E) of the Act.\8\

   \8\ The petitioners claim that Stolt Inc. effectively controls Eicosal through their contractual arrangement. We do not find that the contract between the parties per se establishes clear evidence of affiliation through control. In any event, the issue is moot as the Department has found the two parties to be affiliated by means of stock ownership.

   For purposes of this final determination, the finding of affiliation between Eicosal and Stolt Inc. does not preclude the use of the submitted U.S. sales data, since the Department had already requested that Eicosal report U.S. sales based on the prices charged by Stolt Inc. to the first unaffiliated U.S. customer. We note that in calculating CEP for sales made through affiliated parties (as opposed to unaffiliated consignees), the Department normally reduces the CEP by the amount of the actual selling expenses incurred by the affiliate, plus an amount for profit associated with those selling activities. In this case, we do not have such information for Stolt Inc., because the Department regarded Stolt Inc. as an unaffiliated party through the information-gathering stage. We do not believe that it would be appropriate to draw an adverse inference from this, as Eicosal submitted substantial and voluminous information about its relationship with the Stolt companies in its questionnaire responses. (That the Department developed a greater understanding of this relationship at verification does not imply that Eicosal withheld material evidence at the information-gathering of the proceeding.) Therefore, we have relied on the commission charged by Stolt Inc. to Eicosal in lieu of those selling expenses and the profit attributable to those expenses. However, in the event that an antidumping order is issued in this case and that Eicosal's sales become subject to administrative review, the Department will require that Eicosal submit sales data under the presumption that Eicosal and Stolt Inc. are affiliated parties, and will require the reporting of Stolt Inc.'s actual selling expenses.

   Comment 13: Ordinary Course of Trade.

   Eicosal argues that the Department erred in finding that its sales of vacuum-packed fillets to Japan were made in the ordinary course of trade, and in including these sales in the calculation of CV profit. According to Eicosal, the sales in question involved a small volume of a unique, specialized product, sold over a limited period of time to a single customer. Eicosal disputes the Department's finding in the preliminary determination that these sales were made continuously throughout the POI, contending that there were no shipments of vacuum- packed fillets in March 1997, and adding that all shipments of vacuum- packed fillets ended shortly after the end of the POI.

   The petitioners argue that the Department correctly found in its preliminary determination that Eicosal's sales of vacuum-packed fillets were made in the ordinary course of trade, as these sales were made continuously through the POI, involved significant quantities, and were not done on a test basis.

   DOC Position: We agree with the petitioners, and continue to find Eicosal's sales of vacuum-packed fillets to have been made in the ordinary course of trade.

   Section 773(a)(1)(B) of the Act provides that the Department may use third-country prices as the basis for normal value only where such prices are made in the ordinary course of trade. Prior to the preliminary determination, both Mares Australes and Eicosal argued that their respective sales of vacuum-packed fillets had been made outside the ordinary course of trade. In our preliminary determination, we found that Mares Australes' single sale of that merchandise had been made outside the ordinary course of trade, as the sale had involved a minute quantity of product sold on a test basis. In contrast, we found that Eicosal's sales of vacuum-packed fillets had been made within the ordinary course of trade, as they had been made regularly throughout the POI, and not on a test basis. See Preliminary Issues Memorandum at 12.

   The objections now raised by Eicosal do not warrant a reversal of our preliminary finding. While sales of vacuum-packed fillets may represent a small percentage of total sales, the absolute amount of these sales (several thousand kilograms) is not insignificant.

   [[Page 31424]]

   Also, Eicosal's claim that sales of vacuum-packed fillets were intermittent throughout the POI is not persuasive, since these sales were suspended only for the last month of the period, and resumed a month thereafter. In view of the volume of merchandise involved, the fact that the merchandise was sold regularly throughout the POI, and the lack of evidence that the sales were made on a sample basis, we continue to find that the sales in question were made in the ordinary course of trade.

   Comment 14: Advertising Expense.

   Eicosal argues that, in the preliminary determination, the Department incorrectly found an advertising expense incurred by Eicosal for its participation in the Japan/Chile centennial celebration to be a general promotional expense, and treated it as an indirect selling expense. Eicosal argues that this advertising expense (specifically, a fee that allowed it to display the celebration logo on its boxes of salmon), should instead be treated as a direct selling expense. Eicosal states that the expense meets the Department's two-prong test for classification of advertising expenses as direct expenses, as set forth in Antifriction Bearings (other than Tapered Roller Bearings) and Parts Thereof from France, Germany, Italy, Japan, Singapore, and the United Kingdom; Final Results of Antidumping Duty Administrative Reviews, 62 FR 2081, 2102 (January 15, 1997) (AFBs 94/95), namely that: (1) the expense be incurred directly in conjunction with sales of the foreign like product; and (2) the advertising be directed towards the customers' customer. Eicosal acknowledges that the promotional logo was displayed on boxes of seafood products other than fresh Atlantic salmon, but argues that a portion of the expenses nonetheless was incurred in direct connection with sales of subject merchandise. Further, Eicosal contends that these expenses do not meet the CIT's definition of ``general image'' advertising set forth in Brother Industries v. United States, 540 F. Supp 1341, 1366 (Ct. Int'l Trade 1982), aff'd, 713 F.2d 1568 (Fed. Cir. 1983), cert. denied, 465 U.S. 1022 (1984) (Brother Industries), i.e., such advertising is ``more in the nature of making consumers aware of the company's concern for consumers and the quality of its workmanship and product in general'' than in the nature of touting a specific product. Eicosal contends that because the promotional logos in question are applied to particular products, they constitute specific product advertising.

   The petitioners respond that the display of the centennial celebration logo on boxes of fresh Atlantic salmon does not specifically promote the sale of that product, but rather promotes goodwill between Chile and Japan, and therefore the associated expense cannot be treated as direct.

   DOC Position: We agree with the petitioners. The expenses in question do not meet the criteria for direct expenses, as described in AFBs 94/95. The nature of the centennial celebration was to promote goodwill, thereby promoting Eicosal's corporate image.

   The promotional logo applied to the boxes of fresh Atlantic salmon did not refer to salmon, nor even to Eicosal's general product lines. Therefore, we have continued to classify the expenses in question as indirect expenses.

   Comment 15: Adjustment to Cash Deposit Rate for Re-Exports to Canada.

   Eicosal argues that its cash deposit rate should be adjusted to account for the fact that it routinely re-exports a portion of its U.S. inventory of salmon to Canada. According to Eicosal, entries that result in re-exportation are not liable to assessment of antidumping duties, yet U.S. importers must post antidumping cash deposits for all entries into the United States, since there is no way to identify at the time of entry those products that are ultimately sold to Canada. In view of this, Eicosal argues, the Department should lower the cash deposit rate so that the total deposits collected do not exceed the total duties ultimately assessed on sales of subject merchandise.

   Petitioners argue that it would be improper to lower Eicosal's calculated deposit rate to account for a hypothetical volume of U.S. entries that might be re-exported to Canada in the future.

   DOC Position: We agree with the petitioners. For the reasons explained with respect to Comment 10 above (regarding similar arguments made by Aguas Claras), it is not appropriate to adjust the cash deposit rate for Eicosal to account for possible future entries of subject merchandise that might be re-exported to Canada in the future.

   Sales Issues--Mares Australes

   Comment 16: Unreconciled Revenues.

   The petitioners note that there is a discrepancy between the total value of sales in the database submitted by Mares Australes and the total value of sales in the database submitted by Mares Australes' consignees. To account for this discrepancy, the petitioners request that the Department reduce CEP prices by the ratio of the unreconciled sales amount to the total value of Mares Australes' sales.

   Mares Australes responds that the discrepancy noted by the petitioners was identified during verification in Chile, and was accounted for almost entirely at the outset of the subsequent CEP verification. Further, Mares Australes argues that the total value of sales of the consignee's database (which was the database relied on by the Department for its preliminary determination) was fully verified, and maintains that any remaining discrepancy with Mares Australes' initial database is insignificant.

   DOC Position: We agree with Mares Australes. The small discrepancy between the two databases found at verification in Santiago was almost entirely accounted for at the outset of the CEP verification. The remaining discrepancy is an insignificant amount, particularly given that it involves a comparison of databases maintained by separate companies at different points in the distribution chain.

   Comment 17: Canadian Sales Included in U.S. Sales Database.

   The petitioners argue that sales to Canada by one of Mares Australes' consignees should be removed from the U.S. sales database.

   Mares Australes argues that in the normal course of business it is not informed of the ultimate destination of merchandise shipped to the United States for consignment resale. According to Mares Australes, the Department's practice is to determine the market of destination according to the producer/exporter's knowledge of destination at the time of sale, and therefore the sales in question are properly included in the U.S. sales database.

   DOC Position: We disagree with Mares Australes. Even if Mares Australes was not aware at the time of sale that the transactions involved Canadian customers, the fact remains that Mares Australes' consignee clearly identified the transactions as Canadian sales in its submitted database.

   The Department's ``exporter knowledge'' rule is typically applied where the respondent ships merchandise to a reseller and is aware at the time of sale that the merchandise is ultimately destined for the United States. In this case, Mares Australes' sales to both the United States and Canada are made through consignees, who set the terms of sale on behalf of Mares Australes, and have ultimate knowledge of the location of the customer. In preparing its sales database, Mares Australes obtained a sales listing from its consignees that listed the location of the customer. Since the sales database identifies

   [[Page 31425]]

   certain transactions as sales to Canada, and since this information reflects the knowledge of the consignee (acting on behalf of the exporter), at the time of sale, the transactions in question are unarguably Canadian sales. Therefore, we have excluded these transactions from the U.S. sales database.

   Comment 18: Unreconciled Claim Adjustments.

   The petitioners contend that, at verification, the Department found that it could not link certain quality claim expenses incurred by the consignee to sales of subject merchandise. According to the petitioners, the Department should not assume that the consignee absorbed the expense of the quality claims, as this would be tantamount to application of ``beneficial facts available.'' The petitioners argue that, instead, the Department should assume that Mares Australes bore the full amount of the quality claim expense, and reduce U.S. price by that amount.

   Mares Australes responds that, while the resellers' books may not permit linkage of specific quality claims to specific sales, all quality claim expenses charged by the consignee to Mares Australes have been captured in the submitted sales database. According to Mares Australes, claim expenses absorbed by the consignee should not be deducted from U.S. price, as they do not affect the net return to the respondent.

   DOC Position: We agree with Mares Australes. At verification, we observed that a number of quality claims were charged by the consignee to Mares Australes. While some of these claims could not be linked to specific transactions due to the nature of the consignees' books, they resulted in an allocated reduction to U.S. price for groupings of sales. Other quality claims were absorbed by the consignee. Such claims are not expenses of the respondent and do not reduce the revenue received by the respondent; rather, they are normal expenses of the consignee, and are covered by the commission charged by the consignee on the sale.

   Sales Issues--Marine Harvest

   Comment 19: Accruals for Rebates.

   The petitioners claim that Marine Harvest did not report certain rebates for co-op advertising accrued on its U.S. expense ledgers during the POI, and failed to provide evidence to support its claim that the co-op advertising program in question was canceled before any rebates were granted. The petitioners request that, as adverse facts available, the Department reduce Marine Harvest's U.S. prices by the highest amount accrued on Marine Harvest's expense ledgers.

   Marine Harvest responds that the co-op advertising program in question never proceeded beyond the ``good idea'' stage, and that no rebates were ever paid. Citing Certain Corrosion-Resistant Carbon Steel Flat Products and Certain Cut-to-Length Carbon Steel Plate From Canada, 63 FR 12725 (March 16, 1998), Marine Harvest argues that the Department's practice is to not adjust prices for such accruals.

   DOC Position: We agree with the petitioners. At verification, we found that Marine Harvest had made accruals for anticipated rebates to be paid to one of its customers during the POI. While we found no evidence that Marine Harvest had paid these rebates to the customer, we observed that Marine Harvest had not reversed these accruals as of the time of verification. Therefore, Marine Harvest's books indicated that the respondent anticipated that such payments would be made.

   The case cited by Marine Harvest involves claims of accrued (but unpaid) rebates for comparison market sales, and not for U.S. sales. In this and other cases involving such claims for adjustments to normal value, the Department has required that the respondent demonstrate that there is evidence of a contractual obligation for the payment of such rebates, or that there is a historical record of such rebates having been paid regularly in the past. Id. at 12740-41; see also Final Determination of Sales at Less Than Fair Value; Gray Portland Cement and Clinker From Japan, 56 FR 12156, 12168 (March 22, 1991); Final Determination of Sales at Less Than Fair Value; Color Television Receivers From Taiwan, 49 FR 7628, 7637 (March 1, 1984). If the Department did not require such evidence, respondents could record accruals on their books for fictitious expenses, artificially reduce normal value, and then reverse the accruals after the antidumping proceeding was ended.

   We do not know of, and the parties have not cited to, any case where the Department has found accrued but unpaid expenses corresponding to U.S. sales, as opposed to comparison market sales. Given the fact that the expense in question involves U.S. sales, we believe that it is incumbent on the respondent to demonstrate that the expense accrued on its books will not result in a rebate payment. At verification, the respondent did not provide any such evidence. The only evidence on the record is the respondent's accrual of these expenses on its books. In view of this, we have reduced U.S. price for the customer in question by the amount of the unreported accrued rebates. Because Marine Harvest has been a cooperative respondent, and with the single exception of this unreported accrued rebate, has been generally very thorough in its reporting of sales and expenses, we have not applied adverse facts available. Instead, we have reduced U.S. price by the rebate amounts actually accrued.

   Comment 20: Level of Trade/CEP Offset for Marine Harvest.

   The petitioners argue that the Department should not make a CEP offset for Marine Harvest's sales in the Japanese market. According to the petitioners, the level of trade in Japan is less advanced than the level of trade of U.S. sales, because Marine Harvest's U.S. sales affiliate engages in a wider variety of sales activities than does Marine Harvest's Japanese sales affiliate. As a secondary point, the petitioners contend that since sales to Japan are made exclusively to trading companies, the Department should find that there are separate levels of trade for U.S. sales involving retailers versus supermarkets/ distributors and make a level-of-trade adjustment for any comparisons of U.S. sales to retailers to Japanese sales.

   Marine Harvest argues that a CEP offset for Japanese sales is appropriate. According to Marine Harvest, the level of trade of sales to Japan is more advanced than the level of trade to the United States, since the sales activities performed by the U.S. reseller correspond to selling expenses already adjusted for as reductions to the CEP, and therefore cannot be considered in the comparison of selling functions performed by the sales affiliates in the two markets. Marine Harvest contends that its Japanese sales affiliate performs significant selling functions.

   Marine Harvest does not address the petitioners' request that the Department find the existence of different levels of trade in the U.S. market and make an LOT adjustment for comparisons of U.S. sales to retailers to Japanese sales.

   DOC Position: We agree with Marine Harvest that a CEP offset is appropriate. In the preliminary determination, we found a single level of trade in the Japanese market and a single level of trade in the U.S. market. We also found that the level of trade of sales to Japan is more advanced than the level of trade to the U.S. See Preliminary Determination at 2670. Verification has borne out that finding. At verification, we found that Marine Harvest's Japanese affiliate is engaged in a variety of selling functions including negotiation of terms of sale, visits to customers, handling of quality claims, and promotion of Marine Harvest's

   [[Page 31426]]

   products. See Marine Harvest Sales Verification Report at 12. To the extent that Marine Harvest's U.S. affiliate performs such functions, the associated expenses have already been adjusted for as reductions to the CEP.‹SUP›9‹/SUP› Therefore, we continue to find that the level of trade of the Japanese market is more advanced than that of the U.S. market.

   \9\ As noted in Comment 9, supra, petitioners claim that the CIT recently overturned the Department's practice of comparing the level of trade of comparison market sales to a constructed level of trade for CEP sales. See Borden et al. v. United States, cited in petitioners' case brief at 83. The Department is still considering the Court's remand order.

   With respect to the petitioners' request that the Department find separate levels of trade in the United States, we note first that petitioners have not offered any reasons for the Department to deviate from its analysis in the preliminary determination. Since (1) the LOT of the Japanese sales is more advanced than the LOT of U.S. sales, (2) there is only one LOT in the Japanese market, (3) Marine Harvest does not sell salmon nor any other product at a different level of trade in Japan, and (4) the data submitted by the other respondents do not permit quantification of differences in level of trade, we find that an LOT adjustment cannot be made. Therefore, we have continued to make a CEP offset.

   Comment 21: Commingling of Different Grades of Salmon.

   According to the petitioners, Marine Harvest has admitted that it commingled premium and super-premium salmon on shipments to the United States. The petitioners argue that, therefore, even if the Department accepts that there is a legitimate distinction between the two grades in the Japanese market, it should nonetheless average Japanese sales prices of premium and super-premium salmon.

   Marine Harvest contends that it is rare that U.S. shipments of premium salmon will contain some super-premium salmon in the mix, and that such sales are in any case properly identified as being of premium grade, since they include only about five percent super-premium salmon.

   DOC Position: As explained above in Comment 1, we have not distinguished between super-premium and premium salmon. Accordingly, this issue is moot.

   Cost Issues--General

   Comment 22: Major Inputs.

   The Association argues that, in its final determination, the Department should not use transfer prices to value transactions between companies and their affiliated processors and feed producers. Instead, the Association suggests that, for Eicosal and Marine Harvest, the Department rely on the affiliated suppliers' costs to value processing services and feed for purposes of computing cost of production and constructed value.

   The Association contends that the so-called ``transactions disregarded'' and ``major input'' rules under sections 773(f) (2) and (3) of the Act do not apply in this instance because the two companies' affiliated suppliers are separate legal entities in form only and that, in substance, these suppliers operate as divisions of a single entity. According to the Association, the record demonstrates that Eicosal and Eicomar, and Marine Harvest and Marifarms/Marine Feeds are more than mere ``affiliated persons'' as defined by section 771(33) of the Act. As evidence of this, the Association points out that Eicosal and Marine Harvest are each part of wholly-owned, commonly controlled, vertically integrated salmon production operations with the same accounting systems and under the same management.

   The Association asserts that the Department has not allowed the legal form of an entity to distort the calculation of dumping margins in other areas of the law. The Association notes that, for instance, in Certain Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products from Korea: Final Results of Antidumping Duty Administrative Reviews, 62 FR 18404, 18430 (April 15, 1997) (Steel Flat Products from Korea) (Comment 19), the Department chose not to impose the major input rule where it treated respondent companies as a single entity for purposes of reporting sales of the subject merchandise. The Association further points to the Department's practice of calculating financial expenses on a consolidated basis in support of its argument that Eicosal and Marine Harvest and their respective affiliated suppliers should be treated as single entities for purposes of valuing inter-company transactions.

   In addition, the Association argues that generally accepted accounting principles suggest that the Department should treat the companies and their affiliated suppliers as single entities. Specifically, the Association notes that U.S. and international financial accounting principles require all companies that hold controlling interests in other companies to consolidate the results of their operations with those of their subsidiaries. This practice, the Association observes, has the effect of treating consolidated companies as a single entity, since all profits and losses on transactions between the companies are eliminated. The Association contends that the respective parent companies of Eicosal and Marine Harvest each follow these accounting principles in the ordinary course of business and prepare consolidated financial statements covering all of their controlled subsidiaries. Thus, the Association argues, the Department should value affiliated-party transactions at cost in the same way they are recorded in the ordinary course of business in the companies' audited, consolidated financial statements.

   With respect to a third salmon producer, Mares Australes, the Association argues that the Department should use a market price instead of the higher transfer price in valuing feed purchases from its affiliated feed producer Trouw Chile, S.A. (Trouw Chile). According to the Association, the relevant provision of the antidumping statute provides for the use of market price to value inputs from affiliated parties ``if, in the case of any element of value required to be considered, the amount representing that element does not fairly reflect the amount usually reflected in sales of merchandise under consideration in the market under consideration.'' See section 773(f)(2) of the Act. Therefore, the Association believes that the statutory provision at issue provides for the use of market price whenever the transfer price does not fairly reflect the amount usually reflected in sales of the subject merchandise. The objective of the affiliated party rule is to ensure that COP is appropriately calculated and not distorted by decisions between affiliated parties as to where to book the profits on the production of the input, suggests the Association.

   The petitioners assert that, in dealing with transactions between affiliated companies under sections 773(f) (2) and (3) of the Act, it is the Department's practice to value major inputs, like processing and feed, at the higher of the transfer price, market price, or actual production cost. Indeed, according to the petitioners, Eicosal and Marine Harvest's argument that the Department may make an exception to its normal practice in the case of ``close affiliates'' is inconsistent with the statutory scheme as drafted by Congress. The petitioners maintain that the Department must reject Eicosal and Marine Harvest's argument to base affiliated-party purchases on cost rather than on the higher transfer price amounts.

   [[Page 31427]]

   The petitioners disagree with the two respondents' reliance on Steel Flat Products from Korea, noting that, unlike Eicosal, Marine Harvest, and their respective affiliates, all of the Korean companies involved in that case produced the subject merchandise and, thus, had been ``collapsed'' by the Department for purposes of reporting sales and computing a single antidumping duty margin. Similarly, the petitioners reject respondents' argument with respect to the Department's practice of computing financial expenses based on consolidated financial statement data. The petitioners observe that, in contrast to debt which is dispersed throughout the consolidated companies, inter-company profit is generated at different points in the production process and by the sales process specific to each product, customer and market. The petitioners also contend that because the Department conducts a two-market price analysis in antidumping cases, some profit must be built into comparison market sales so that respondents do not allocate away all comparison market profit for dumping purposes.

   With respect to respondents' arguments that U.S. and international accounting principles call for treating Eicosal, Marine Harvest and their affiliates as single entities, the petitioners contend that these accounting principles do not in any way outweigh the provisions of the antidumping statute. The petitioners argue that the Department must therefore apply the statutory provisions for ``fair value'' and ``major inputs'' for Eicosal and Marine Harvest in the final determination.

   With regard to the Association's claim that the Department should rely on market prices for Mares Australes, the petitioners assert that this claim is inconsistent with the Department's normal establishment of arm's-length transactions.

   DOC Position: We disagree with the Association with respect to our application of the major input rule for Eicosal, Marine Harvest and Mares Australes. In order to value processing services and feed purchased by these companies from their affiliated suppliers, we have continued to rely on the higher of transfer prices, market value, or the affiliate's cost of production in accordance with sections 773(f)(2) and (3) of the Act.

   As noted in the comments from both respondents and the petitioners, section 773(f)(2) and (3) of the Act prescribes how the Department is to treat affiliated-party transactions in its calculation of cost of production and constructed value. With respect to major inputs purchased from affiliated suppliers (in this instance, salmon processing and feed), the Department's practice is that such inputs will normally be valued at the higher of the affiliated party's transfer price, the market price of the inputs, or the actual costs incurred by the affiliated supplier in producing the inputs.

   Since implementation of the URAA, the Department has consistently applied this interpretation (see, e.g., Small Diameter Circular Seamless Carbon and Alloy Steel Standard, Line and Pressure Pipe From Germany: Final Results of Antidumping Duty Administrative Review, 63 FR 13217, 13218 (March 18, 1998)(Comment 1), and Silicomanganese from Brazil; Final Results of Antidumping Duty Administrative Review, 62 FR 37869, 37871 (July 15, 1997) (Silicomanganese from Brazil)(Comment 3)), making exception in only those cases wherein it treats respondents as a single entity for purposes of sales reporting and calculating an antidumping margin (see, e.g., Steel Flat Products from Korea (Comment 19)). Relying solely on cost in the latter case flows logically from the overall calculation methodology being employed.

   All of the parties in question are separate legal entities in Chile, responsible for maintaining their own books and records. In contrast to Steel Flat Products from Korea, the Department is applying its normal company-specific calculation methodology. Therefore, there is no basis for establishing an exception to the ``major input rule.'' Accordingly, sections 773(f)(2) and (3) of the Act apply to the transactions between these companies.

   Further, we disagree with respondents' argument that the principles that guide the Department to treat groups of affiliated companies as a single entity for purposes of calculating financial expenses should apply to other elements of cost of production. The Department's practice regarding the calculation of financial expenses based on the consolidated financial statements of the parent company is well established and has been upheld by the courts. See, e.g., E.I. DuPont de Nemours & Company v. United States, Slip Op. 98-7, Court No. 96-11- 02509 (January 29, 1998)(upholding the Department's application of its long-standing policy of calculating interest expense from the borrowing cost incurred by the consolidated group of companies rather than the individual producer). The Department's practice with respect to calculating financial expenses is for a completely different purpose, i.e., to ensure that consolidated companies do not direct actual interest costs away from producers of subject merchandise and to producers of non-subject merchandise. On the other hand, under the major input rule, the statute requires that we review affiliated-party purchases in order to determine that they reasonably reflect a fair value.

   Although generally accepted accounting principles usually require that a company's financial statements be consolidated with all companies in which it owns a controlling interest, these consolidated financial statements do not alter the manufacturing costs associated with producing the subject merchandise as recorded by the entity producing the subject merchandise.

   Consistent with our general practice, outlined above, we disagree with Mares Australes that a market price rather than the transfer price it pays its affiliate should be used to value feed purchases from Trouw Chile. The Department will use the transfer price which normally reflects Mares Australes' purchases of the input, unless the transfer price does not reflect a fair value in the market under consideration. Therefore, we continue to rely on transfer prices in order to value feed purchased from Mares Australes' affiliated supplier, Trouw Chile.

   Comment 23: Perishability.

   The Association argues that the Department erroneously determined in the preliminary determination that salmon was not a highly perishable agricultural product for purposes of determining ``substantial quantities'' of sales below cost in the cost test. The Association contends that the test for ``high perishability'' is whether a product has a short shelf life, noting that the Department has found products with significantly longer shelf lives than salmon, ‹SUP›10‹/SUP› to be highly perishable. According to the Association, the petitioners themselves have attested to the high perishability of salmon before the International Trade Commission (ITC).

   \10\ Although it did not make this specific point in its case briefs, at the public hearing the Association referenced a determination involving a 1983 Department finding that potatoes from Canada are highly perishable. The Association noted that salmon have much shorter shelf lives than potatoes. See Transcript of Case Hearing at 59-61 (April 28, 1998).

   Further, although the Association acknowledges that the Department did not find salmon to be highly perishable in the LTFV investigation of Fresh and Chilled Atlantic Salmon from Norway (Salmon from Norway), it contends that

   [[Page 31428]]

   that precedent is not controlling. According to the Association, Norwegian producers and exporters of salmon were different entities, and the Department's focus in that case was whether live farmed salmon was highly perishable for producers (who sold that salmon to exporters). The Association argues that the respondents in this case are integrated producers/exporters, such that the Department is not examining any sales of live salmon as sold by producers; rather, the merchandise in question consists entirely of dressed fish sold by the producer/exporter. Therefore, the Association contends, any alleged control over harvest timing is irrelevant, since once salmon are dressed and/or filleted, they become inherently perishable.

   Finally, the Association claims that the sales data submitted in this investigation indicate that salmon prices fall significantly due to inevitable perishability problems after harvesting. As evidence, the Association submits a graphical illustration of U.S. and Canadian price trends over the shelf life of salmon, based on data submitted by Aguas Claras in its sales databases.

   The petitioners argue that salmon should not be considered a highly perishable agricultural product for purposes of the cost test. According to the petitioners, the Department's precedent established in Salmon from Norway (i.e., that salmon is not a highly perishable product) is controlling in the instant investigation. The petitioners disagree with the Association's claim that, due to the integration of producers and exporters in the Chilean salmon industry, Salmon from Norway is inapplicable. According to the petitioners, that high degree of integration in the Chilean salmon industry enhances the respondents' control over harvesting and distribution schedules.

   More generally, the petitioners contend that a product can only be deemed to be highly perishable if the producer has very little flexibility in controlling the timing of harvesting, and if this lack of control normally and inevitably results in sales below cost for the industry. According to the petitioners, salmon harvests can be delayed by as many as 15 months, such that the respondents can fine-tune harvest timing so as to avoid the need to make sales below cost.

   The petitioners further argue that verification revealed that sales below cost are not an inevitable aspect of salmon production, and that Chilean salmon producers have not demonstrated that they suffer from perishability problems in bringing their product to market.

   DOC Position: We do not disagree with the Association's statement that, once harvested, salmon is a perishable product that does not have a long shelf life. However, the issue with respect to the ``substantial quantities'' portion of the cost test is whether salmon is a product that the respondents can expect to sell routinely in the comparison market at prices below the cost of production due to the highly perishable nature of the product. We disagree with the Association's contentions in this regard and find that fresh Atlantic salmon is not a highly perishable agricultural product for purposes of the ``substantial quantities'' test.

   In Salmon from Norway, the Department found that the respondents had sufficient control over harvest timing and distribution such that perishability was not a concern, as the salmon were brought to market before freshness was compromised. Although the Association contends that the Department's focus in that case was on live salmon as sold by producers to exporters, the Department in fact found that salmon was not highly perishable either with respect to producers or exporters, whether live or harvested. The Department concluded:

   Norwegian salmon farmers have the ability to control the time of sale of their output by ``holding over'' inventory and, since January 1990, by freezing fresh salmon. Regarding respondents' assertion that salmon is perishable in the hands of the exporters, the Department found at verification that the opposite is true. Exporters coordinate their salmon requirements in weekly telephone conferences with their customers, with farmers, and with other exporters. By doing so, exporters can communicate their salmon requirements two weeks into the future so that farmers can begin to ``starve'' (prepare for harvest) the salmon two weeks prior to harvest. Accordingly, there appears to be no perishability problem at the exporter level.

   See Salmon from Norway at 7673.

   The record of the instant investigation, including our findings at verification, suggests that perishability is even less of a problem for the Chilean respondents than for the Norwegian respondents. The Chilean respondents are integrated producers/exporters, so that their production and harvesting schedules are more easily coordinated. Moreover, the respondents sell to a small number of importers in their respective comparison markets, with whom they closely coordinate both production and distribution. Shipments to third-country markets are made directly to the customer, without the involvement of consignees or affiliated resellers.‹SUP›11‹/SUP› As the salmon are shipped, the terms of the sale are set, and the sale is consummated. Therefore, perishability does not become a factor in the respondents' pricing.

   \11\ The single exception is Aguas Claras, which made sales to Canada out of its U.S. affiliate's inventory. However, at verification Aguas Claras asserted that it sells merchandise affected by perishability problems in the United States and not in Canada due to the longer transportation times required for Canadian sales. See Aguas Claras Sales Verification Report at 6. Thus, to the extent that Aguas Claras makes significant sales below cost in the Canadian market, it is for reasons other than perishability.

   Our verifications bear out these findings. For instance, Marine Harvest sells to a total of three customers in Japan, and the majority of sales are made to a single customer. According to company officials, because Marine Harvest Chile's sales to Japan are arranged in close consultation with Japanese customers, it is exceptionally rare for Marine Harvest Chile to make sales below cost to the Japanese market due to perishability concerns. See Marine Harvest Sales Verification Report at 4-5. The other respondents similarly are able to coordinate closely their shipments with their customers. In the case of Eicosal, its Japanese customers reportedly will purchase all the high-quality salmon that Eicosal can produce. See letter from Eicosal to the Department of Commerce, transmitting Supplemental Section A Questionnaire Response (November 18, 1997), at 3. Moreover, in describing its production and sales process at verification, Eicosal stated that it conducts negotiations for Japanese sales before the salmon are harvested. See Eicosal Sales Verification Report at 7. Similarly, Mares Australes has stated that its two Japanese importers inform them of their requirements a month in advance, and that one of its importers even provides ``exact requirements by shipment.'' See letter from Mares Australes to the Department of Commerce, transmitting Supplemental Section A & B Questionnaire Responses (November 3, 1997), at 12.

   As for the Association's argument that the Department has found products with longer shelf lives than salmon (such as potatoes) to be highly perishable, we note that shelf life is not the sole criterion in determining whether an agricultural product is highly perishable for purposes of the cost test. Rather, as explained above, the issue is whether salmon is a highly perishable product that the respondents can expect to routinely sell in the comparison market at prices below the cost of production.‹SUP›12‹/SUP›

   [[Page 31429]]

   Given the facts of this case, we have found that fresh Atlantic salmon does not meet that standard.

   \12\ With respect to the Association's reference to the Department's finding that potatoes (which have longer shelf lives than salmon) are a perishable product, we note that the underlying case dates back sixteen years, and the notice of final determination in that case does not set forth any details of the Department's analysis of perishability with respect to potatoes. See Final Determination of Sales at Less Than Fair Value; Fall-Harvested Round White Potatoes From Canada, 48 FR 51669, 51669 (November 10, 1983). In any event, there is no bright line ``shelf-life'' test to define high perishability, and the determination of whether a product is highly perishable for purposes of the cost test is necessarily specific to the facts of each case.

   In view of the record evidence that salmon is not a highly perishable product for purposes of the cost test, we do not find any basis to warrant the application of a higher threshold for the ``substantial quantities'' aspect of the cost test.

   Comment 24: Exchange Rate Losses.

   The Association argues that, in calculating financial expenses for COP and CV, the Department must include only those exchange rate losses that are attributable to loans used to finance salmon production during the POI. While it acknowledges the Department's normal practice of calculating general expenses, including financial expenses, based on each respondent's fiscal year data, the Association maintains that, in this case, such a practice would overstate the actual financial expenses incurred by the salmon producers due to the effects of exchange rate losses incurred during 1996. Specifically, the Association points to the fact that a shift in the Chilean peso/U.S. dollar exchange rate during the first part of 1996 was responsible for the major portion of the exchange losses incurred by the producers in connection with their dollar-denominated debt. These losses, adds the Association, were reported by the salmon producers in their 1996 financial statements, the same financial statements used by the Department to compute financial expenses for COP and CV. The Association notes, however, that during the actual months of the POI, the change in the peso/dollar exchange rate was significantly less than that of the full calendar year 1996. Thus, according to the Association, where the Department determines to include exchange rate losses in financial expenses, it should compute such losses based on the actual POI and not the company's 1996 fiscal year, in effect, limiting its analysis of exchange rate gains and losses to the POI so as to match these costs to sales during the POI.

   As support for its position, the Association argues that exchange rate gains and losses differ from other types of G&A expenses and interest expense in that the former may fluctuate significantly from month to month, causing considerable changes in the amount of gain or loss recognized as a cost. Moreover, according to the Association, the Department has acknowledged the distortion caused by exchange losses and its practice of calculating financial expenses based on full-year financial statement information. As evidence of this, the Association points to the Final Determination of Sales at Less Than Fair Value Oil Country Tubular Goods from Mexico, 60 FR 33567, 33572 (June 28, 1995) (OCTG from Mexico) in which the Department chose not to use financial statement data to compute financial expenses because devaluation of the Mexican peso made the information unrepresentative of costs during the POI.

   In addition to considering only the exchange losses incurred during the POI, the Association also urges the Department to exclude from COP and CV a portion of the losses on loans allocable to financing sales and accounts receivable. The Association argues that because the companies finance all of their operations, including both production and sales activities, part of the exchange loss arising from dollar- denominated debt must be attributed to the companies' non-production activities. If the Department chooses not to allocate a portion of the exchange loss to sales activities and accounts receivable, the Association contends that it should reexamine its treatment of exchange gains arising from foreign currency receivables by treating all such gains as an offset to foreign exchange losses.

   The petitioners argue that the Department must continue to calculate financial expenses based on the salmon producers' 1996 financial statement data, and not use the POI data as suggested by the Association. According to the petitioners, consistent with the Department's practice, the fiscal year information provides the most accurate and reasonable basis for estimating the actual expenses incurred, including exchange gains and losses. The petitioners point also to Gray Portland Cement and Clinker from Mexico: Final Results of Antidumping Duty Administrative Review, 62 FR 17148, 17160 (April 9, 1997), in which the Department determined that exchange gains and losses arising from the respondent's foreign currency debt were, indeed, related to production and therefore properly included in the calculation of financing expenses. Lastly, the petitioners call attention to the fact that the Department's practice of including foreign exchange gains and losses in financial expenses has been upheld by the CIT in Micron Technology, Inc v. United States, 893 F. Supp. 21 (CIT 1995).

   DOC Position: Our practice is to calculate general expenses, including financial expenses, based on the full fiscal year's information that most closely corresponds to the period of investigation or review. See, e.g., Final Results of Antidumping Duty Administrative Review: Silicon Metal From Brazil, 63 FR 6899, 6906 (February 11, 1998) (Comment 16). Contrary to the Association's claim, general expenses often vary greatly from month to month. By considering general expense information for the fiscal year, however, the Department is able to ensure that it has reasonably captured all of the expenses associated with the respondent's complete business and accounting cycle. In particular, we note that the year-end financial statement data are generally the most accurate reflection of a company's results because these data include complete year-end accruals and other adjusting entries that are often posted only at year-end. In addition, the year-end statements are often audited, or at a minimum, reviewed by outside accountants, which provides additional assurance as to the accuracy of the data presented and the accounting principles used to compile those data.

   Here, the Association suggests that the Department isolate one specific expense, foreign exchange losses, which it contends would be lower if the Department departs from its normal methodology and shifts the calculation period for foreign exchange losses on loans by three months. While that may be the case, it is difficult to accept the Association's rationale in light of the fact that they have offered no information as to the effect that the three-month shift would have on all other costs incurred by the companies, certain of which may indeed be higher than those of the 1996 fiscal year. Thus, we do not consider it appropriate for the Department to abandon its normal practice for a single expense (foreign exchange losses) when the rationale for doing so is little more than the fact that such expense would be lower if calculated over a different period.

   With respect to the Association's reliance on OCTG from Mexico as a departure from the Department's general practice of using fiscal year data, we note that, in that case, the respondent's financial expense ratio was based on best information available (the predecessor to facts available).

   [[Page 31430]]

   Specifically, the investigation in that case encompassed a six-month period from January through June 1994. The respondent's 1994 financial statements were provided by the petitioners, after the respondent claimed that these statements were not available. The financial statements showed the effects of the massive devaluation of the Mexican peso sustained in late December of 1994, several months subsequent to the POI. As discussed more fully in OCTG from Mexico, the Department used an adverse inference in its calculation of interest expense, while declining to include the full amount of the peso collapse. While the Association has characterized the change in the Chilean peso rate during the fiscal year as ``four and one-half times'' that of the POI, this reflects a change of from 1 to 4.4 percent. This change does not begin to equate to the massive currency devaluation noted in OCTG from Mexico. Finally, we note that the choice of adverse facts available (or its predecessor best information available) provides no guidance with respect to the Department's preferred methods for calculating actual expenses.

   As to the Association's assertion that exchange losses should be attributed to the accounts receivable balance, this is inconsistent with our practice. The Department has an established practice of including currency translation gains and losses on foreign-currency denominated loans in COP and CV because they reflect an actual increase in the amount of local currency that will have to be paid to retire the foreign-currency denominated loan balances. See, e.g., SRAMs from Korea (Comment 4). We allocate the financial expenses based on the cost of goods sold and, thus, these expenses are reflected as a cost of production, and not a selling expense. We do not consider exchange gains and losses from sales transactions to be related to the manufacturing activities of the company. See, e.g., Notice of Final Determination of Sales at Less Than Fair Value: Steel Wire Rod From Trinidad and Tobago, 63 FR 9177, 9181 (February 24, 1998) (Comment 4).

   For this final determination, we have included in the cost of production the amortized portion of foreign exchange losses resulting from foreign-currency denominated loans as part of the financial expenses. The foreign exchange losses on loans reported in the consolidated financial statements were amortized over the average remaining life of the loans on a straight-line basis.

   Comment 25: CV Imputed Credit.

   The Association argues that the Department's methodology for comparing U.S. prices to CV does not properly account for imputed credit expenses in the comparison market. The Association believes that the Department should either deduct an amount for imputed credit from CV, as it has done in recent cases, or should exclude from COP financial expenses the amount allocable to financing accounts receivable, as it did under the old law.

   Further, for Camanchaca, the only producer that did not have a comparison market, the Association argues that, if the Department continues to use the weighted-average selling and profit rates of the other four respondents in this investigation, the Department should apply the weighted-average comparison market imputed credit of the other four producers.

   The petitioners do not rebut the Association's comments on this issue.

   DOC Position: We agree with the Association that a ``circumstance of sale'' adjustment for imputed credit should be made to CV. The Department ``uses imputed credit expenses to measure the effect of specific respondent selling practices in the United States and the comparison market.'' See Stainless Steel Wire Rods from France (Comment 5). Thus, in order to make a fair comparison, we have deducted imputed credit from CV as a COS adjustment in this final determination.

   Comment 26: Allocation of Financial Expenses Based on Assets.

   The Association asks the Department to consider the special circumstances of three salmon producers--Eicosal, Camanchaca, and Aguas Claras--in calculating financial expenses for COP and CV. According to the Association, certain characteristics unique to these companies' operations require that the Department modify its normal method of computing consolidated financial expenses based on the ratio of net financial expenses to cost of goods sold during the period.

   In the case of Eicosal, the Association contends that the Department must recognize the very different capital requirements of-- and disproportionate generation of financial expenses by--Eicosal and its affiliated processor, Eicomar. That is, in the Association's view, the Department must allocate consolidated financial expenses between Eicosal and Eicomar based on the relative value of fixed assets held by each company. The Association maintains that this allocation is necessary in order to avoid significant distortions in the calculation of financial expenses due to the fact that Eicomar, as a seafood processor, requires substantially greater amounts of capital for equipment than does Eicosal, which conducts the salmon farming operations. In support of this view, the Association cites the Final Determination of Sales at Less Than Fair Value of Dynamic Random Access Memory Semiconductors of One Megabit and Above from the Republic of Korea, 58 FR 15467, 15471 (March 23, 1993)(DRAMS from Korea) where, before calculating a respondent's net financial expense ratio for COP and CV, the Department first allocated financial expenses to various divisions within the corporation based on the relative value of fixed assets within each division.

   The Association also requests that the Department make a fixed asset-based allocation of financial expenses for Camanchaca as well. In this instance, the Association points out that Camanchaca is involved in many fish and seafood-related operations other than the production of fresh Atlantic salmon. According to the Association, Camanchaca's operations are divided into six distinct production areas, each locally administered and having its own capital requirements. The Association maintains that unless financial expenses are first allocated to Camanchaca's production area on the basis of fixed asset value, the Department's normal method of computing such expenses will significantly distort the actual capital costs incurred by the company's salmon production operations.

   Finally, in the case of Aguas Claras, the Association argues that the Department's financial expense calculation fails to take account of the company's frozen and smoked salmon operations. Specifically, the Association observes that, in addition to fresh salmon, Aguas Claras produces and holds in inventory a large amount of frozen and smoked salmon products. According to the Association, before it can accurately capture the financial expenses of fresh Atlantic salmon, the Department must first allocate a portion of total financial expense to frozen and smoked salmon in recognition of the costs incurred to finance these products in inventory. The Association contends that such an allocation would be consistent with the Department's imputation of inventory carrying costs in antidumping cases.

   The petitioners argue that the Department should follow its normal methodology and calculate financial expenses as a ratio of each company's cost of goods sold. According to the petitioners there is no reason in this case for the Department to allocate interest on the basis of inventory or fixed assets as suggested by the Association. The petitioners further

   [[Page 31431]]

   point out that Camanchaca and Aguas Claras improperly reduced their submitted financial expenses associated with the imputed cost of carrying their accounts receivable and ending inventory.

   DOC Position: We disagree with the Association that the facts of the case require us to depart from our general practice of calculating financial expenses based on a ratio of the foreign producer's net expenses to its cost of goods sold. In this case, each of the three respondents proposes alternative methods for calculating financial expenses which they believe best represent the unique circumstances of their operations. In effect, these calculations allocate interest charges to certain assets which the companies contend are not associated with subject merchandise, and thus, have the effect of lowering the interest expense for subject merchandise. The fact that the results of these calculations differ from the normal cost-of-sales- based calculation does not in any way suggest that the Department's longstanding practice of calculating financial expenses is inaccurate or unreasonable. In fact, the Courts have upheld as reasonable the Department's practice of calculating financial expenses based on the consolidated group as a whole, notwithstanding the fact that any non- respondent member of the Group may have been involved in a different line of business or held assets having values substantially different from those of the respondent company. See, e.g., E.I. DuPont de Nemours & Co. v. United States, Slip op. 98-7, Court No. 96-11-02509 (January 29, 1998)(where the Court noted that the Department's calculation of financial expenses reasonably reflects the actual costs incurred by the respondent) and Gulf States Tube Division v. United States, Slip op. 97-124, Court No. 95-09-01125 (August 29, 1997) at 31 (where in light of the fact that the statute provides no specific guidance for the calculation of financial expenses, the Court recognized as reasonable the Department's allocation of such expenses based on the respondent's consolidated group).

   With respect to the Association's citation to DRAMS from Korea, we note that while the Department relied on an asset-based allocation methodology in the investigation phase of that case, we have since reconsidered this approach. Specifically, although the CIT upheld the Department's interest calculation in that proceeding (Micron Technologies, Inc. v. United States), in a recent investigation involving the same respondent companies from the DRAMS from Korea proceeding, Final Determination of Sales at Less Than Fair Value: Static Random Access Memory Semiconductors From the Republic of Korea, 63 FR 8934, 8938 (February 23, 1998)(SRAMS from Korea), the Department described why it was unnecessary to follow the fixed asset based allocation methodology for financial expenses that had been used in the DRAMS from Korea proceeding. See SRAMS from Korea at 8938 (General Comment 2). (``We have reconsidered this issue for the final determination and concluded that because the COGS includes a proportional amount of the depreciation of the assets used in the production of the merchandise, allocation of financing expenses on the basis of COGS distributed proportionately more interest expense to those products having higher capital investment.'') Thus, as in this case, the Department recognized that its normal method of calculating financial expenses on the basis of cost of goods sold, without special allocations to specific divisions or assets, provides a reasonable measure of the costs incurred for the merchandise.

   Further, we have not allowed the respondents to offset financial expenses for the claimed cost of holding accounts receivable and inventory. The statute directs the Department to calculate selling, general and administrative costs, including financial expense, based upon the actual experience of the company. See section 773(b)(3)(B) and section 773(e)(2)(A) of the Act. Under the pre-URAA law, we allowed offsets to financial expense for accounts receivable and finished goods inventory to account for the fact that we calculated CV inclusive of amounts imputed for credit and inventory carrying costs. Consistent with the provisions of the new law, however, we now base financial expense for COP and CV on the amounts incurred by the respondents, and do not account for imputed expenses as actual costs for the calculation of CV. Therefore, it is no longer appropriate to reduce the financial expenses by the accounts receivable and inventory offsets as suggested by the Association. See, e.g., Steel Flat Products From Korea at 18422 (Comment 6); Final Determination of Sales at Less Than Fair Value: Certain Pasta From Italy, 61 FR 30326, 30361 (June 14, 1996).

   Comment 27: Inflation.

   The Association contends that the Department should not adjust the respondents' reported cost of production and constructed value figures to account for the effects of Chilean inflation on salmon stock costs. Although it recognizes that such an adjustment would be consistent with Chilean accounting principles, the Association points out that inflation in the country ranged only between six and eight percent during the period over which the respondents calculated their reported salmon costs. This low inflation rate, argues the Association, does not meet the Department's normal threshold for adjusting costs in cases involving significant inflation.

   In support of its position, the Association cites Certain Fresh Cut Flowers from Colombia: Final Results of Antidumping Duty Administrative Reviews, 61 FR 42833, 42845 (August 19, 1996)(Flowers from Colombia) and Final Determination of Sales at Less Than Fair Value: Fresh Cut Roses from Colombia, 60 FR 6980, 6993 (February 6, 1995)(Roses from Colombia), where it contends that the Department's policy is to adjust costs to a constant currency basis only in cases involving high- inflation and, even then, only to adjust expenses related to long-lived fixed assets (i.e., depreciation expense). The Association notes that, consistent with Chilean GAAP, each respondent restated the historical cost of its fixed assets such that the depreciation expense reported for cost of production and constructed value reflected current Chilean peso values during the period of investigation. However, the Association contends that salmon stock is not a fixed asset and, thus, it is inconsistent with past Department practice to also adjust these costs for the low inflation experienced in Chile during the cost calculation period.

   The petitioners, citing Final Determination of Sales at Less Than Fair Value: Canned Pineapple Fruit from Thailand, 60 FR 29553, 29559 (June 5, 1995) (Pineapple from Thailand), claim that the Department should rely on the respondents' normal books and records, kept in accordance with Chilean GAAP, for the calculation of the live fish inventory cost. The petitioners argue that whether inflation in Chile was high or low is irrelevant to the cost calculation because the Department must first look at the respondents' home country GAAP to determine whether such principles reasonably reflect the costs of producing the subject merchandise. In Pineapple from Thailand, the Department stated that normal accounting practices provide an objective standard by which to measure costs, while providing the respondents a predictable basis on which to compute costs. The petitioners further contend that, in this case, the respondents want the Department to reject outright the Chilean GAAP requirements regarding price-level

   [[Page 31432]]

   adjustments to non-monetary assets. Yet, the petitioners note, the respondents have failed to meet their burden of demonstrating that such an adjustment would distort the reported costs. The petitioners assert that the respondents have failed to indicate how their normal books and records, kept in accordance with Chilean GAAP, distort costs. The petitioners argue that the respondents' claim that the cost of live fish inventory are mainly contained within the POI is incorrect because the production cycle of salmon is between two and three years.

   DOC Position: We agree with the petitioners that certain of the salmon producers failed to provide costs which reflected their normal accounting practices of adjusting non-monetary assets for increases in price-levels. The exclusion of these adjustments results in costs which are not reflective of current price levels and, thus, produces an improper match of revenues and expenses.

   The Department's long-standing practice, codified at section 773(f)(1)(A) of the Act, is to rely on data from a respondent's normal books and records where those records are prepared in accordance with home country GAAP and reasonably reflect the costs of producing the merchandise. Normal GAAP accounting practices provide both respondents and the Department a reasonably objective and predictable basis by which to compute costs for the merchandise under investigation. However, in those instances where it is determined that a company's normal accounting practices result in a misallocation of production costs, the Department will adjust the respondent's costs or use alternative calculation methodologies that more accurately capture the actual costs incurred to produce the merchandise. See, e.g., Final Determination of Sales at Less Than Fair Value: New Minivans from Japan, 57 FR 21937, 21952 (May 26, 1992) (Minivans from Japan) (the Department adjusted a respondent's U.S. further manufacturing costs because the company's normal accounting methodology did not result in an accurate measure of production costs); see also, Pineapple from Thailand, 60 FR at 29559.

   In the instant proceeding, the Association asks the Department to reject each salmon producer's normal price-level accounting methodologies used for live fish inventories in favor of costs calculated for purposes of this investigation. As noted, however, the Department's practice is to rely on a respondent's books and records prepared in accordance with its home country GAAP unless these accounting principles do not reasonably reflect costs associated with production of the subject merchandise. As a result, before analyzing any alternative accounting method reported by a respondent during the proceeding, the Department will determine whether it is appropriate to use the respondent's normal GAAP accounting practices in order to calculate the cost of the merchandise.

   In this case, the Department examined whether it was reasonable under Chilean GAAP for the salmon producers to adjust their fish inventory costs to reflect current Chilean peso values corrected for the effects of inflation. Fish stock costs are recorded on the basis of the historical amounts incurred to raise the salmon from eggs to maturity. Similar to fixed assets, however, because fish stock costs are carried on the company books as an asset for two to three years prior to harvest, Chilean GAAP requires that the costs be restated to reflect inflation-adjusted amounts. In examining the companies' books and records at verification, we found that Camanchaca, Aguas Claras and Eicosal had used the recorded price-level adjustment methodology for live fish inventories for at least a number of years. In addition, evidence on the record, i.e., audited financial statements, indicated that each of the three companies' normal price-level adjustment methodologies was accepted by its independent auditors and was consistent with GAAP practiced in Chile.

   Given the fact that the companies' price-level adjustment methodology is consistent with Chilean GAAP and the Association has not shown this practice to distort salmon production costs during the period, we have recalculated each company's fish stock costs to include the price-level adjustment reported in accordance with its normal accounting practices.

   We also found that two of the companies, Mares Australes and Marine Harvest, did not record the price-level adjustment to fish stock costs as they do not prepare financial statements in accordance with Chilean GAAP. Specifically, these companies are subsidiaries of foreign companies that prepare only consolidated financial statements in other countries following accounting principles dictated by the home country GAAP of their respective parent companies. Thus, Mares Australes and Marine Harvest are not required to prepare financial statements in accordance with Chilean GAAP.

   We note that in this case, however, the information provided by Marine Harvest does, in effect, consider the change in the value of the Chilean peso. Marine Harvest's financial data is restated into U.S. dollars monthly as part of its reporting for consolidation purposes. We note that during the cost calculation period the Chilean peso/U.S. dollar exchange rate reflected much of the inflation rate experienced in Chile. Thus, Marine Harvest's reported costs were effectively adjusted for the price-level changes each month, as part of the company's normal accounting.

   With respect to Mares Australes, the case record does not contain information regarding the company's accounting consolidation process with its parent. As part of the consolidation process, however, Mares Australes would have to convert its peso accounting records to the currency in which its parent maintains its normal books and records. Thus, as with Marine Harvest, it is reasonable to conclude that Mares Australes, in effect, accounts for the price-level changes through the currency conversion process of its normal accounting consolidation. Yet, because Mares Australes reported its salmon production costs in pesos for purposes of this investigation, it is necessary for us to reflect the price level changes that are consistent with its currency conversion and consolidation. Accordingly, we have revised Mares Australes' submitted COP and CV figures to reflect price level adjustments based on the inflation index.

   The Association has argued that the salmon producers' normal price- level adjustment methodologies do not reasonably reflect costs due to the low rate of inflation in Chile during the growing period for fresh Atlantic salmon harvested during the POI. Yet, the fact that the level of inflation during the years prior to the POI was not at levels experienced in Chile in the past does not make the price-level adjustment requirements under Chilean GAAP unreasonable.

   Further, the Association's claim that the Department's high- inflation methodology (as stated in Flowers from Colombia and Roses from Colombia) which only requires price-level adjustments for depreciable assets is unfounded. In the specific facts present in those cases, the only restated non-monetary assets which affected the COP and CV were fixed assets, including the flower and rose plants. In this case, as well as in Flowers from Colombia and Roses from Colombia, the costs of the subject merchandise, which were accumulated over years prior to the period of investigation or review, were adjusted for the price-level changes recorded in the company's normal accounting records. Contrary to the

   [[Page 31433]]

   Association's claim, our treatment of the price-level adjustments for the live fish inventory in this case is consistent with our treatment of similar costs in Flowers from Colombia and Roses from Colombia.

   Comment 28: CV Profit for Japanese Market.

   The Association argues that the Department should not base CV profit on sales to the Japanese market without making an appropriate adjustment for differences in the grades sold in the U.S. and Japanese markets. According to the Association, the Department has recognized that there are physical differences between the premium-grade salmon sold in the United States and the super-premium salmon sold in Japan, and has found that it is inappropriate to make price-to-price comparisons of those sales. The Association contends that calculating CV profit based on sales of Japan (which are primarily of super-premium salmon) effectively results in a CV equivalent to the sales price of super-premium salmon in Japan. The Association argues that the use of such a NV would result in an unfair comparison, would be contrary to other case precedent, and would be inconsistent with the Department's stated recognition that price-to-price comparisons of premium to super- premium merchandise are inappropriate.

   The Association proposes that, for Mares Australes (which sold both premium and super-premium salmon in Japan), the Department base CV profit only on sales of premium salmon to Japan. For the other two respondents for whom Japan is the comparison market (and who did not make any sales of premium salmon to Japan), the Association proposes an adjustment based on the percentage difference between Mares Australes' profit rates from sales of the two grades of salmon in Japan.

   Alternatively, the Association proposes that the Department make price-to-price comparisons between premium and super-premium prices with a value-based difference-in-merchandise adjustment, based on the percentage difference between Mares Australes' sales prices for premium and super-premium prices in Japan.

   The petitioners argue that the statute requires that CV profit be based on all sales of the foreign like product made in the ordinary course of trade in the comparison market. According to the petitioners, the statute grants the Department the authority to rely on alternative methods only when such data are unavailable.

   DOC Position: This issue has been rendered moot by the Department's finding, set forth above in response to Comment 1, that there is no significant distinction between premium and super-premium grade salmon for purposes of an antidumping analysis.

   Cost Issues--Eicosal

   Comment 29: Company-Wide G&A.

   The petitioners argue that the Department must recalculate Eicosal's G&A expenses to reflect amounts reported in the company's consolidated financial statements. According to the petitioners, such a calculation would be consistent with the Department's practice of computing G&A expenses of the respondent company as a whole, and not just for those expenses directly related to the manufacture of the product under investigation.

   Eicosal claims that the Department should rely on the submitted G&A rate calculation.

   DOC Position: We agree with the petitioners' assertions that the Department's normal methodology is to calculate G&A based on the producing company as a whole and not just based on G&A expenses related to the production of a particular product. We do not agree, however, that this means that the G&A expenses should be based on amounts reported in the respondent company's consolidated financial statements, as the Department's normal methodology does not rely on consolidated level G&A expense. Thus, we did not calculate Eicosal's G&A rate using the consolidated company financial statements.

   Cost Issues--Mares Australes

   Comment 30: Combined G&A.

   Mares Australes contends that it correctly computed its G&A expenses by combining the expenses of Mares Australes and those of its affiliate, Trouw Chile. According to Mares Australes, the two companies are completely integrated and share common management and administrative operations. Thus, Mares Australes argues, in order to accurately capture the G&A expenses incurred on sales of fresh Atlantic salmon, the Department must compute G&A expenses as if Mares Australes and Trouw Chile were a single integrated business unit.

   The petitioners argue that the Department should recalculate Mares Australes' G&A expenses excluding the G&A expenses of Trouw Chile. According to the petitioners, the Department's general practice is to use the G&A expenses that relate to the operations of the producer (Mares Australes) supplemented, but not commingled, with a portion of G&A expenses from the parent company. Further, the petitioners contend that Mares Australes has reported, in effect, not the G&A expenses incurred to produce salmon, but a G&A ratio which represents the results of a combined fish feed and salmon producer. The petitioners also argue that to the degree it is appropriate for Mares Australes to report feed costs based on the actual costs of its affiliate Trouw Chile, Trouw Chile's actual G&A expenses should be included in determining the COP of feed and its G&A expenses should not be mixed with those of Mares Australes.

   DOC Position: We disagree with Mares Australes regarding the appropriateness of its submitted G&A expense calculation. It is the Department's practice to use the G&A expenses calculated based on information from the producer. See, e.g., OCTG from Mexico at 33573 (Comment 8). Trouw Chile's G&A expenses relate to its cost of producing fish feed, and do not bear upon the general expenses incurred by Mares Australes in producing salmon. For this final determination, we calculated G&A expenses for Mares Australes using amounts recorded in the company's normal books and records, and excluded the submitted information of Trouw Chile.

   Comment 31: Bonus Adjustment.

   Mares Australes argues that the Department should allow its adjustment to its reported labor costs so that they reflect only the cost of bonuses actually paid to employees rather than the amount accrued. Because it accrued a greater expense for employee bonuses than was actually paid out during 1996 and the excess accrual was not reversed at year-end, Mares Australes believes it should be permitted to base the expense on only the cash actually paid for bonuses. Mares Australes further argues that in order to match costs incurred during the POI with sales during the POI, the Department should include in COP only the company's ``actual'' bonus expense.

   The petitioners argue that the Department should disallow Mares Australes' adjustment to bonuses and that the full amount of bonuses recognized should remain in the cost of production of Atlantic salmon. Because Mares Australes has accounted for its fiscal year on the accrual basis, that is, in the normal course of business, it recognized the expenses to be incurred for the period, whether or not yet fully paid, it should be required to report this information to the Department.

   [[Page 31434]]

   DOC Position: We agree with the petitioners that Mares Australes' bonus expense should reflect the amounts recorded in the company's audited financial statements. Mares Australes follows accrual accounting in its normal books and records. We therefore consider it inappropriate to rely on a cash-basis accounting method for bonus payments, a single expense identified by the company. Accordingly, we have included the bonus amount recognized in the company's accounting records in the cost of Atlantic salmon.

   Cost Issues--Marine Harvest

   Comment 32: Major Input.

   Marine Harvest argues that if the Department does not rely on the costs from the company's affiliated feed producer, Marine Feed, it should use only the market prices for feed comparable to Marine Harvest's proprietary feed formula in order to value the affiliated feed purchases. According to Marine Harvest, the salmon harvested during the POI were raised on a diet of a unique proprietary feed that was produced only by Marine Feed. Marine Harvest argues that the feed prices charged by other unaffiliated feed producers cannot be used to value feed inputs produced by Marine Feed because they were for experimental trials produced with alternative feed formulations.

   Marine Harvest further contends that the Department has recognized that any application of the ``major input'' rule must deal with ``identical'' or ``comparable transactions of similar inputs.'' See, e.g., Final Determination of Sales at Less Than Fair Value: Engineered Process Gas Turbo-Compressor Systems, Whether Assembled or Unassembled, and Whether Complete or Incomplete, from Japan, 62 FR 24394, 24411 (May 5, 1997)(Comment 15). Marine Harvest argues that, therefore, any calculation of the market price for feed must be based on unaffiliated producers of Marine Harvest's proprietary feed formula. Marine Harvest also argues that the small amount of feed sold by Marine Feed to unaffiliated purchasers demonstrates that the price charged by Marine Feed to Marine Harvest was an arm's-length market price.

   The petitioners contend that the Department should value salmon feed purchases from Marine Feed at the average price of all unaffiliated purchases. The petitioners argue that there is nothing in the Department's cost verification report that supports Marine Harvest's contention that the average unaffiliated feed price was based on a product formula that could not be compared to the feed that Marine Harvest purchased from Marine Feed.

   DOC Position: As discussed in our response to Comment 22, we have followed our practice of using the higher of transfer price, market value or cost of production when valuing major inputs from affiliated suppliers. Accordingly, we continue to value feed purchased from Marine Harvest's affiliated feed supplier, Marine Feed, based on the market value of the input. As to Marine Harvest's claim that the market value for its purchases from Marine Feed must be based only on purchases from unaffiliated producers of its ``proprietary'' feed formula, we note that this argument was first raised in the company's case brief and, therefore, the Department was unable to examine this claim during its verification of the submitted data. There is no record evidence detailing the recipes for Marine Harvest's affiliated or unaffiliated feed purchases. Further, there is no record evidence that feed produced using Marine Harvest's proprietary formula is not sufficiently similar to feed produced by the unaffiliated companies for purposes of comparing transfer prices to market prices under section 773(f)(2) of the Act. Therefore, we used the weighted average of Marine Harvest's purchases from all unaffiliated feed suppliers in order to value the company's affiliated feed purchases for this final determination.

   Cost Issues--Camanchaca

   Comment 33: Area Management Expenses.

   Camanchaca argues that the Department has double-counted area management expenses in its recalculated G&A ratio. According to Camanchaca, because the company's submitted cost of manufacturing figures already included area management expenses, it was necessary to exclude these amounts from G&A in order to avoid double counting. In addition, Camanchaca claims that the Department's calculation of the company-wide G&A rate includes administration costs for non-salmon producing areas of the company. Camanchaca asserts that the G&A ratio should be calculated based only on areas related to salmon production, and cites as support for its position, the Department's decision in the Final Determination of Sales at Less Than Fair Value: Furfuryl Alcohol From South Africa, 60 FR 22550, 22556 (May 8, 1995) (LTFV determination in Furfuryl Alcohol from South Africa) (Comment 15).

   In rebuttal, the petitioners argue that the Department calculated correctly Camanchaca's G&A expense rate. The petitioners point out that Camanchaca did not follow the instructions in the Department's antidumping questionnaire with respect to reporting of G&A expenses. According to the petitioners, instead of reporting a company-wide G&A rate, Camanchaca shifted expenses from G&A to factory overhead by basing its G&A rate on only the salmon division of the company.

   DOC Position: In recalculating G&A expenses for Camanchaca, we excluded from the company's G&A expenses the local administration costs of Puerto Montt and Tome because these costs were already included in the cost of manufacturing. Additionally, we reduced Camanchaca's company-wide G&A expenses for the amounts reported as indirect selling expenses.

   As to the respondent's citation to the LTFV determination in Furfuryl Alcohol from South Africa case, we do not believe that this case supports Camanchaca's claim that the G&A rate should be calculated based only on areas of the company related to salmon production. In that proceeding, the respondent maintained its normal books and records in such a way that its chemical operations, including subject merchandise, maintained specific G&A accounts in the general ledger. As a result, the company's G&A rate was calculated based on the sum of the overall company G&A expenses, consistent with the Department's normal methodology, and also included certain chemical operations-specific G&A expenses.

   Comment 34: G&A Expenses Allocation Base.

   Camanchaca explains that the cost of goods sold figure used to calculate the G&A and financial expense ratios includes packing cost. Thus, according to Camanchaca, G&A and financial expense ratios should be applied to packing costs, which the company claims would increase the packing expense for U.S. sales.

   DOC Position: We disagree with respondent that the G&A and financial expense ratios should be applied to packing costs. We note that the packing costs are included in the cost of sales denominator used in calculating Camanchaca's G&A and financial expense ratios. Thus, in order to correctly reflect the G&A and financial expenses incurred by Camanchaca, these ratios must be applied to the salmon production costs inclusive of the reported packing expenses. Moreover, in calculating packing costs it is not the

   [[Page 31435]]

   Department's practice to include G&A and financial expenses.

   For this final determination, we have applied the G&A and financial expense ratios to the total of COM and packing costs. See Final Results of Antidumping Duty Administrative Review and Partial Termination of Administrative Review: Circular Welded Non-Alloy Steel Pipe From the Republic of Korea, 62 FR 55574, 55580 (October 27, 1997)

   (Comment 6) where the Department determined the same conclusion for this issue.

   Comment 35: CV Profit Rate for Camanchaca.

   Camanchaca does not have a viable home or third-country market. In the preliminary determination, the Department based normal value for Camanchaca on CV, and based CV profit on a weighted average of the profit rates of the other four Chilean producers on sales of the foreign like product in their respective comparison markets. Camanchaca argues that this method is an arbitrary and unreasonable surrogate for Camanchaca's home market profit. Camanchaca contends that the antidumping law establishes a preference for company-specific data in the calculation of profit for CV, and that the average profit realized by the four other respondents in the Japanese and Canadian markets is not a reasonable surrogate for Camanchaca's home market profit, because those respondents have very different costs, expenses, and profit levels.

   Camanchaca argues that, instead, the Department should rely on Camanchaca's average profit rate from total worldwide sales, as reflected in the company's 1995 and 1996 audited financial statements. Camanchaca states that the Department has accepted the use of a company's overall worldwide profit under similar circumstances in other cases, provided that the overall profit rate reflects sales of the same general category as the foreign like product. According to Camanchaca, its operations are all fish and seafood-related, and are all related within the same general category of merchandise as fresh Atlantic salmon, so that the company's overall profit would be a reasonable and representative surrogate for home market profit from the sales of salmon.

   The petitioners respond that the Department's use of an average of the profit for the other four respondents as a surrogate for Camanchaca's profit on the foreign like product is both reasonable and consistent with statutory requirements and Department practice. According to the petitioners, it would be inappropriate to use Camanchaca's worldwide profit, as that profit would reflect sales of merchandise other than the foreign like product, as well as sales made outside the POI. The petitioners note that Camanchaca has argued with respect to other issues that costs incurred in relation to other merchandise are vastly different from costs incurred on fresh Atlantic salmon, and that costs incurred outside the POI are not representative of POI costs.

   The petitioners further contend that the cases cited by Camanchaca are not on point because, in those cases, the Department had acknowledged that the respondent's worldwide profit was the most appropriate basis for profit based on the record of that case.

   DOC Position: We have continued to calculate the surrogate profit rate for Camanchaca based on the weighted average of the profit rates of the other respondents.

   As explained in detail in the preliminary determination, the Department must calculate profit for Camanchaca in accordance with section 773(e)(2)(B)(iii) of the Act, which allows for profit to be based on ``any other reasonable method.'' Given the fact pattern in this case, we find that the use of the weighted average of the profit rates of the other respondents is a reasonable method. That weighted- average rate is based on POI sales of the foreign like product, the reliability of which the Department has ascertained through verification. Camanchaca has not provided any specific reason why the profit rates of the other respondents are unreliable, stating only that each of the other four respondents has ``different costs, expenses, and profit levels.'' See Association's Case Brief at II-52. We do not believe that differences in the various profit rates render an average of those rates an unreliable surrogate profit; on the contrary, the very purpose of an average rate is to capture the range of profit experienced by the other parties to the proceeding.

   Moreover, we believe that it would be far less reasonable in this case to rely on Camanchaca's worldwide profit for 1995 and 1996 as a surrogate profit. First, Camanchaca's only significant market for fresh Atlantic salmon is the United States. To the extent that Camanchaca's profit on the sale of fresh Atlantic salmon has a significant weight in the company's overall profit, it is based in large part on U.S. sales that are subject to an antidumping investigation, and therefore inherently suspect. Second, as the petitioners correctly point out, Camanchaca has acknowledged with respect to other issues that costs incurred in relation to other merchandise are vastly different from costs incurred on fresh Atlantic salmon (see Comment 26, below), and that costs incurred outside the POI are not representative of POI costs (see Comment 24, above). These assertions by Camanchaca cast further doubt on the representativeness of Camanchaca's worldwide profit for a period largely outside the POI.

   In view of the above, we believe that the use of the weighted average of the profit rates of the other respondents is not only reasonable (thus meeting the standard required by statute), but also preferable to the alternative methodology proposed by Camanchaca. Therefore, as in our preliminary determination, we have continued to calculate Camanchaca's profit, as facts available under section 773(e)(2)(b)(iii) of the Act, based on the profits realized by the other four respondents in sales to their respective comparison markets.

   Cost Issues--Aguas Claras

   Comment 36: Feed Costs.

   Aguas Claras maintains that, while it agrees with the Department's conclusion that the company miscalculated the amount of discount on feed purchased from its supplier, EWOS Chile S.A. (EWOS), the amount of the correction in the Department's cost verification report overstates the actual amount of the error.

   The petitioners contend that the Department should disallow the feed purchase discount paid by EWOS for reasons that are proprietary in nature. Additionally, the petitioners argue that the Department should not allow Aguas Claras to reduce its feed costs for the EWOS discount because the company had knowledge of an impending trade case when it entered into the EWOS feed agreement. Furthermore, the petitioners claim that Aguas Claras applied the feed discount to salmon which were harvested before the contract was entered into and, therefore, these fish could not have consumed any EWOS feed.

   DOC Position: We disagree with Aguas Claras' claim that our adjustment to EWOS' feed discount overstates the actual amount of the company's calculation error. The amount of the discount in question was identified in Article 15 of the feed supplier contract between Aguas Claras and EWOS. Aguas Claras initially calculated its cost of EWOS- supplied feed using a methodology that tied the discount to specific feed purchases. The contract, however, does not contain any such specific provisions relating the discount to feed purchases. In fact, provisions of the contract specify only the period for which it is in effect. To correct Aguas

   [[Page 31436]]

   Claras' calculation error, we amortized the discount specified in Article 15 over the life of the contract and reduced feed cost by only the portion of the discount that was amortized within the POI. We then allocated this amount to individual fish groups based on each groups' relative biomass.

   Comment 37: Unreported Costs.

   Aguas Claras argues that the Department's cost verification report erroneously concluded that the respondent had not reported in its submitted COP and CV certain packing and ice costs that were recorded outside the company's normal cost accounting system. Aguas Claras claims that it included the amount of these costs related to salmon production in the minor corrections presented at the beginning of verification.

   The petitioners state that the Department should include in COP and CV the packing and ice costs that Aguas Claras' failed to report.

   DOC Position: We agree with the respondent. We reexamined the information on the record and determined that Aguas Claras did, in fact, include the packing and ice costs in question in the revised COP and CV figures it submitted as minor corrections at the beginning of verification. Therefore, we have not made any additional adjustment for these costs. See Aguas Claras Cost Verification Report at exhibits B25 (the overall reconciliation) and B1 (the minor corrections exhibit).

   Comment 38: Sale of Investment.

   Aguas Claras claims that because Salmofood S.A. and Antarfrio Invertec S.A. were involved in the production and processing of Atlantic salmon, it is correct to reduce the company's G&A expenses with the gain earned from the sale of its investment in the two affiliates. Aguas Claras argues that its shareholdings in the two companies were not simply passive investments but, instead, represent joint ventures related to the production of fresh Atlantic salmon. Aguas Claras asserts that there is no practical difference between the sale of fixed assets of a feed mill or processing plant, which it claims the Department recognizes in calculating G&A expenses, and the sale of shares in such a feed mill or processing company.

   The petitioners argue that Aguas Claras incorrectly reduced G&A expenses for its gain on the sale of common stock in Antarfrio Invertec and Salmofood. The petitioners state that Aguas Claras did not sell the assets of these companies but instead sold only its equity investment in the companies. The petitioners claim that the gain on the sale of common stock is not a part of the day-to-day business of producing salmon. In support of its argument, the petitioners indicate that the gain was shown on Aguas Claras' income statement as ``other income.'' Therefore, the petitioners claim that Aguas Claras itself confirmed that the gain was from an investment and not related to the production of subject merchandise. The petitioners allege that the sales of the affiliated companies were not conducted for bona fide commercial reasons, but to influence the antidumping investigation.

   DOC Position: For the final determination in this case, we have not reduced Aguas Claras G&A expense for the amount of gain that the company received from its sales of Salmofood and Antarfrio Invertec. It is the Department's practice to consider the disposal of fixed assets used to produce the merchandise under investigation to be a normal part of a company's operations. Thus, the Department typically accounts for the gains or losses generated from these transactions as part of G&A expense in the COP and CV calculations. See, e.g., Minivans from Japan at 21943. However, the Department considers the transfer of an equity interest in another company as a sale of an investment, which is unrelated to the production activities for G&A expenses. Neither is the gain or loss from an investment activity considered part of financial expenses, since the investment is unrelated to financing the company's working capital. See, e.g., Final Determination of Sales at Less than Fair Value: Oil Country Tubular Goods from Korea, 60 FR 33561, 33567 (June 28, 1995). Moreover, in this case, we disagree with Aguas Claras' characterization of its sale of common stock in Salmofood and Antarfrio Invertec as the equivalent of a disposal of fixed assets related to the company's salmon production. Specifically, the sale of stock in a company is, indeed, the sale of an interest in all assets of the company.

   Comment 39: Cost of Idle Facility.

   Aguas Claras argues that because the cost of the idled salmon smoking plant facility related solely to the production of non-subject merchandise, it properly excluded these costs from the reported G&A expenses. Aguas Claras cites several cases where the Department excluded the costs associated with idled or inactive facilities where those facilities produced non-subject merchandise.

   The petitioners contend that the costs associated with the idle facilities were incorrectly excluded from G&A.

   DOC Position: We agree with respondent that the costs of the idled salmon smoking plant should be excluded from the G&A expenses of the company. The Department's general practice recognizes that all costs incurred during a period should be absorbed by the company's sales of all products during that same period. As we stated in Silicomanganese from Brazil at 37871, we consider idle facility costs to be period costs (i.e., costs that are more closely related to the accounting period rather than the current manufacturing costs). While it is the Department's general practice to include the cost of shutdowns and idle assets in the COP and CV, in this case we determined that the salmon smoking facilities were idle for only a short time and that the smoking facilities later resumed production during the POI. Therefore, the costs associated with this temporary shutdown of the smoking plant are more appropriately absorbed by the smoked salmon products sold during the POI, rather than absorbed by all products.

   Comment 40: Calculation of CV Indirect Selling expenses for Aguas Claras.

   Aguas Claras contends that the Department erred in including in CV a fixed amount of selling expenses for different products, rather than an amount proportionate to the cost of manufacturing of each product. Specifically, Aguas Claras notes that it sold both salmon fillets and whole salmon in the Canadian market and claims that, on a per-pound basis, salmon fillets are a higher value product than whole salmon. Aguas Claras contends that, by assigning all products the same per-unit amount of CV indirect selling expenses regardless of the value of the product, the Department's methodology is distortive. Aguas Claras proposes that the Department calculate a weighted-average selling expense ratio and, in computing CV, increase the cost of manufacturing of each product by this ratio, such that selling expenses are proportionate to costs.

   The petitioners respond that, in view of the problems encountered at verification in determining the value of Aguas Claras' sales to Canada (see Comment 7 above), the Department should continue to apply a fixed per-pound weighted-average selling expense to CV for all products.

   DOC Position: We agree with Aguas Claras, and have recalculated CV selling expenses as a percentage of cost of production, thus ensuring that the selling expenses for higher value-added products are proportionately higher than the selling expenses apportioned to

   [[Page 31437]]

   lower value-added products. This is consistent with the methodology used in Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, Germany, Italy, Japan, Romania, Singapore, Sweden, and The United Kingdom: Notice of Preliminary Results of Antidumping Duty Administrative Reviews and Partial Termination of Administrative Reviews, 63 FR 6512 (February 9, 1998).

   We do not agree with the petitioners' argument that, due to shortcomings in Aguas Claras' recordkeeping discovered at verification, it would be more appropriate to apply a fixed average selling expense to all products. However, we cannot address the specifics of the petitioners' argument in this public forum, as a meaningful discussion is only possible by means of reference to business proprietary information. We have addressed the petitioners' argument in a separate memo to the file, which has been placed on the official record, and served upon parties with access to such information under administrative protective order.

   We note that, although only Aguas Claras requested that the Department recalculate CV indirect selling expenses, to ensure consistency in our calculations for the other respondents we have also revised their CV indirect selling expenses on the same basis.

   Continuation of Suspension of Liquidation

   In accordance with section 735(c)(4)(B) of the Act, we are directing the Customs Service to continue suspending liquidation of all entries of fresh Atlantic salmon from Chile, except for subject merchandise produced and exported by Camanchaca and Marine Harvest (which have de minimis weighted-average margins), that are entered, or withdrawn from warehouse, for consumption on or after January 16, 1998 (the date of publication of the preliminary determination in the Federal Register). The Customs Service shall continue to require a cash deposit or the posting of a bond equal to the weighted-average amount by which the normal value exceeds the EP or CEP, as indicated in the chart below. These instructions suspending liquidation will remain in effect until further notice.

   The weighted-average dumping margins are as follows:

   Weighted- average Exporter/manufacturer

   margin percentage

   Aguas Claras................................................

   8.27 Camanchaca..................................................

   0.21 Eicosal..................................................... 10.91 Mares Australes.............................................

   2.24 Marine Harvest..............................................

   1.36 All Others..................................................

   5.19

   Section 735(c)(5)(A) of the Act directs the Department to exclude all zero and de minimis weighted-average dumping margins, as well as dumping margins determined entirely under facts available under section 776 of the Act, from the calculation of the ``all others'' rate. As explained above in Comment 5, we have therefore excluded the de minimis dumping margins for Camanchaca and Marine Harvest from the calculation of the ``all others'' rate. No dumping margins were based entirely on facts available.

   ITC Notification

   In accordance with section 735(d) of the Act, we have notified the ITC of our determination. As our final determination is affirmative, the ITC will, within 45 days, determine whether these imports are materially injuring, or threaten material injury to, the U.S. industry. If the ITC determines that material injury or threat of material injury does not exist, the proceeding will be terminated and all securities posted will be refunded or canceled. If the ITC determines that such injury does exist, the Department will issue an antidumping duty order directing the Customs Service to assess antidumping duties on all imports of the subject merchandise entered for consumption on or after the effective date of the suspension of liquidation.

   This determination is published pursuant to sections 735(d) and 777(i) of the Act.

   Dated: June 1, 1998. Robert S. LaRussa, Assistant Secretary for Import Administration.

   [FR Doc. 98-15183Filed6-8-98; 8:45 am]

   BILLING CODE 3510-DS-P