Antidumping: Dynamic random access memory semiconductors of one megabit or above (DRAMs) from— Korea,

 
CONTENT

[Federal Register: December 14, 1999 (Volume 64, Number 239)]

[Notices]

[Page 69694-69718]

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

[DOCID:fr14de99-35]

DEPARTMENT OF COMMERCE

International Trade Administration

[A-580-812]

Dynamic Random Access Memory Semiconductors of One Megabit or Above From the Republic of Korea: Final Results of Antidumping Duty Administrative Review and Determination Not To Revoke the Order in Part

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

ACTION: Notice of final results of antidumping duty administrative review and determination not to revoke the order in part.

SUMMARY: On June 8, 1998, the Department of Commerce (``the Department'') published the preliminary results of its administrative review of the antidumping duty order on dynamic random access memory semiconductors of one megabit or above (``DRAMs'') from the Republic of Korea (``Korea''). The review covers two manufacturers/exporters of the subject merchandise to the United States and one reseller for the period May 1, 1997, through April 30, 1998. The two manufacturers/ exporters are Hyundai Electronics Industries, Co. (``Hyundai''), and LG Semicon Co., Ltd. (``LG''). The reseller is the G5 Corporation (``G5'').

As a result of our analysis of the comments received, we have changed the results from those presented in our preliminary results of review.

EFFECTIVE DATE: December 14, 1999.

FOR FURTHER INFORMATION CONTACT: Thomas Futtner, Alexander Amdur (``Hyundai''), or John Conniff (``LG''), AD/CVD Enforcement, Group II, Office IV, Import Administration, International Trade Administration, U.S. Department of Commerce, 14th Street and Constitution Avenue, N.W., Washington, D.C. 20230; telephone: (202) 482-3814, (202) 482-5346, and (202) 482-1009, respectively.

SUPPLEMENTARY INFORMATION:

Applicable Statute and Regulations

Unless otherwise stated, all citations to the Tariff Act of 1930, as amended (``the Act''), are references to the provisions as of January 1, 1995, the effective date of the amendments made to the Act by the Uruguay Round Agreements Act (``URAA''). In addition, unless otherwise indicated, all references to the Department's regulations are to 19 CFR 351 (1998).

Background

On June 8, 1999, the Department published in the Federal Register (64 FR 30481) the preliminary results of its administrative review of the antidumping duty order on DRAMs from Korea. On September 13, 1999, we released information to interest parties pertaining to possible unreported sales by LG. On October 7, 1999, LG and an interested party submitted factual information relevant to this issue. We also gave interested parties an opportunity to comment on this information and our preliminary review results.

The petitioner, Micron Technology, Inc. (``Micron''), Hyundai, and LG submitted case briefs on October 21, 1999, and rebuttal briefs on October 28, 1999. We held both public and closed hearings on November 4, 1999. We have now completed this administrative review in accordance with section 751(a) of the Act.

Scope of Review

Imports covered by the review are shipments of DRAMs from Korea. Included in the scope are assembled and unassembled DRAMs. Assembled DRAMs include all package types. Unassembled DRAMs include processed wafers, uncut die, and cut die. Processed wafers produced in Korea, but packaged or assembled into memory modules in a third country, are included in the scope; wafers produced in a third country and assembled or packaged in Korea are not included in the scope.

The scope of this review includes memory modules. A memory module is a collection of DRAMs, the sole function of which is memory. Modules include single in-line processing modules (``SIPs''), single in-line memory modules (``SIMMs''), or other collections of DRAMs, whether unmounted or mounted on a circuit board. Modules that contain other parts that are needed to support the function of memory are covered. Only those modules which contain additional items which alter the function of the module to something other than memory, such as video graphics adapter (``VGA'') boards and cards, are not included in the scope. The scope of this review also includes video random access memory semiconductors (``VRAMS''), as well as any future packaging and assembling of DRAMs; and, removable memory modules placed on motherboards, with or without a central processing unit (``CPU''), unless the importer of motherboards certifies with the Customs Service that neither it nor a party related to it or under contract to it will remove the modules from the motherboards after importation. The scope of this review does not include DRAMs or memory modules that are reimported for repair or replacement.

The DRAMS and modules subject to this review are currently classifiable under subheadings 8471.50.0085, 8471.91.8085, 8542.11.0024, 8542.11.8026, 8542.13.8034, 8471.50.4000, 8473.30.1000, 8542.11.0026, 8542.11.8034, 8471.50.8095, 8473.30.4000, 8542.11.0034, 8542.13.8005, 8471.91.0090, 8473.30.8000, 8542.11.8001, 8542.13.8024, 8471.91.4000, 8542.11.0001, 8542.11.8024 and 8542.13.8026 of the Harmonized Tariff Schedule of the United States (``HTSUS''). Although the HTSUS subheadings are provided for convenience and customs purposes, the

[[Page 69695]]

Department's written description of the scope of this review remains dispositive.

Determination Not To Revoke

LG submitted a request for revocation from the order covering DRAMs from Korea pursuant to 19 CFR 351.222(b)(2). Under the Department's regulations, the Department may revoke an order, in part, if the Secretary concludes that: (1) [o]ne or more producers or resellers covered by the order have sold the merchandise at not less than

[normal] value for a period of at least three consecutive years; (2)

[i] t is not likely that those persons will in the future sell the merchandise at less than normal value (``NV''); and (3) the producers or resellers agree in writing to the immediate reinstatement of the order, as long as any producer or reseller is subject to the order, if the Secretary concludes that the producer or reseller, subsequent to the revocation, sold the merchandise at less than NV. See 19 CFR 351.222(b)(2). In this case, LG does not meet the first criterion for revocation. The Department found that LG sold the subject merchandise at less than NV during the previous review period. See DRAMs from the Republic of Korea: Final Results of Antidumping Duty Administrative Review, 63 FR 50867 (September 23, 1998) (``Final Results 1998''). Since LG has not met the first criterion for revocation, i.e., zero or de-minimis margins for three consecutive reviews, the Department need not reach a conclusion with respect to the other criteria. Therefore, on this basis, we have determined not to revoke the Korean DRAM antidumping duty order with respect to LG. In light of this decision, interested party comments on revocation are moot and will not be addressed further in these final results.

Facts Available (``FA'')

In accordance with section 776(a) of the Act, we have determined that the use of adverse FA is warranted for LG and G5 for these final results of review.

  1. Application of FA

    Section 776(a) of the Act provides that, if an interested party withholds information that has been requested by the Department, fails to provide such information in a timely manner or in the form or manner requested, significantly impedes a proceeding under the antidumping statute, or provides information which cannot be verified, the Department shall use, subject to sections 782(d) and (e), facts otherwise available in reaching the applicable determination. In this review, as described in detail below, the above-referenced companies failed to provide the necessary information in the form and manner requested, and, in some instances, the submitted information could not be verified. Thus, pursuant to section 776(a) of the Act, the Department is required to apply, subject to section 782(d), facts otherwise available.

    Section 782(d) of the Act provides that, if the Department determines that a response to a request for information does not comply with the request, the Department will inform the person submitting the response of the nature of the deficiency and shall, to the extent practicable, provide that person the opportunity to remedy or explain the deficiency. If that person submits further information that continues to be unsatisfactory, or this information is not submitted within the applicable time limits, the Department may, subject to section 782(e), disregard all or part of the original and subsequent responses, as appropriate.

    Pursuant to section 782(e) of the Act, notwithstanding the Department's determination that the submitted information is ``deficient'' under section 782(d) of the Act, the Department shall not decline to consider such information if all of the following requirements are satisfied: (1) The information is submitted by the established deadline; (2) the information can be verified; (3) the information is not so incomplete that it cannot serve as a reliable basis for reaching the applicable determination; (4) the interested party has demonstrated that it acted to the best of its ability; and (5) the information can be used without undue difficulties.

  2. Selection of FA

    In selecting from among the facts otherwise available, section 776(b) of the Act authorizes the Department to use an adverse inference if the Department finds that an interested party failed to cooperate by not acting to the best of its ability to comply with the request for information. See, e.g., Certain Welded Carbon Steel Pipes and Tubes From Thailand: Final Results of Antidumping Duty Administrative Review, 62 FR 53808, 53819-20 (Oct. 16, 1997) (Pipe and Tubes From Thailand). In this segment of the proceeding, the Department has determined that it is appropriate to apply in these final review results total adverse facts available to both LG and G5.

    G5

    For purposes of the preliminary results, the Department concluded that, because G5 failed to respond to the Department's questionnaire, a determination based on total adverse FA was warranted for this company. We, accordingly, assigned an adverse FA rate and articulated detailed reasons for our decision in Dynamic Random Access Memory Semiconductors of One Megabit or Above From the Republic of Korea: Preliminary Results of Antidumping Duty Administrative Review and Notice of Intent Not To Revoke Order in Part, 64 FR 30481 (June 8, 1999) (``Preliminary Results'').

    For the final results, no interested party comments were submitted regarding this issue and we continue to find that G5's failure to respond to the Department's questionnaire in this review demonstrates that it failed to cooperate by not acting to the best of its ability. Thus, consistent with the Department's practice in cases where a respondent fails to respond to the Department's questionnaire, in selecting FA for G5 in this review, an adverse inference is warranted. Therefore, we are assigning G5 an adverse FA rate of 10.44 percent, the rate calculated for Hyundai in this review and the highest margin from any segment of the proceeding related to DRAMS from Korea.

    LG

    Based on information obtained from Customs, the Department preliminarily determined, as it had in the prior review, that numerous sales which LG had reported as third-country sales, were in fact sales to the United States. See Preliminary Results. For the final results, we have considered interested party comments (see the Department Position to LG Comment 1) and continue to find that sales which LG had reported as third-country sales, were in fact sales to the United States. See Memorandum for Holly A. Kuga, from John Conniff regarding Dynamic Random Access Memory Semiconductors of One Megabit and Above (DRAMs) from the Republic of Korea--LG Sales through Mexico, December 3, 1999.

    Similarly, on January 4, 1999, the Department received an e-mail from a former LG employee stating that LG was shipping subject merchandise from Korea to the United States through a customer in Europe and these shipments were being made with the knowledge and support of LG's senior management.

    [[Page 69696]]

    At verification, LG submitted information related to these sales which had been made by its German subsidiary, LG Germany (``LGSG''). Subsequently, the Department queried Customs to determine if any of the sales by LGSG to the European customer in question had entered the United States. Customs data revealed entries covering Korean DRAMs imports into the United States by the European customer's affiliate in the United States. The quantities and values of DRAMs shown in the entries were substantially identical to the quantities and values of DRAMs reflected on the invoices between LGSG and the European customer in question. Documentation from randomly selected sample entries covering transactions between the European customer and its U.S. operation confirm that the DRAMs in question were manufactured in Korea by LG.

    In August 1999, information was provided to the Department by a former LG employee familiar with and responsible for worldwide sales to the customer in question during the period of review (``POR''). He stated that at the time that he sold DRAMs to the customer in question in Europe, he had knowledge the DRAMS were ultimately destined for the customer's operation in the United States. See Memorandum for Holly A. Kuga, from John Conniff regarding Dynamic Random Access Memory Semiconductors of One Megabit and Above (DRAMs) from the Republic of Korea--LG Sales Through Germany, December 6, 1999 (``LG Sales Through Germany Memo'').

    On September 13, 1999, LG was provided with information collected by the Department related to this matter and on September 22, 1999, LG was provided an opportunity to submit factual information and comments concerning this issue. See Letter from the Department to Michael House, Esq. regarding Dynamic Random Access Memory Semiconductors of One Megabit (DRAMs) from the Republic of Korea, September 22, 1999 (``September 22, 1999, letter''). On October 7, 1999, LG submitted factual information and on October 21, 1999, and October 28, 1999, LG also presented comments and arguments on this matter in its case briefs and rebuttal briefs.

    Based on the record evidence the Department concluded that LG knew at the time it sold the subject DRAMS, that the merchandise was destined for consumption in the United States. See LG Sales Through Germany Memo. As LG did not report these sales, in accordance with section 782(d) of the Act, the Department provided LG with the opportunity to explain its deficiencies with respect to unreported U.S. sales. See September 22, 1999, letter. However, LG failed to correct these deficiencies. Thus, the Department is required, under section 782(d) to apply, subject to section 782(d) of the Act, FA.

    We further determine that LG failed to satisfy several of the requirements enunciated by 782(e) of the Act. First, LG failed to report a significant portion of the company's U.S. sales data. Second, because the unreported sales are significant, LG's U.S. sales data is so incomplete that it cannot serve as a reliable basis for reaching the applicable determination pursuant to subsection (e)(3). Third, LG did not demonstrate that it acted to the best of its ability in providing the necessary information under subsection (e)(4). Fourth, given the incompleteness of LG's responses, the information could not be used without undue difficulties, as required by subsection (e)(5). We thus find that LG did not act to the best of its ability to comply with the request for information under section 776(b) and that, under section 776(b), an adverse inference is warranted. Therefore, we are assigning LG an adverse FA rate of 10.44 percent, the rate calculated for Hyundai in this review, which is the highest margin from any of the proceedings related to DRAMS from Korea.

    Duty Absorption

    On July 27, 1998, the petitioner requested that the Department determine whether antidumping duties had been absorbed during the POR. Section 751(a)(4) of the Act provides for the Department, if requested, to determine during an administrative review initiated two or four years after the publication of the order, whether antidumping duties have been absorbed by a foreign producer or exporter, if the subject merchandise is sold in the United States through an affiliated importer. In this case, both Hyundai and LG sold to the United States through importers that are affiliated within the meaning of sections 751(a)(4) and 771(33) of the Act.

    Section 351.213(j)(2) of the Department's regulations provides that for transition orders (i.e., orders in effect on January 1, 1995), the Department will conduct duty absorption reviews, if requested, for administrative reviews initiated in 1996 or 1998. Because the order underlying this review was issued prior to January 1, 1995, and this review was initiated in 1998, we will make a duty absorption determination in this segment of the proceeding.

    On January 26, 1999, the Department requested evidence that unaffiliated purchasers will ultimately pay the antidumping duties to be assessed on entries during the review period. Neither Hyundai nor LG provided any evidence in response to the Department's request. Accordingly, based on the record, we cannot conclude that the unaffiliated purchaser in the United States will ultimately pay the assessed duty. Therefore, we find that antidumping duties have been absorbed by the producer or exporter during the POR. For further discussion, see DOC position to general comment 5.

    Fair Value Comparisons

    Unless otherwise noted, to determine whether sales of subject merchandise from Korea to the United States were made at less than fair value, we compared the Constructed Export Price (``CEP'') to the NV, as described in the ``Constructed Export Price'' and ``Normal Value'' sections of the preliminary results of review notice. See Dynamic Random Access Memory Semiconductors (``DRAMs'') of One Megabit or Above from the Republic of Korea, 64 FR 40481, (June 8, 1999) (``Preliminary Results'').

    Interested Party Comments

    General Comments

    Comment 1: Deferral of Research and Development (``R&D'') Expenses. Hyundai and LG argue that the Department erred in rejecting their accounting methodology for the amortization and deferral of R&D expenses. Hyundai and LG, citing to Micron Technology v. United States, 893 F. Supp. 21, 28 (U.S. Court of International Trade (``CIT'') 1995) (``Micron I''), The Thai Pineapple Public Co., Ltd. v. United States, 187 F.3d 1362, 1366 (Fed. Cir. 1999), NTN Bearing Corp. v. United States, 17 CIT 713, 826 F. Supp. 1435, 1441 (1993), Ipsco, Inc. v. United States, 12 CIT 384, 687 F. Supp. 633, 636 & n.3 (1988), Color Television Receivers from Korea, 53 FR 24975, 24982 (July 1, 1988) (``CTVs from Korea''), and Gilbert B. Kaplan, Marie Parker, et. al., Cost Analysis Under the Antidumping Law, 21 George Wash. J. Int'l L. & Econ. 357, 373-74 (1988), contend that the accounting methodology at issue is in conformity with Article 70.5 of Korean generally accepted accounting principles (``GAAP''), and under section 773(f)(1)(A) of the Act, the Department must accept this methodology unless it finds that the reported costs are distortive. Hyundai contends that the Act's preference for the exporting

    [[Page 69697]]

    country GAAP reflects Article 2.2.1.1 of the WTO Antidumping Agreement, and that the Department has followed such a preference in numerous cases, including Steel Wire Rod from Canada, 63 FR 9182 (February 28, 1998), Collated Roofing Nails from Korea, 62 FR 51420, 51423 (October 1, 1997); Ferrosilicon from Brazil, 62 FR 43504, 43511 (August 14, 1997); Polyethylene Terephthalate Film, Sheet, and Strip from the Republic of Korea, 61 FR 35177, 35179 (July 5, 1996); Certain Corrosion-Resistant Carbon Steel Flat Products from Korea, 61 FR 18547, 18568 (April 26, 1996); and Industrial Nitrocellulose from France, 48 FR 21615, 21617 (May 13, 1983).

    Hyundai and LG further contend that there is no basis for finding that Hyundai's reported R&D costs are distortive. Hyundai and LG state that in Micron I, 893 F. Supp. at 29, the CIT specifically held that, for the DRAM industry, the amortization of R&D expenses, as allowed by Korean GAAP, is not distortive, and that the three to five-year amortization period allowed by Korean GAAP was more reasonable than the Department's expensing methodology. LG states that the Court's logic in Micron I applies to LG's deferral of certain R&D expenses until the related projects achieve commercial realization, as this methodology allows R&D costs to be allocated over the commercial life of the product. Hyundai also states that, in DRAMs from Korea, 61 FR 20216, 20219 (May 6, 1996) (``Final Results 1996''), the first administrative review of this proceeding, the Department complied with this ruling in its treatment of LG's R&D expenses.

    Hyundai points out that, in CTVs from Korea, 53 FR at 24982, Certain Welded Stainless Steel Pipe from the Republic of Korea, 57 FR 53693 (November 12, 1992) (``Pipe from Korea''), and Polyethylene Terephthalate Film, Sheet, and Strip from the Republic of Korea, 56 FR 16305 (April 22, 1991) (``PET Film from Korea''), the Department explicitly allowed the amortization of R&D expenses pursuant to Korean GAAP. Hyundai maintains that amortizing R&D expenses is just as appropriate in the present case as it was in those cases.

    Hyundai also maintains that the Department, in Steel Wire Rod from Canada, 63 FR 9182, 9187 (February 24, 1998), recognized that it is not distortive for a company to defer expenses that will benefit future operations. Hyundai states that, in much the same manner, it is reasonable for Hyundai to spread R&D costs over future periods because Hyundai's R&D expenses for the development of new generations of products will benefit future periods by providing sales revenues for improved products.

    Hyundai further argues that the amortization and deferral of R&D expenses under Korean GAAP conforms to the principles of International Accounting Standard (``IAS'') No. 9, which is intended to match costs with products that benefit from those expenditures, and not, as the Department suggests, to ``alleviate losses listed on a company's financial statement.'' Hyundai states that the Department explicitly endorsed IAS No. 9 as the basis for amortizing R&D expenses for products, including semiconductors, in Erasable Programmable Read Only Memories (EPROMs) from Japan, 51 FR 39680, 39682 (October 30, 1986), and Cellular Mobile Telephones and Subassemblies from Japan, 50 FR 45447, 45453 (October 31, 1985) (``Cell Phones from Japan''), and did not rely upon U.S. GAAP to reject amortization of R&D until it issued its decision (which was subsequently overturned by the CIT) in Dynamic Random Access Memory Semiconductors of One Megabit and Above from Korea, 58 FR 15467, 15472 (March 23, 1993) (``Final Determination''). Hyundai also notes that the principles of IAS No. 9 are recognized in Canadian and British accounting standards.

    Hyundai contends that, in view of its ``virtual isolation,'' the treatment of R&D under U.S. GAAP should not be automatically accepted as the standard for determining whether costs are distorted under section 773(f)(1)(A) of the Act. Hyundai notes that U.S. GAAP requires the expensing of R&D expenditures because of the ``presumed'' absence of a relationship between R&D expenditures and subsequent benefits, whereas Hyundai's own experience demonstrates the direct link between R&D expenditures and the revenues derived from the sale of later generations of DRAMs. Hyundai also notes that the U.S. practice of expensing R&D in Financial Accounting Standards Board (``FASB'') Standard No. 2 has been criticized by accounting experts, such as Baruch Lev and Theodore Sougiannis in ``The Capitalization, Amortization, and Value-Relevance of R&D,'' 21 Journal of Accounting & Economics, 107, 134 (1996), and is under review by the FASB. Hyundai states that the FASB has proposed to abandon the U.S. GAAP requirement for expensing in-process R&D acquired in a corporate acquisition in the year of acquisition, and require the amortization of such R&D. Hyundai maintains that although the FASB subsequently tabled this proposal, the FASB has plans to eventually consider the treatment of all R&D, and the Department cannot use a standard with such an uncertain future to judge the validity of Korean GAAP.

    Hyundai also argues that its method of recognizing R&D is consistent with both accounting theory and the SAA. Hyundai notes that Eiden Hendricksen, in Accounting Theory, (Irwin 1992), at 650, endorses the matching of R&D expenses with ``the period benefitted,'' and the SAA, at 835, specifically condones allocating R&D costs over current and future production in order to match the expenditures with the production that benefits from the expenditures. Hyundai contends that its R&D methodology is particularly appropriate for the semiconductor industry in general, and Hyundai in particular. Hyundai states that the nature of its R&D activities, its emphasis on development of specific products, and the steady flow of next generation products, contrary to the rationale of FASB No. 2, produce ``direct and immediate'' benefits. Hyundai argues that a large part of its 1997 R&D expenditures were for products that were to be sold in 1998 and 1999, while the other part of its R&D expenditures involves products that are expected by the Semiconductor Industry Association to be available within the next five years. Hyundai concludes that its treatment of R&D reasonably reflects the cost of producing the subject merchandise.

    Hyundai additionally contends that the Department improperly rejected Hyundai's accounting treatment of R&D expenses on the grounds that Hyundai, under the SAA at 834, had not demonstrated that it had historically utilized such a methodology. Hyundai states that behind this statement in the SAA is the need to justify the appropriate period for amortizing expenses that benefit future production. Hyundai, citing to the Micron I decision, contends that there is a history in the Korean semiconductor industry of amortizing R&D expenses over five years. Hyundai also adds that the SAA at 834 is directed at changes in depreciation, and not R&D, methodology.

    Hyundai further maintains that, in accordance with other standards stated by the Department, its new R&D methodology better reflects the actual costs incurred in producing the subject merchandise than the prior methodology because it matches the cost of R&D with the products that benefit from the R&D. Hyundai also

    [[Page 69698]]

    states that there is no risk that its R&D costs will never be reflected in the dumping calculations. Hyundai notes that Korean GAAP provides for the unamortized, deferred balance of R&D costs to be expensed immediately if the possibility of realizing revenue from an R&D project is remote. Hyundai asserts that the Department's statement that, under this provision, Hyundai ``could potentially . . . never recognize any of the R&D expenses'' deferred implies that Hyundai's auditors would allow the company to violate GAAP by deferring R&D expenses indefinitely, and ignores the fact that Hyundai stated in its questionnaire response that it is following this requirement. Hyundai adds that all R&D costs that were incurred in prior years have already been captured by the Department in the cost calculations of this review and prior reviews, and all current R&D expenditures will be captured in this review and subsequent reviews.

    LG contends that the Department's decision at issue is also inconsistent with the CIT's decision after remand in Micron Technology v. United States, Slip Op. 99-51 (June 16, 1996) (``Micron II''). LG, citing to Micron II, Slip Op. 99-51 at 5, states that Court found that the Department's concern over costs that would never be included in any review was a ``red herring,'' and that such concerns in the present case are similarly misplaced, as the R&D costs that are deferred and amortized in this review period will be captured in subsequent periods. LG also states that the Court, in the same decision, found that the Department's concern that LG would change its accounting procedures to achieve favorable antidumping treatment made ``little, if any, business sense'' (see Id. at 6), and in this review, the Department has even acknowledged that LG did not change its R&D accounting method for antidumping purposes.

    LG also states that, since it changed its accounting methodology for R&D expenses in the normal course of business, it must show only, as in PET Film from Korea, 56 FR at 16312-13, that its methodology is not distortive, and does not have to justify its change in methodology. In this regard, LG points out that the CIT ruled in Micron I that amortization of DRAM R&D costs is not distortive, and is more reasonable than expensing R&D costs in the year incurred.

    LG further states that its R&D expense is lower in 1997, as noted by the Department, purely as a result of the transition in methodologies from expensing to amortizing, since in the transition year there are no prior years' R&D expenses subject to amortization. LG notes that all prior years' R&D expenses have already been included by the Department in prior review periods, and argues that the CIT, in the Micron II decision, ruled that the Department may not penalize LG for changing accounting methodologies in the normal course of business. LG states that, under the Department's reasoning, a company would never be able to change from expensing any cost to amortizing that cost, because the cost in a transition year would always be reduced from prior years as a result of the transition.

    LG also points out that the Department's concerns about its complete deferral of certain R&D expenses are misplaced. LG states that all of the deferred R&D projects are those of which there was no current production, and no related revenue. LG also asserts that the Department's position that R&D related to these future generation projects' benefits current production is not supported by the record.

    Micron contends that the Department correctly rejected Hyundai and LG's accounting method for R&D expenses as distortive of costs, and issued a legally sound determination on this issue. Micron, citing to the SAA at 834, Certain Small Business Telephone Systems and Subassemblies Thereof From Korea, 54 FR 53141, 53149 (December 27, 1989), Certain Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products From Korea, 64 FR 12927, 12944 (March 16, 1999), and Foam Extruded PVC and Polystyrene Framing Stock From the United Kingdom, 61 FR 51411, 51418 (October 2, 1996), states that the issue is not whether the Department has allowed R&D to be expensed or amortized in other cases, but whether a company's own records, even when kept in accordance with local GAAP, distort costs. Micron, citing to the SAA at 834, Mechanical Transfer Presses From Japan, 55 FR 335 (January 4, 1990), and Static Random Access Memory Semiconductors From Taiwan, 63 FR 8909, 8921-22 (February 23, 1998) (``SRAMs from Taiwan''), further states that the Department determines whether costs are distortive by looking to U.S. GAAP for the industry in question, and the Department, in the instant case, fully explained how the respondents' accounting for R&D was distortive under U.S. GAAP.

    Micron maintains that the Court, in Micron I, 893 F. Supp. at 29, did not rule as a matter of law either that the Department must amortize R&D costs, or that Korean GAAP reasonably reflects R&D costs; rather, Micron points out that the Court ruled that Department had failed to articulate a reasoned analysis in support of its decision to expense R&D. Micron also states that the respondents' claim that they amortize R&D over the period of the DRAM ``life cycle'' noted by the Court in Micron I in order to correspond to the life cycle of their products is a ``post-hoc rationalization'' that is not supported by documents prepared by the respondents in the normal course of business. Micron adds that Korean GAAP does not specifically relate the amortization period to the life cycle of the product, and notes that a company could be conducting R&D on a product with a much shorter life cycle than DRAMs, amortize the costs over five years, and still be within Korean GAAP. Micron also notes, that, in any case, FASB No. 2 clearly states there is no direct relationship between R&D expenditures and future benefits.

    Micron further distinguishes the present case from Micron I. First, Micron maintains that, unlike the situation in Micron I, the respondents, in addition to amortization, adopted a new, ``inherently uncertain'' accounting approach to R&D by completely deferring R&D costs until they ``arbitrarily foresee any possibility of realizing revenue.'' Micron also notes that the Department fully explained how this approach affected the respondent's reported costs. Second, Micron argues that, unlike the situation in Micron I, the respondents have repeatedly changed their accounting methodologies for R&D throughout the course of this proceeding, and distorted costs, in order to affect their apparent profitability. In specific regards to Hyundai, Micron states that Hyundai offered no substantive reason for this accounting change, and simply changed its methodology without any change in the nature of its R&D. Third, in Micron I, because the history of the respondent's accounting changes was not before the Court, the Court had no basis to consider, as it would now, the respondent's inability to show that they have ``historically utilized'' such allocations, as required by the SAA at 834.

    Micron disagrees with LG that the Department's treatment of LG's R&D expenses conflicts with the Micron II decision. Micron states that this decision is irrelevant to LG's situation because it only concerned R&D that was previously incurred but not expensed, and prior to this review period, LG had no such R&D expenses.

    Micron further argues, notwithstanding the issue of whether the Department, under the Act, should

    [[Page 69699]]

    defer to international accounting standards over U.S. GAAP when the two standards differ, that the international standards Hyundai discusses are not inconsistent with the Department's analysis. Micron states that both U.S. GAAP, under FASB No. 2, and international standards, under IAS 9, both recognize that R&D expenses, in at least some instances, cannot be matched to revenues because the benefits of R&D cannot be discerned at the time the costs are incurred. In relating this guideline to Hyundai and LG, Micron states that neither respondent demonstrated during 1996 and 1997 that they could have associated future revenues and current revenue with any reasonable certainty, especially in light of the downturn in the DRAM market and the economic crisis in South Korea at that time. Micron further generally states that R&D may benefit future generations, but at the time that R&D is incurred, one cannot tell when or whether R&D will produce a commercially successful result; and amortizing R&D expense over a number of years is a ``self-serving guess.''

    DOC Position: Section 773(f)(1)(A) of the Act, directs to the Department to rely ``on the records of the exporter or producer of the merchandise, if such records are kept in accordance with the GAAP of the exporting country (or the producing country where appropriate) and reasonably reflect the costs associated with production and sale of the merchandise.'' Section 773(f)(1)(A) of the Act also states that the Department will consider whether ``such allocations have been historically used by the exporter or the producer.'' Further, as explained in the SAA, ``[t]he exporter or producer will be expected to demonstrate that it has historically utilized such allocations, particularly with regard to the establishment of appropriate amortization and depreciation periods and allowances for capital expenditures and other development costs.'' See SAA at 834. See also Final Results 1998, 63 FR at 50871.

    We agree with Hyundai and LG that their method of amortizing and deferring R&D costs is permissible with Korean GAAP, and that their previous method of expensing all current period R&D expenses in the year incurred is also in accordance with Korean GAAP. However, Hyundai's and LG's practice of continually changing between these methods distorts the cost calculation in an antidumping analysis. As explained in the Department's Memorandum on ``Whether to Accept the Reported Research & Development Expenses of Hyundai Electronics Industries Co., Ltd. and LG Semicon, Ltd.,'' dated June 1, 1999 (``R&D Memo''), Hyundai and LG have repeatedly changed their accounting method for R&D expenses throughout the course of these proceedings (i.e., from capitalizing and amortizing, to expensing in the year incurred, and now back to capitalizing and amortizing) and are now deferring certain R&D expenses indefinitely. See R&D Memo at 2. As a result, the respondents recognize, in relation to amounts that would be recognized if either method was constantly applied, aberrationally high amounts of R&D expense in some years, and aberrationally low amounts of R&D expense in other years, that do not reasonably reflect the costs of producing the subject merchandise. For example, in the first administrative review of this proceeding, LG changed its method for recognizing R&D expenses from capitalizing and amortizing R&D expenses over five years to expensing in full in the current year. See DRAMs from Korea, 61 FR 20216, 20219 (May 6, 1996) (``Final Results 1996'') and Micron II, Slip Op. 99-551. In that year, LG recognized, in addition to its current year R&D expense, R&D expenses from its balance sheet which it had capitalized in prior years (as part of its capitalizing and amortizing methodology) and not yet amortized and recognized on its income statement. Consequently, in that year, LG recognized the full amount of R&D expenses incurred in that current year (under its expensing methodology) as well as all of the previously unamortized and unrecognized amounts of R&D expenses remaining on its balance sheet from prior years. LG thus recognized in that year significantly higher than normal amounts of R&D expenses than it would have under the consistent application of either methodology.

    In the current review period, Hyundai and LG have changed their accounting methodology for R&D expenses once again, this time back to capitalizing and amortizing their R&D expenses over five years. As a result, the respondents recognize (and have reported to the Department) less than one-fifth of their current year's R&D costs. With the adoption of this new methodology, the respondents next year will recognize approximately one-fifth of that year's R&D expense and approximately one-fifth from the current review period, and will not recognize the equivalent of a full year's R&D expense until at least the fifth year. Thus, because of inconsistent accounting treatment, the respondents are recognizing an aberrationally low amount of R&D expenses.

    A second methodology that further distorts Hyundai's and LG's reported costs is their new practice of indefinitely capitalizing certain R&D costs. Apart from R&D costs that are amortized over five years, Hyundai and LG are now completely deferring R&D costs for certain long-term projects until they realize revenues from these projects, or until they foresee no possibility of realizing revenue from these projects. While in prior years, the respondents recognized all of this type of R&D expense in the year incurred, under the new methodology, none of this R&D expense is recognized in the current year. Moreover, this methodology is contrary to the principle of conservatism in accounting where an expense is recognized when incurred if the probability of associated revenue is remote or uncertain. Therefore, we find that, for dumping purposes, this methodology does not reasonably reflect the cost of producing the subject merchandise.

    Hyundai and LG, by continually changing their R&D accounting methodologies, are manipulating the magnitude of the R&D expenses that they are recognizing, and reporting to the Department. This switching of methodologies can lead to distortions for antidumping purposes because the fluctuating costs tend to overstate per unit amounts in one period and understate these amounts in other periods. The CIT has noted the distortion that such changes in R&D accounting methodologies can cause. In Micron II (which relates to the first review of this proceeding, when LG switched from amortizing to expensing R&D costs currently), the Court ruled that it was distortive for the Department to include in its calculations, as LG included in its own books and records, both the current year's R&D expenses and the unamortized amount of prior years R&D expenses. See Micron II, Slip Op. 99-551. In the same manner that the CIT believes that the amount of R&D expenses that LG recognized, and the Department included in its calculations in the first review (i.e., one full current year amount, plus prior capitalized amounts), was overstated, the amount of R&D expenses that Hyundai and LG recognized in the current review (i.e., less than one- fifth of one year's R&D expense) is understated.

    The Court, in Micron II, specifically stated that ``the object of the cost of production exercise is*** to capture***those expenses that reasonably and accurately reflect a respondent's actual production costs for

    [[Page 69700]]

    a period of review.'' Micron II, Slip Op. 99-551, at 6. However, by abruptly switching to amortizing and deferring R&D expenses, Hyundai and LG are not capturing those expenses that reasonably and accurately reflect their actual R&D costs for this POR. As a result of their constantly changing of R&D methodologies, their latest method of capitalization of R&D produces a distorted and meaningless (for the cost of production exercise) result that does not reasonably reflect the actual cost of producing the subject merchandise.

    We have therefore determined that is appropriate to recognize for antidumping purposes all of Hyundai's and LG's 1997 R&D expenses in order to reasonably and accurately reflect their actual R&D costs for a given year. The Department also believes that, in general, recognizing the current year's R&D expenses is a reasonable method to recognize R&D expenses. This methodology is consistent with both Korean and U.S. GAAP, and is the same methodology that Hyundai and LG have been following for the past several years.

    Moreover, as Hyundai recognizes, the Department's practice is to consider, among other factors, international accounting standards for determining the reasonableness of a cost accounting methodology. While IAS No. 9 principally provides for the amortization of R&D expenses, IAS No. 9 also states that costs should be recognized as an expense on a systematic basis, which directly contradicts Hyundai's and LG's practice of continually changing how they recognize R&D expenses.

    We disagree with Hyundai that the SAA at 835 (on non-recurring costs) specifically supports Hyundai's argument that SAA prefers the amortization of R&D expenses. The SAA at 835 states that Commerce associates expenditures with all production benefitting from the expenditure, and gives R&D costs as an example of an expenditure that Commerce may allocate over current and future production. In some limited instances, consistent with the SAA at 835, it may be appropriate to allocate certain R&D costs for items that have alternative future uses (and benefits) over future production. The Department, in specific reference to the section of the SAA at issue, stated in the preamble to its final regulations that ``the allocation of nonrecurring costs, such as R&D costs, for purposes of computing COP and CV is dependent on case-specific factors.'' Antidumping Duties; Countervailing Duties; Final Rule, 62 FR 27296, 27362 (May 19, 1997) (``Final Rule''). Moreover, in its proposed rules, the Department, also in reference to the SAA at 835, specifically stated that:

    * * *there is no guarantee that * * * [R&D] * * * costs, if incurred to develop a new product or production process, would hold any future benefit to a company. To the contrary, after many months of costly research, a manufacturer could find its new product technologically useless due to the efforts of its competitors. In that case, the amounts incurred for R&D would not benefit the producer in terms of future product sales. Under these circumstances, the R&D expenditures must be recognized as a expense in the year incurred rather than amortized to some future periods.

    See Antidumping Duties; Countervailing Duties; Notice of Proposed Rule Making and Request for Public Comments, 61 FR 7308, 7342 (February 27, 1996) (``Proposed Rule'') (emphasis added).

    We also disagree with Hyundai that it has demonstrated, pursuant to the SAA, that it has historically utilized its new R&D accounting methodologies. While both Hyundai and LG previously amortized R&D, they have not done so consistently, and for the last several years have been expensing R&D currently. See Final Results 1996, DRAMs from Korea, 62 FR 965 (January 7, 1997) (``Final Results 1997 (I)'') (final results of second review), DRAMs from Korea, 62 FR 39809 (July 24, 1997) (``Final Results 1997 (II)'') (final results of third review), and Final Results 1998. Moreover, there is no evidence on the record that LG or Hyundai (prior to 1996) ever completely deferred R&D expenses. Additionally, the SAA at 834 is not, as Hyundai claims, directed at changes in depreciation, but only discusses depreciation as an example of how a company's records might not fairly allocate costs.

    We disagree with both Hyundai and LG that the Department's decision to reject their R&D accounting methodologies is contrary to the Micron I decision. First, in Micron I, the Court ruled that the Department ``failed to articulate a reasoned analysis justifying the departure from its established practice of amortizing those R&D expenses.'' See Micron I at 28. In contrast, in the present case, the Department has specifically articulated how amortizing and deferring R&D expenses is distortive. Second, the Department's methodology of expensing R&D is no longer a ``departure from its established practice.'' The Department's established practice is to expense semiconductor R&D currently. While the Department, prior to the Final Determination, in the cases cited by Hyundai and LG (i.e., CTVs from Korea, Pipe from Korea, and PET Film from Korea), allowed respondents to amortize R&D, the Department, for at least the last six years, and throughout the course of this proceeding, has constantly required that respondents recognize R&D expenses currently. See, e.g., Final Determination, 58 FR at 15472 (Department rejected amortization of R&D), Final Results 1996, Final Results 1997 (I), Final Results 1997 (II), and Final Results 1998 (Department accepted expensing of R&D currently). See also SRAMs from Korea, 63 FR 8934 (February 23, 1998) and SRAMs from Taiwan (Department accepted expensing of R&D currently) and Dynamic Random Access Memory Semiconductors of One Megabit or Above from Taiwan, 64 FR 56308, 56319 (October 19, 1999) (Department agreed that ``R&D costs should be expensed as incurred''). Third, the substance of the issues in Micron I and the present case are different. The Micron I decision concerned only the respondents' amortization of R&D expenses, while the present case also involves the respondents' practice of continually changing how they recognize R&D expenses.

    Hyundai's citation to Steel Wire Rod from Canada also is misplaced. Steel Wire Rod from Canada, 63 FR at 9187, concerned the amortization of certain costs relating to a furnace conversion, and not the amortization and deferral of R&D costs, as in the present case.

    We disagree with LG that the Department would never allow a company to change from expensing any cost to amortizing that cost because of the reduced cost recognized in the transition year. The Department evaluates any such change on a case by case basis. In the present case, as explained above, we found that the reduced R&D cost recognized by Hyundai and LG through the amortization and deferral of their R&D expenses, and resulting allocation of R&D expenses to merchandise, does not reasonably reflect the cost of producing the subject merchandise.

    Comment 2: Cross-Fertilization of R&D. Hyundai and LG argue that the Department erred in including the cost of R&D performed for non- memory and non-DRAM semiconductor products, respectively, in the cost of their DRAMs. Hyundai, citing section 773(e) of the Act, and LG, citing section 773(f)(1)(A) of the Act, argue that this decision violates the statutory requirements under the Act that Commerce calculates constructed value (``CV'') based on the production and general expenses related to the subject merchandise only. Hyundai, citing High-Tenacity Rayon Filament Yarn from Germany, 60 FR

    [[Page 69701]]

    15897, 15899 (March 28, 1995), Large Power Transformers from Japan, 57 FR 45767, 45768 (October 5, 1992), Sweaters Wholly or in Chief Weight of Man-Made Fiber from the Republic of Korea, 55 FR 32659, 32671 (August 10, 1990), Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof from France, et al., 61 FR 66472, 66491 (December 17, 1996)), Oil Country Tubular Goods from Argentina, 60 FR 33539, 33549 (June 28, 1995), High Information Content Flat Panel Displays and Display Glass Therefor from Japan, 56 FR 32376, 32386 (July 16, 1991), and Certain Small Business Telephone Systems and Subassemblies Thereof from Taiwan, 54 FR 42543, 42548 (October 17, 1989); and LG, additionally citing Lightweight Polyester Filament Fabrics from Korea, 48 FR 49,679, 49,681 (Oct. 27, 1983), Shop Towels from Bangladesh, 57 FR 3996, 3998-99(Feb. 3, 1992), Erasable Programmable Read Only Memories from Japan, 51 FR 39,680, 39,685 (Oct. 30, 1986) and Nippon Pillow Block Sales Co. v. United States, 17 CIT 276, 820 F. Supp. 1444 (1993); further argue that, consistent with the statute, the Department's long-standing practice has been to calculate the R&D expense component of CV on the most product-specific basis possible, and to exclude those R&D expenses that do not relate to the production of subject merchandise. LG, citing Cost Analysis Under the Antidumping Law, 21 George Wash. J. Int'l L. & Econ. at 389, notes that the Department has applied this practice without distinction to semiconductor cases, including 64K DRAMs From Japan, 51 FR 15,943, 15,948 (April 29, 1986)

    Hyundai and LG add that, in Micron I, 893 F. Supp. 21, 27, the CIT reversed the same decision to include non-subject R&D in the investigation of this case because it was unsupported by substantial evidence. Hyundai and LG contend that nothing has changed in this review that would justify a different result. Hyundai and LG maintain the Department has provided no factual support for its cross- fertilization theory that ``the subject merchandise benefits from R&D expenditures earmarked for non-subject merchandise.'' Hyundai and LG also maintain that the only cross-fertilization that may occur is that, for Hyundai, non-memory, and for LG, SRAM, semiconductor R&D may benefit from more advanced DRAM R&D.

    Hyundai and LG state the Memorandum from Dr. Murzy Jhabvala to U.S. Department of Commerce/Office of Antidumping Compliance regarding Cross-Fertilization of R&D of Semiconductor Memory Devices (``September 1997 Jhabvala Memo''), on file in the CRU, provides no substantial evidence to justify the Department's decision. Hyundai and LG contend that the September 1997 Jhabvala Memo is general and conclusory, and argue that the September 1997 Jhabvala Memo provides no evidence specific to their R&D and operations. Hyundai specifically states that its R&D for non-memory devices in its System IC Laboratory does not benefit memory devices because of the fundamental differences in function, design, and production between non-memory and memory products. LG specifically argues that it demonstrated to the Department at verification that its DRAM production operations do not derive any benefit from the R&D conducted for other products. LG also argues that Mr. Jhabvala's qualifications as set forth in his letter do not reveal any experience in DRAM R&D and production, and that other letters by actual experts on the record contradict his opinion.

    Micron argues that the Department properly accounted for the cross- fertilization of semiconductor R&D. Micron, citing to Final Results 1996, 61 FR at 20217-18, Final Results 1997 (2), 62 FR at 967, Final Results 1997 (3), 63 FR at 39823, Final Results 1998, 63 FR at 50870, SRAMs from Korea, 63 FR at 8938, SRAMs from Taiwan, 63 FR at 8925, and DRAMs from Taiwan, 64 FR at 56319, states that the Department has found that semiconductor industry R&D has a significant ``cross-fertilizing'' effect for R&D relating to all semiconductor products. Micron further argues that all of the respondents' arguments, including those relating to the Micron I decision, were rejected previously by the Department in Final Results 1998 and SRAMs from Korea.

    Micron also states that the September 1997 Jhabvala Memo, and information that Micron placed on the record, support the Department's finding on this matter. Micron also notes that the respondents have not included in their case briefs any direct citations to any expert opinion to support their arguments since the record evidence supports the Department's position.

    Micron contends that the Department has not departed from its practice, or its statutory obligations, in this issue. Rather, as the Department explained in SRAMs from Korea, 63 FR at 8940, the cost calculations in the present review are product-specific. Micron further points out that Hyundai, by proposing to allocate R&D on broad product lines, memory and non-memory, has acknowledged that its R&D expenses cannot be allocated on a model-specific or subject merchandise-specific basis. Moreover, Micron notes that the Department stated in Final Results 1998, 63 FR at 50870, that it is not bound by the way a company such as Hyundai categorizes its costs.

    DOC Position: We disagree with Hyundai and LG and have allocated all semiconductor R&D expenses over the total semiconductor cost of goods sold. This allocation methodology is fully consistent with the antidumping statute and the R&D calculations we have used throughout the Korean and Taiwan DRAMs and SRAMs proceedings.

    In SRAMs from Korea, we noted that, as a result of the forward- looking nature of R&D activities, we could not predict every instance where SRAM R&D may influence logic products or where logic R&D may influence SRAM products. As a result, we requested that Dr. Murzy Jhabvala, a semiconductor device engineer at the National Aeronautics and Space Administration with twenty-four years of experience, state his views regarding any potential overlap or cross-fertilization of R&D efforts in the semiconductor industry. In fact, Dr. Jhabvala had identified in another semiconductor proceeding before the Department areas where R&D from one type of semiconductor product influenced another semiconductor product. These statements, including the September 1997 Jhabvala Memo, are on the record of this review. In a statement prepared for the SRAMs Final Determination, Dr. Jhabvala stated that:

    SRAMs represent along with DRAMs the culmination of semiconductor research and development. Both families of devices have benefitted from the advances in photo lithographic techniques to print the fine geometries (the state-of-the-art steppers) required for the high density of transistors * * * In addition to achieve higher access speeds bipolar (ECL or TTL) output amplifiers are incorporated directly on chip with the CMOS SRAM memory array, a process known as BiCMOS. Further efforts to improve speed have resulted in the combination of the bipolar ECL technology with CMOS technology with silicon on insulator (SOI) technology.

    Clearly, three distinct areas of semiconductor technology are converging to benefit the SRAM device performance. There are other instances where previous technology and the efforts expended to develop that technology occurs in the SRAM technology. Some examples of these are the use of thin film transistors (TFTs) in SRAMs, advanced metal interconnect systems, anisotropic etching and filling techniques for trenching and planarization (CMP) and implant technology for retrograde wells.

    See September 8, 1997, Memorandum from Murzy Jhabvala to U.S. Department of Commerce regarding ``Cross Fertilization of Research and

    [[Page 69702]]

    Development of Semiconductor Memory Devices'' (``September 1997 Jhabvala Memo'').

    In SRAMs from Korea, we disagreed with Hyundai's contention that we must follow Hyundai's normal accounting records which categorize R&D expenses by project and product. See SRAMs from Korea, 63 FR at 8940. We disagree with similar contentions from LG and Hyundai in this review. As we have said in the past (see, e.g., Final Results 1998, 63 FR at 50870), we are not bound by the way a company categorizes its costs, R&D projects, or laboratory facilities, or by the company's accounting records that we review at verification if they do not reasonably reflect the costs attributable to production of the subject merchandise. Moreover, the mere fact that R&D projects for memory and non-memory products may be run in different laboratories, the fact that process and product research for memory and non-memory products may be distinguished, and the fact that each of the respondents may account for these R&D projects separately in their respective books and records, does not address the issue of cross-fertilization in semiconductor R&D. The existence of cross-fertilization in semiconductor R&D is the central theme of Dr. Jhabvala's many statements to the Department. Dr. Jhabvala offers various examples in those statements to illustrate that, regardless of the accounting or laboratory arrangements, the research results or developments in the processes and technologies used in the production and development of one semiconductor family can be (and are) used in the production and development of other semiconductor families. Dr. Jhabvala goes so far as to state that it would be ``unrealistic to expect researchers to work in complete technical isolation constantly reinventing technology that might already exist.'' See September 1997 Jhabvala Memo. Hyundai, in contrast to LG, does not contest the Department's position that all R&D for memory semiconductor projects, including SRAMs, benefits DRAMs. Given these facts, we do not believe that the reported expenses for DRAM R&D projects reasonably reflect the appropriate cost of producing the subject merchandise. As a result, we have continued to allocate all semiconductor R&D expenses over the total semiconductor cost of goods sold, a methodology which does not overstate costs, but which we believe reasonably and accurately identifies the R&D expenses attributable to subject merchandise. ‹SUP›1‹/SUP›

    \1\ In SRAMs from Korea, 63 FR at 8940, and SRAMs from Taiwan, 63 FR at 8925, we also disagreed with those same expert opinions regarding semiconductor R&D that LG submitted in this review, and for the reasons stated above, continue to disagree with those opinions.

    This methodology is not a change in the Department's approach to this issue. It is the Department's long-standing practice where costs benefit more than one product to allocate those costs to all the products which they benefit. See, e.g., SRAMs from Korea, 63 FR at 8940. This methodology results in the calculation of product-specific costs consistent with sections 773(e) and 773(f)(1)(A) of the Act because we have determined that DRAM-specific R&D account entries do not by themselves reflect all costs associated with the production and sale of subject merchandise.

    Comment 3: Level of Trade (``LOT'')/CEP Offset. Micron argues that in two recent decisions the CIT held that the Department must perform its LOT analysis based on unadjusted starting prices both for the U.S. sales used to calculate CEP as well as the home market sales used to calculate NV. See Borden, Inc. v. United States, 22 CIT __, 4 F. Supp. 2d 1221 (1998); Micron Technology, Inc. v. United States, 23 CIT __, 40 F. Supp. 2d 481, 485-86 (1999). In the Preliminary Results, the Department failed to do this, and instead analyzed the LOT of the home market sales based on the unadjusted starting prices of those sales, while analyzing the LOT of the U.S. sales based on the ``level of the constructed sale from the exporter to the [affiliated] importer,'' i.e., the prices after adjustment for U.S.-related selling expenses.

    Using this analysis, the Department determined, for both Hyundai and LG, that the home market and U.S. sales were made at different LOTs. In the absence of sales at more than one LOT in the comparison market, the Department found it could not quantify a LOT adjustment, and granted a CEP offset adjustment to each of the three respondents. See id. The Department declines to follow Borden on the grounds that it ``is not a final decision.'' See Level of Trade Memorandum, dated May 27, 1999. However, as the Borden and Micron decisions both establish, the Department's current practice is in conflict with the requirements of the statute. When the Department conducts a corrected LOT analysis, based on unadjusted starting prices in both the U.S. and the comparison markets, it will find that the comparison market sales made by Hyundai and LG were not made at a more advanced LOT than their sales in the United States, and therefore there is no basis for granting either a LOT adjustment or a CEP offset.

    Hyundai argues that the Department has ruled that Hyundai has been entitled to a CEP offset in each administrative review and argues that there are no factual reasons why the Department should reverse its long-standing practice. The Department has consistently ruled that the LOT of CEP sales must be based on the adjusted CEP price, not on the CEP starting price as advocated by Micron. See DRAMs from Taiwan 64 FR at 56313 (October 19, 1999). Hyundai argues that Petitioner's reliance on Borden is based on an incorrect interpretation of the statute. The court in Borden stated that the adjustments to CEP must be disregarded in defining the LOT of the CEP for purposes of the offset. However, the adjustments authorized under section 772(b) are an integral part of the definition of the statute and must be adhered to when determining the adjusted CEP price for comparisons and conducting the LOT analysis. Hyundai argues that the adoption of the court's reasoning in the Borden case would result in an unfair and distorted price comparison that is contrary to Congressional intent.

    Hyundai argues that it has established that there is a difference between the LOT in the home market and the CEP LOT. All of Hyundai's U.S. sales are on a CEP basis and its home market sales are at a more advanced stage of distribution than the CEP sales. Therefore, a CEP offset is appropriate under the provisions of the statute.

    LG asserts that the Department made a CEP offset correctly. LG also maintains that the Department should not apply the Borden case to the instant review. According to LG, the court held mistakenly that the Department's adjustments to CEP starting prices (by removing certain expenses) are inconsistent with section 773(a)(7) of the Act. LG claims that the court believed that such adjustments distort the LOT analysis and that this ``pre-adjustment'' creates an automatic CEP offset in addition to any CEP offset or LOT adjustment made after a comparison of adjusted CEP to HM price. LG contends that the Department's methodology does not create a ``pre-adjustment'' and correctly removes from the starting U.S. price only those expenses related to the resale transaction between the U.S. affiliate and the unaffiliated U.S. customer.

    DOC Position: The Department agrees with Hyundai and LG. We have

    [[Page 69703]]

    consistently stated that the statute and the SAA support analyzing the LOT of CEP sales at the constructed level after expenses associated with economic activities in the United States have been deducted, pursuant to section 772(d) of the Act. In the preamble to our proposed regulations, we stated:

    With respect to the identification of levels of trade, some commentators argued that, consistent with past practice, the Department should base level of trade on the starting price for both export price EP and CEP sales . . . The Department believes that this proposal is not supported by the SAA. If the starting price is used for all U.S. sales, the Department's ability to make meaningful comparisons at the same level of trade (or appropriate adjustments for differences in levels of trade) would be severely undermined in cases involving CEP sales. As noted by other commentators, using the starting price to determine the level of trade of both types of U.S. sales would result in a finding of different levels of trade for an EP sale and a CEP sale adjusted to a price that reflected the same selling functions. Accordingly, the regulations specify that the level of trade analyzed for EP sales is that of the starting price, and for CEP sales it is the constructed level of trade of the price after the deduction of U.S. selling expenses and profit.

    See Proposed Rule, 61 FR at 7347.

    Consistent with the above position, in those cases where a LOT comparison is warranted and possible, the Department normally evaluates the LOT for CEP sales based on the price after adjustments are made under section 772(d) of the Act. See, e.g., Large Newspaper Printing Presses and Components Thereof, Whether Assembled or Unassembled, From Japan: Notice of Final Determination of Sales at Less Than Fair Value, 61 FR 38139, 38143 (July 23, 1996). We note that, in every case decided under the revised antidumping statute, we have consistently adhered to this interpretation of the SAA and of the Act. See, e.g., Aramid Fiber Formed of Poly Para-Phenylene Terephthalamide from the Netherlands; Preliminary Results of Antidumping Duty Administrative Review, 61 FR 15766, 15768 (April 9, 1996); Certain Stainless Steel Wire Rods from France; Preliminary Result of Antidumping Duty Administrative Review, 61 FR 8915, 8916 (March 6, 1996); and Antifriction Bearings (Other Than Tapered Roller Bearings) and parts Thereof from France, et al., Preliminary Results of Antidumping Duty Administrative Review, 61 FR 25713, 35718-23 (July 8, 1996).

    In this case, in accordance with the above precedent, our instructions in the questionnaire issued to respondents stated that constructed LOT should be used. All respondents adequately documented the differences in selling functions in the home and in the U.S. markets. Therefore, the Department's decision to grant a CEP offset to Hyundai and LG was consistent with the statute and the Department's practice, and was supported by substantial evidence on the record.

    We disagree with the petitioner's interpretation of Borden and of its impact on our current practice. In Borden, the Court held that the Department's practice to base the LOT comparisons of CEP sales after CEP deductions is an impermissible interpretation of section 772(d) of the Act. See Borden, 4 F. Supp. 2d at 1236-38; see also Micron, 40 F. Supp. 2d at 485-86. The Department believes, however, that its practice is in full compliance with the statute, and that the court decision does not contain a persuasive statutory analysis. Because Borden is not a final and conclusive decision, the Department has continued to follow its normal practice of adjusting CEP under section 772(d) of the Act, prior to starting a LOT analysis, as articulated in the regulations at section 351.412. Accordingly, consistent with the Preliminary Determination, we will continue to analyze the LOT based on adjusted CEP prices, rather than the starting CEP prices.

    Comment 4: Exchange Rate Methodology. Hyundai and LG argue that the Department failed to consider the rapid decline in the value of the Korean won during the POR when it converted won to U.S. dollars. The respondents state that, in Policy Bulletin 96-1, the Department acknowledged the need to apply daily exchange rates in administrative reviews as well as investigations during periods of substantial exchange rate depreciation. The respondents further point out that in two recent administrative reviews, Steel Wire Rope from Korea, 63 FR 67662, 67665 (December 8, 1998), (unchanged at Steel Wire Rope from the Republic of Korea, 64 FR 17995 (April 13, 1999)) (``Steel Wire Rope from Korea''), and Certain Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products from Korea, 64 FR 48767, 48774 (September 8, 1999), and several investigations, specifically Stainless Steel Sheet and Strip in Coils from the Republic of Korea, 64 FR 30664, 30670 (June 8, 1999), Stainless Steel Round Wire from Korea, 64 FR 17342, 17343 (April 9, 1999), Stainless Steel Plate in Coils (``SSPC'') from the Republic of Korea, 64 FR 15444, 15446 (March 31, 1999), and Emulsion Styrene- Butadiene Rubber from the Republic of Korea, 64 FR 14865, 14867-8 (March 29, 1999) (``ESBR from Korea''), the Department applied the modified exchange rate methodology. The respondents contend that the same circumstances in these cases apply to this case because the POR includes the period (at the end of 1997) when the won lost over 40 percent of its value, and there is no reason why the Department should adopt any different treatment in this case. Hyundai specifically maintains that the Department should use daily won-dollar exchange rates for home market sales matched to U.S. sales occurring between November 1 and December 31, 1997; and should also use the average of the January 1 through February 28, 1998 exchange rate as a benchmark for U.S. sales occurring between those dates.

    Micron states that the Department properly applied its standard exchange rate methodology in the preliminary results of the review, and it should adhere to that standard methodology in the final results.

    DOC Position: We agree with the respondents, in part. Section 773A(a) of the Act directs the Department to use a daily exchange rate to convert foreign currencies into U.S. dollars unless the daily rate involves a fluctuation. The Department considers a ``fluctuation'' to exist when the daily exchange rate differs from the benchmark rate by 2.25 percent or more. The benchmark is defined as the moving average of rates for the past 40 business days. When we determine a fluctuation to have existed, we generally substitute the benchmark rate for the daily rate, in accordance with established practice. (An exception to this rule is described below.) (For an explanation of this method, see Policy Bulletin 96-1: Currency Conversions (61 FR 9434, March 8, 1996).)

    Our analysis of dollar-Korean-won exchange rates demonstrates that the Korean won declined rapidly in November and December 1997. Specifically, the won declined more than 40 percent over this two-month period. The decline was, in both speed and magnitude, many times more severe than any change in the dollar-won exchange rate during recent years, and it did not rebound significantly in a short time. As such, we determine that the decline in the won during November and December 1997 was of such magnitude that the dollar-won exchange rate cannot reasonably be viewed as having simply fluctuated at that time, i.e., as having experienced only a momentary drop in value relative to the normal benchmark. Accordingly, the Department used actual daily exchange rates exclusively in November and

    [[Page 69704]]

    December 1997. See Notice of Final Determination of Sales at Less Than Fair Value: Stainless Steel Sheet and Strip from the Republic of Korea, 64 FR 30664, 30670 (June 8, 1999) (``SSSS from Korea'').

    We note, however, that we have refined our methodology somewhat from that applied in SSSS from Korea. We recognize that, following a large and precipitous decline in the value of a currency, a period may exist wherein it is unclear whether further declines are a continuation of the large and precipitous decline or merely fluctuations. Under the circumstances of this case, such uncertainty may have existed following the large, precipitous drop in November and December 1997. Thus, we devised a methodology for identifying the point following a precipitous drop at which it is reasonable to presume that rates, more than 2.25 percent from the benchmark, were merely fluctuating. Following the precipitous drop in November and December 1997, we continued to use only actual daily rates until the daily rates were not more than 2.25 percent below the average of the 20 previous daily rates for five consecutive days. At that point, we determined that the pattern of daily rates no longer reasonably precluded the possibility that they were merely ``fluctuating''. Using a 20-day average for this purpose provides a reasonable indication that it is no longer necessary to refrain from using the normal methodology, while avoiding the use of daily rates exclusively for an excessive period of time. Accordingly, from the first of these five days, we resumed classifying daily rates as ``fluctuating'' or ``normal'' in accordance with our standard practice, except that we began with a 20-day benchmark and on each succeeding day added a daily rate to the average until the normal 40- day average was restored as the benchmark. See Notice of Final Results of Antidumping Duty Administrative Review: Certain Welded Carbon Steel Pipes and Tubes from Thailand, 64 FR 56759, 56763, October 21, 1999. See also Polyethylene Terephthalate Film, Sheet and Strip From Korea: Final Results of Antidumping Duty Administrative Review and Notice of Intent Not To Revoke in Part, 64 FR 62648, 62649 (November 17, 1999).

    Applying this methodology in the instant case, we used daily rates from November 3, 1997, through January 13, 1998. We then resumed the use of our normal methodology, starting with a benchmark based on the average of the 20 reported daily rates from January 14, 1998. We used the normal 40-day benchmark from February 12, 1998, to the close of the review period.

    Comment 5: Duty Absorption. Hyundai contends that the finding of duty absorption is null and void because the Department had no authority to conduct a duty absorption inquiry in this administrative review. Hyundai maintains that section 751(a)(4) of the Act ``explicitly limits'' duty absorption inquiries to administrative reviews initiated 2 years or 4 years after the publication of an antidumping duty order, and that this review was initiated on June 29, 1998, five years after publication of the antidumping duty order in this case. Hyundai also maintains that section 351.213(j)(2) of the Department's regulations, which provides for the Department to conduct duty absorption inquiries for transition orders (as defined in section 751(c)(6) of the Act) in reviews initiated in 1996 or 1998, cannot authorize the conduct of a duty absorption inquiry. Hyundai states that this regulation is ``directly contradicted'' by section 751(a)(4) of the Act, which makes no exception for transition orders.

    Hyundai further argues that section 751(c)(6) of the Act, which defines transition orders, only applies to section 751(c) of the Act, which establishes procedures for the conduct of sunset reviews. Hyundai states that section 751(c)(6) of the Act has no relationship to administrative reviews conducted under section 751(a) of the Act, nor to duty absorption inquiries conducted under section 751(a)(4) of the Act.

    LG argues that the Department may not lawfully presume that duties have been absorbed by LG without record evidence to support this conclusion. LG, citing to Extruded Rubber Thread from Malaysia, 63 FR 2752, 2757 (March 16, 1998) and Report to the House and Senate Appropriations Committees: The Efficacy of Antidumping Measures in Related Importer Situations (January 30, 1998), at 3, states that Department presumes that absorption is occurring where a dumping margin is found unless the U.S. affiliate's customers have promised in writing to pay any antidumping duties imposed on the merchandise. LG further states, citing to Id. at 4 and Certain Cut-to-Length Carbon Steel Plate from Belgium, 63 FR 2959, 2963-64 (January 20, 1998), that the Department has never found an instance of such a written agreement that is acceptable. LG argues that it defies commercial reality to expect a customer to agree to assume such a liability for antidumping duties, and the Department's establishment of an ``effectively irrebuttable'' presumption that duties are being absorbed ``makes a mockery'' of the duty absorption inquiry entrusted to the Department by Congress.

    Micron argues that the Department correctly made a finding of duty absorption for Hyundai and LG. Micron notes that both respondents imported the subject merchandise through their affiliated U.S. importers, and therefore, the antidumping duties assessed as a result of this review will be paid, in the first instance, by those affiliated importers. Micron also points out that the Department provided Hyundai and LG with an opportunity to submit evidence that unaffiliated purchasers will pay the antidumping duties to be assessed on entries during the review period, and neither party submitted such evidence. Micron maintains that, in the absence of any evidence, and in the light of the ``commercial reality'' noted by LG under which no unaffiliated customer would assume the liability for these assessments, the Department can only reasonably conclude, based on record evidence, that duties will be absorbed by Hyundai and LG.

    Micron argues that since LG explicitly declined to submit any evidence on this matter, it cannot now argue that the Department is employing an ``inappropriately high evidentiary standard.'' Micron also states, in reference to Hyundai's arguments, that section 751(a)(4) of the Act does not limit the Department's authority to make a duty absorption inquiry in the context of administrative reviews conducted in those years not referenced by this section. Micron further argues that, since Hyundai does not allege any harm arising out of the Department's conduct of this inquiry, it has no standing to complain that the Department's conduct is ultra vires.

    Micron contends that the Department explained in Final Regulations, 62 FR at 27317-18, that its interpretation of section 751(a)(4) of the Act, under section 351.213(j)(2) of the Department's regulations, is necessary to carry out the legislative intent of the statute, i.e., to provide the relevant information to the International Trade Commission (``ITC'') in connection with its conduct of sunset reviews. Micron further contends that if the Department had adopted Hyundai's interpretation of the Act, then the Department would have had the option of conducting a duty absorption inquiry, pursuant to Micron's July 21, 1997 request, in the fourth review of this proceeding.

    DOC Position: We agree with Micron. With regard to the time frame in which we are conducting this review, section

    [[Page 69705]]

    351.213(j)(1) of our regulations, in accordance with section 751(a)(4) of the Act, provides for the conduct, upon request, of absorption inquiries in reviews initiated two and four years after the publication of an antidumping duty order. With respect to transition orders, the preamble to the proposed antidumping regulations explains that reviews initiated in 1996 will be considered initiated in the second year, and reviews initiated in 1998 will be considered initiated in the fourth year. Notice of Proposed Rulemaking and Request for Public Comments, 61 FR 7308, 7317 (February 27, 1996). Because this order has been in effect since 1993, this is a transition order in accordance with section 751(c)(6)(C) of the Act. This being a review initiated in 1998 and a request having been made, we have made a duty absorption determination as part of this administrative review.

    We believe that Congress intended that the ITC would consider the issue of duty absorption in all sunset reviews. In this regard, the statutory provision requiring the consideration of duty absorption does not distinguish between antidumping orders issued after January 1, 1995, and transition orders. See section 752(a)(1)(D) of the Act. Moreover, in all of the legislative history, Congress explained the implications of affirmative duty-absorption findings and clearly contemplated that such findings would be considered in all sunset reviews. See S. Rep.103-412 at 50 (1994). See also H. Rep. 103-826 at 60-61 (1994) (``Commerce will inform the Commission of its findings regarding duty absorption, and the Commission will take such findings into account in determining whether injury is likely to continue or recur if an order were revoked''). Thus, we have made a duty absorption determination as part of this administrative review. See Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, Germany, Italy, Japan, Romania, Sweden, and the United Kingdom: Final Results of Antidumping Duty Administrative Reviews, 64 FR 35590, 35601 (July 1, 1999) (``AFBs'').

    In considering methodologies that might be used for a duty absorption inquiry, the Department sought to adopt one that would comply with the statute, as well as one that would be administrable within the time frame of a review period and still provide respondents with a sufficient opportunity to cure any deficiencies. The method the Department adopted accomplishes these goals. As the Department explained in AFBs, 64 FR at 35601, the ``existence of a margin raises an initial presumption that the respondent and its affiliated importer(s) are absorbing the duty.'' This is a reasonable presumption because the continued existence of dumping margins indicates that the producer and its affiliated U.S. importer have not adjusted their prices to eliminate dumping. If the producer has not set its price to the first unaffiliated U.S. customer high enough to eliminate dumping and the affiliated importer is liable for payment of the antidumping duties, it is reasonable to presume that the producer is absorbing the antidumping duties. The reasonableness of this presumption is also reflected in the SAA at 885, which states that ``the affiliated importer may choose to pay the antidumping duty rather than eliminate the dumping.'' In sum, the existence of dumping gives rise to a reasonable presumption that the affiliated importer is absorbing dumping duties.

    As in previous cases where the Department has found duty absorption (see, e.g., AFBs), this is an instance where the existence of a margin raises an initial presumption that the respondent and its affiliated importer(s) are absorbing the duty. As such, the burden of producing evidence to the contrary shifts to the respondent. See Creswell Trading Co., Inc. v. United States, 15 F.3d 1054 (CAFC 1994). Here the respondents have not placed evidence on the record, despite being given ample time to do so, in support of their position that they and their affiliated importer(s) are not absorbing the duties. In fact, as noted by Micron, LG explicitly refused to do so. Therefore, because Hyundai and LG submitted no information showing that their respective affiliated importer is not absorbing the duties for this POR, we find that duty absorption occurred.

    Comment 6: Cash Deposit Rate. Micron argues that the Department should establish a single cash deposit rate for both Hyundai and LG in the final results of this review, at a minimum, by weight averaging the combined dumping margins for Hyundai and LG. Micron asserts that, at the end of the current POR, LG was acquired by Hyundai and renamed Hyundai Microelectronics, and based on comments in a letter submitted by LG, the merger is to be completed in October 1999. Micron contends that one company will control both Hyundai and LG's fabrication facilities, and that company could choose to designate all of its exports to the United States as being from whichever respondent in the fifth review turns out to have the lower duty deposit rate.

    Hyundai argues that it is entitled to its own cash deposit rate based exclusively on the Department's calculations for Hyundai. Hyundai contends that the single rate advocated by Micron would incorporate any adverse FA margin that the Department might impose on LG, and Hyundai should not be penalized for any action taken with respect to LG. Hyundai maintains that, whatever decisions the Department may make concerning LG, Hyundai had no involvement in the matters alleged. Hyundai states that there has been no indication of any diversion or unreported sales by Hyundai in the six consecutive times that Hyundai has been verified since this case began.

    Hyundai further argues that the Department must disregard Micron's argument because it relates to ``matters that allegedly might occur well beyond'' the end of the current POR. Hyundai contends that Micron's speculation as to what might happen after the merger has no support on the record. Hyundai also contends that if one company controls both companies' fabrication facilities, the cash deposit rate would be the rate that applies to the controlling company, regardless of which fab produced the DRAMs or how the company chose to ``designate'' the origin of the exports.

    DOC Position: We are concerned about the implications of the pending merger of Hyundai and LG on the efficacy of the antidumping duty order on DRAMs from Korea. However, pursuant to Micron's November 12, 1999 request, we initiated a changed circumstances review under section 751(b) of the Act to address the cash deposit issue. Because we have initiated a separate segment of this proceeding to address the cash deposit issue, we will continue to issue Hyundai and LG their own cash deposit rates for these final results.

    Comment 7: All Others Rate. Micron claims that the Department should correct the ``all others'' rate to reflect the revised 4.55 percent rate calculated by the Department following judicial review of the original less than fair value (``LTFV'') determination. See Micron Technology, Inc. v. United States, 117 F.3d 1386 (Fed. Cir. 1997). In its notice of the preliminary results of review the Department states that the ``all others'' cash deposit rate ``will be 3.85 percent, the `all others' rate established in the LTFV investigation.'' Preliminary Results, 64 FR at 30486. Therefore, the Department's revised ``all others'' rate of 4.55 percent has become final and should be reflected in these final results.

    [[Page 69706]]

    No rebuttal briefs were filedin regards to this issue.

    DOC Position: We agree with the petitioner and have corrected the ``all others'' rate to reflect the revised 4.55 percent rate calculated by the Department following judicial review of the original LTFV investigation.

    Company-Specific Issues

    1. Hyundai

      Comment 1: Use of Cost of Goods Sold to Calculate R&D Ratio. Hyundai states that the Department greatly overstated the per unit R&D costs allocated to each product by calculating Hyundai's R&D ratio as a percentage of cost of goods sold (``COGS'') rather than as a percentage of cost of manufacturing (``COM''). Hyundai, citing to High Information Content Flat Panel Displays and Display Glass Therefor from Japan, 56 FR 32376, 32382 (July 16, 1991), contends that since the R&D ratio is applied to the COM, the denominator for the R&D ratio should also be the COM. Hyundai points out that although the Department has used COGS as the denominator for the R&D ratio in other proceedings including DRAMs, it has not stated any reason why COGS is a more acceptable denominator than COM.

      Hyundai maintains that it is inconsistent to apply a COGS-based percentage to calculate Hyundai's R&D cost since R&D is considered by the Department as an element of the COM, and the Department applies the R&D ratio to the total COM of each product. Hyundai notes that the Department used to classify certain R&D costs as G&A expenses, but now classifies all R&D costs as manufacturing costs. Hyundai states that, in contrast to R&D expenses, G&A expenses, which support both sales and production, could reasonably be compared to COGS to calculate the G&A ratio.

      Hyundai, citing the Department's Antidumping Manual at 48, and DRAMs from Taiwan, 64 FR at 56312, further notes that, in the G&A ratio calculation, the Department adjusts COGS to make it equivalent to COM in order to reflect the same category of costs as the per unit COM to which this ratio is applied. Hyundai argues that if it is appropriate to adjust the COGS, when COGS is used as the denominator for a ratio calculation, to align it more closely to the COM, then it is accurate and appropriate to use COM itself as the denominator, which has been calculated in the same manner as the per unit COM of each product. Hyundai also contends that the Department's minor distinction between COGS and COM indicate that the use of COGS as the denominator is merely for administrative convenience.

      Hyundai further argues that the propriety of the Department's practice of using COGS to represent COM depends entirely on the presumption that the COGS during a period is a reasonably close approximation of the COM. Hyundai contends that, in the DRAM industry, which has a consistent trend toward higher density products and strong learning curve effects on costs, this presumption does not hold. Hyundai explains that, during a period of ``rapid generational progress,'' the cost of (new generation) DRAMs that are produced during this period is higher than the cost of (old generation) DRAMs that are sold during this period. Hyundai states that consequently, when total current R&D expenses are calculated as a percentage of COGS rather than as a percentage of COM, the R&D ratio is inflated, and Hyundai's total R&D costs are overstated.

      Finally, Hyundai argues that the application of the Department's standard cost-of-production (``COP'') completeness test demonstrates that the use of COGS as the denominator in the R&D ratio calculation significantly overstates Hyundai's R&D expenses. Hyundai states that multiplying Hyundai's total semiconductor COM by the Department's calculated R&D ratio results in a total R&D expense that greatly exceeds Hyundai's actual R&D expense.

      Micron argues that the Department was correct to use the COGS instead of COM for its R&D calculation, as is the Department's practice. Micron contends that Hyundai never complained before about the Department's use of COGS instead of COM in its R&D calculation ratio, and is only complaining now because of the difference between its COGS and COM figures. Micron also contends that the difference in these figures is not due, as Hyundai claims, to the change in the density of the DRAMs that it produced, but is due to a certain proprietary item in its cost accounting records.

      DOC Position: We agree with the petitioner. The Department's practice, as we have carried out throughout this proceeding, is to calculate the R&D ratio by dividing a respondent's R&D expense by the respondent's COGS. See e.g., Final Determination, 58 FR at 15470, Final Results 1997 (I), 62 FR at 967, Final Results 1997 (II), 62 FR at 39823, and Final Results 1998, 63 FR 50870. See also DRAMs from Taiwan, 64 FR at 56312. We calculate this ratio based on the COGS, or a modified COGS, and not the COM, because R&D expenses, like G&A expenses, are incurred for the products sold during a period, rather than the products manufactured during a period. Furthermore, we believe that evaluating whether to use COGS or COM as the denominator in the R&D ratio from one review segment to the next would eliminate significant consistency and predictability in our calculations.

      We also agree with the petitioner that the record does not support Hyundai's assertions that the difference between its COGS and COM is due to the change in the densities that it produces. Rather, this difference is due, in part, to the proprietary accounting item referenced by Micron.

      Comment 2: Double-Counting of R&D. Hyundai argues that the Department double-counted certain R&D expenses incurred by Hyundai Electronics America, Inc. (``Hyundai America''). Hyundai states that the Department included Hyundai America's actual expense for certain R&D projects, as well the amount that Hyundai paid to Hyundai America to reimburse it for these expenses. Hyundai explains that it did not itself offset these expenses in its response because these expenses were classified as long-term R&D, and not included by Hyundai in the current year R&D calculation. Hyundai states that the double-counting occurred when the Department included all of Hyundai's expenditures on long-term projects in current year production costs. Consequently, Hyundai maintains that if the Department decides to continue expensing all of Hyundai's R&D expenses incurred in 1997, then the Department should either exclude the expenses at issue from Hyundai's R&D costs, or offset Hyundai America's R&D expenditures, to avoid double-counting.

      Micron did not comment on this issue.

      DOC Position: We disagree with Hyundai that the record evidence demonstrates that we double-counted certain R&D expenses incurred by Hyundai America, and reimbursed by Hyundai. Hyundai America's financial statement indicates that Hyundai America received revenue from Hyundai for R&D services (see exhibit 17 of Hyundai's October 8, 1998 section A response). However, the record evidence does not demonstrate that Hyundai's R&D expenses include any payments to Hyundai America. Therefore, we cannot confirm that any R&D expense has been double- counted and have made no changes in our calculations with respect to this issue from our preliminary results.

      Comment 3: Offset for Long-Term Interest Income. Hyundai argues that the Department improperly denied an offset

      [[Page 69707]]

      for long-term interest income earned on restricted bank deposits. Hyundai states that the interest income at issue was earned by Hyundai on (1) collateral that Hyundai is required to maintain on deposit at banks in order to obtain loans to finance current operations; and (2) deposits with insurance companies that can only be used to pay retirement benefits. Hyundai states that, in both cases, the income is not derived from long-term investments, but is directly tied to the current operations of the company. Citing to Final Determination, 58 FR at 15473, and The Timken Company. v. United States, 809 F. Supp. 121, at 125 (1992), Hyundai contends that the Department grants adjustments for interest income that is earned on compensating deposits because such income is related to current operations; and that the Department previously granted Hyundai an adjustment for interest income related to current operations. Hyundai maintains that, since the income derived from these deposits reduces the cost of the related loans, the interest earned from those deposits should be used to offset Hyundai's interest expense.

      Hyundai explains that some of the deposits at issue are not investments, but a prerequisite for receiving loans from the Korea Development Bank. Hyundai states that the use of compensating deposits enabled Hyundai to receive a lower effective interest rate from the banks, and thus directly affected the interest expense that Hyundai incurred for financing current operations during the POR. Hyundai maintains that, since these deposits are an integral part of the relevant loans, and since the Department considers all financing costs to be related to current operations, then the interest earned on the deposits is directly related to current operations, regardless of the period of time over which the deposits are maintained.

      Hyundai further explains that the other deposits at issue are held at life insurance companies to fund accrued severance benefits. Hyundai states that it makes these deposits in order to claim the total amount of severance benefits as a tax-deductible expense. Hyundai contends that, since the severance benefits themselves are included in the labor expense element of Hyundai's COM, the income earned from these deposits of severance benefits is directly related to current operations.

      Hyundai, citing to Gulf States Tube Div. Of Quanex Corp. v. United States, 981 F. Supp. 630, 643 (CIT 1997) and Recent Stitches in the Department of Commerce's Cost of Production Analysis: MMF Sweaters Antidumping Case and Commerce's Treatment of Interest Expense, 25 George Wash. J. Int'l L (1991), also contends that the Department allows respondents an offset for interest income on long-term deposits that are related to current operations. Hyundai also notes that, in fact, the Department has granted offsets for long-term ``compensating'' deposits because such deposits are related to the cost of borrowing funds for current operations.

      Micron argues that the Department should not make any adjustment for Hyundai's claimed offset for interest income. Micron, citing to Final Determination, 58 FR at 15473, contends that the Department only previously granted Hyundai an offset for interest income earned on short-term assets. Micron further maintains that a respondent must show at verification that deposits are compensating balances tied to loans in order to receive an offset for interest earned on such deposits, and that the verification report and preliminary calculation memorandum indicate that the Department was not satisfied that Hyundai had shown this.

      Micron also contends that the severance insurance deposits are not connected to loans at all, but represent Hyundai's funding of accrued severance benefits. Micron states that the classification of the insurance balance as a restricted deposit does not qualify it as a compensating balance (for a loan). Micron concludes that Hyundai has not demonstrated that the interest earned on the insurance deposit is in any way tied to interest expense, and that the Department should continue to exclude the claimed interest income from the offset to Hyundai's interest expense rate.

      DOC Position: We agree with Hyundai in part, and with Micron in part. Upon further review of cost verification exhibit 2, and Hyundai's 1997 consolidated financial statement, which specifically mentions that ``certain bank deposits are pledged as collateral for long-term debt,'' we agree with Hyundai that its long-term restricted deposits at issue with the Korea Development Bank are an integral part of certain loans from that Bank. Since the income derived from these deposits are directly related to specific loans, the interest earned from those deposits should be used to offset Hyundai's interest expense for the same loans. Additionally, we agree with Micron that the severance insurance deposits are long-term investments that represent Hyundai's funding of accrued severance benefits. These severance insurance deposits are simply a source of funds from which Hyundai funds severance benefits, and are only held by Hyundai as restricted deposits to allow Hyundai to claim a tax deduction. Accordingly, we have only offset the interest from the deposits at issue with the Korean Development Bank against Hyundai's interest expense.

    2. LG

      Comment 1: LG's Knowledge of U.S. Sales: Mexico. LG asserts that, in its relationship with a Mexican client, it requested a variety of proof of delivery (``POD's'') to determine that the DRAMs were actually destined for Mexico. LG states that it believed reasonably and in good faith that this customer was a legitimate third country purchaser of LG DRAMs with ample ability to consume them and to market LG's products in Mexico, Latin America, as well as in other third country markets.

      LG alleges that taken together, the facts on the record demonstrate that it was the unsuspecting victim of Customs fraud, including the alteration and falsification of LG invoices, and the unlawful diversion of LG products into the United States. LG claims that there is overwhelming evidence that it did not have knowledge. Thus, LG contends that the Department has no lawful basis to attribute the illegally diverted shipments to LG and to include such shipments as ``unreported U.S. sales'' in the calculation of LG's dumping margin.

      To support its argument, LG cites Tapered Roller Bearings from China, where the petitioner in the case argued that the Department should reclassify third country transactions, placed by a U.S. firm, as U.S. sales. The Department disagreed and considered these transactions as sales to a third country, stating ``the Act requires that the producer of the merchandise know, at the time of sale to the reseller, the country to which the reseller intends to export the merchandise in order for the Department to treat sales to a reseller as sales to the United States.'' LG points out that the Department made no inquiry as to whether the respondent ``should have known'' that the goods were destined for U.S. consumption.

      According to LG, the Department applies its knowledge test to a respondent at the time of sale, not later. For example, in Notice of Final Determination of Sales at Less Than Fair Value: Manganese Sulfate From the People's Republic of China, 60 FR 51255 (1995) (``Manganese Sulfate''), the Department determined that a transaction was not a U.S. sale where the respondent learned that the merchandise it sold to a third country trading company was ultimately destined for the United States ``at the time of shipment, after the sale had already been made.'' Similarly, in Pure

      [[Page 69708]]

      Magnesium from the Russian Federation (``Magnesium''), the Department treated certain sales as third country exports, even though the respondent later learned that some of these exports were sold to U.S. customers, because ``this knowledge always came after (the respondent) had sold the merchandise.'' LG maintains that in neither of these cases did the Department suggest that the sales could have been deemed U.S. sales by the respondent if the respondent ``should have known'' the ultimate destination was the United States, notwithstanding evidence on the record from which such a conclusion might have been drawn.

      According to LG, the CIT, in NSK Ltd. v. United States (``NSK''), explicitly confirmed that the antidumping law mandates the knowledge test as applied by the Department in these prior cases. In NSK, the Department classified Honda as a reseller after concluding that Honda's Japanese suppliers were unaware of the ultimate destination of the merchandise they sold to Honda. The Court agreed, emphasizing the statutory language and legislative history upon which the Department's knowledge test is based. Specifically, the Court quoted the portion of the statute which states that U.S. purchase price is ``the price at which merchandise is purchased, or agreed to be purchased, prior to the date of importation, from a reseller or the manufacturer or producer of the merchandise for exportation to the United States.'' The Court cited the legislative history to this provision:

      If a producer knew that the merchandise was intended for sale to an unrelated purchaser in the United States under terms of sale fixed on or before the date of importation, the producer's sale price to an unrelated middleman will be used as the purchase price.

      LG states that the NSK Court acknowledged that the Department's knowledge test requires a ``high standard'' which could possibly be exploited by ``the `perfect' scenario, where a reseller hides the ultimate destination of its purchases from its foreign suppliers,'' but found nevertheless that ``such a standard is necessary to fulfill the statutory intent that purchase price be based on sales of goods sold abroad with the intent of being exported to the U.S.'' According to LG, both the stress on the word ``knew'' in the quoted legislative history and the Court's emphasis on the ``intent'' of the seller make clear that the U.S. sales may not be attributed to the foreign producer if the producer did not actually know that the United States was their destination, even if in retrospect it appears to an outside observer that the producer should have known that the goods would reach the United States. The Court ruled that this sort of generalized suspicion is not sufficient: ``the suppliers must have knowledge that particular sales are destined for import into the U.S.''

      LG asserts that, in INA Walzlager-Schaeffler KG v. United States (``INA''), the CIT noted again the strict requirement for particularized knowledge before U.S. sales may be attributed to a foreign producer. After reviewing the Department's ``knew or should have known'' test, the Court reiterated that this test applies only under the NV section of the law:

      This decision is not intended to alter the standard for imputed knowledge pursuant to 19 U.S.C. 1677a(b). As both FAG and Commerce acknowledge, section 1677a(b) requires knowledge that the merchandise is purchased from a reseller for exportation to the United States * * *. Under 19 U.S.C. 1677b(a) it is not necessary for the respondent to have knowledge that all of the merchandise sold is destined for the United States in order to impute knowledge that the sales were not intended for home consumption.

      LG states that, in Tapered Roller Bearings and Parts Thereof from China (``TRBs II''), the Department had occasion to consider the ``perfect'' scenario envisioned by the Court in NSK. In TRBs II, the petitioner argued that suppliers ``knew or had reason to know'' that sales to a reseller were destined for the United States because TRBs sold to the U.S. market were all identified with the supplier's trade name, constituting ``sufficient evidence on the record for the Department to impute knowledge on behalf of the suppliers.'' In response, the reseller argued that ``NSK requires the Department to find evidence of actual knowledge that particular sales were destined for importation into the United States.'' In response to these arguments, the Department held:

      Lacking evidence of actual knowledge that particular sales were destined for the United States, we cannot assume such knowledge, regardless of general knowledge that some merchandise was intended for exportation to the United States.

      LG argues that the Department in TRBs II recognized that even when a reseller hides the ultimate destination of its purchases from its foreign supplier, and there is no evidence that the foreign supplier had actual knowledge that particular sales were destined for the United States, the Department may not attribute the resulting U.S. sales to the supplier, regardless of whether the supplier ``should have known'' or ``had reason to know'' that the goods would be resold into the United States.

      Finally, in Certain Cut-to-Length Carbon Steel Plate from Ukraine, 62 FR 61,754, 61,760 (Nov. 19, 1997) (``Ukraine Plate''), the Department excluded diverted sales where the respondent producer, Ilyich, argued that it made the sales ``believing they were destined to third countries and had no knowledge that these sales were ultimately destined for the United States.'' The Department stated ``[i]t is the Department's practice to include as U.S. sales only those sales known by the producer/exporter to be destined for the United States at the time of sale and delivery'' and determined that ``these originally non- U.S. bound shipments were delivered to the U.S. without prior knowledge of Ilyich. Therefore, consistent with * * * Department practice, we have not included the pirated sales in the final margin calculation.''

      LG states that the Department cited Yue Pak, Ltd. v. United States, Slip Op. 96-65 at 9 (``CIT''), aff'd. 1997 U.S. App. LEXIS 5425 (Fed. Cir. 1997) (``Yue Pak'') in the fourth review final results in support of its contention that U.S. resales may be charged to a supplier that did not know, but should have known, that the United States was the ultimate destination of its shipments. But, according to LG, Yue Pak fails to support the Department's legal theory. First, no party in that case argued that under the antidumping law ``should have known'' is not a sufficient standard for attribution of U.S. sales, and the Court was thus not presented with the relevant issue. Second, from the facts cited by the Court, it is apparent that the record evidence established that the suppliers knew that the merchandise was destined for the United States; the Court's references to whether the suppliers should have known are thus dicta. Third, neither of the cases cited by the Court for its interpretation of section 772(b) of the Act, 19 U.S.C. 1677a(b), see 20 CIT at 498, support the ``should have known'' theory; rather, both cases clearly state that the supplier must know that the merchandise is destined for the United States. Finally, to the extent that Yue Pak is contrary, it has been overruled by the CIT's later disposition of this issue in NSK and INA, both of which make clear that section 772(b) of the Act, 19 U.S.C. 1677a(b) and its legislative history allow U.S. sales to be charged to a producer only ``[i]f a producer knew that the merchandise was intended for sale to an unrelated purchaser in the United States.''

      [[Page 69709]]

      In sum, LG asserts that the CIT has clearly and consistently held, in line with the consistent practice of the Department, that the Department may treat third country sales as U.S. sales of a producer only if the producer knew, at the time of sale, that those particular sales were destined for the United States. LG contends that there is no evidence on the record that they actually knew that its sales were destined for the United States. Thus, it was unlawful for the Department in the preliminary results to consider the diverted shipments to have been U.S. sales of LG.

      LG argues that the circumstances of the sales to its Mexican customer did not provide LG with knowledge that the diverted sales were destined for importation into the United States. According to LG, in the fourth review final results, as in other cases cited above, the Department suggested that the diverted transactions be deemed U.S. sales of LG, regardless of whether LG knew the shipments' final destination, because LG sometimes dealt directly with the customer's U.S. parent company and because it arranged for the DRAMs to be shipped in bond through the United States. This argument, LG maintains, is contrary to numerous Department precedents.

      LG contends that the shipment route chosen by LG's customer is simply not relevant to the question of whether LG knew the ultimate destination of the merchandise. According to LG, the relevant inquiry is what LG knew at the time of sale about the goods' final destination, not what route the goods traveled to get there.

      LG asserts that the Department has many precedents on this issue which uniformly hold that dealings with a U.S. company or a shipment route through the United States do not transform a third country sale into a U.S. sale. LG maintains that, in Magnesium, the Department ``found nothing to indicate any unreported instances of merchandise being sold with the knowledge at the time of sale that the ultimate destination was the United States,'' and refused to reclassify as U.S. sales certain third-country sales with purchase orders placed by a U.S. firm, determining that purchase orders placed by a U.S. company do not constitute evidence that the respondent had knowledge the sale was destined for the United States. Likewise, in Manganese Sulfate from China, the Department determined that a transaction was not a U.S. sale even where (1) the bill of lading listed the destination as a U.S. port; (2) PRC Customs export statistics' printout of exports to the United States showed that this shipment was sent to the United States; and (3) correspondence from a New York company regarding this shipment was dated before the issuance of the sales contract.

      LG argues, contrary to the Department's conclusion that LG's sales to the Mexican customer ``were shipped by LG directly to the United States,'' that the undisputed facts on the record show that the customer's purchases from LG were transported by its freight forwarder from arrival in the United States in bond, just as the Mexican customer had agreed. LG asserts that the Department is prohibited from treating goods transported in bond as U.S. sales for purposes of the antidumping law because in-bond entries do not enter the Customs territory of the United States for consumption. LG maintains that in Titanium Metals Corp. v. United States, 19 CIT 1143, 901 F. Supp. 362, 364 (1995) (``Titanium Sponge''), the CIT has explicitly confirmed this point, holding that goods entered for transportation in bond are not entered for consumption, and thus cannot be included in the dumping margin calculation, because the law restricts the assessment of antidumping duties to merchandise entered or withdrawn from warehouse for consumption.

      Further, LG claims that the Department's interpretation of Persulfates from the PRC (``Persulfates'') is wrong and entirely contradicts the proposition that sales may not be considered to be U.S. sales by the supplier unless (i) the supplier had actual knowledge that the goods would be resold to the United States, and (ii) the goods actually entered the Customs territory of the United States.

      LG argues what dictated the result of Persulfates was not that the producer shipped the goods to the United States, but that the producer shipped the goods to the United States knowing that the customer planned to enter the goods into the United States, rather than ship them in bond to a third country. LG further asserts that the Department has described its decision in Persulfates as turning on the producer's knowledge, stating that ``in cases where evidence exists that a supplier had knowledge that the ultimate destination of the merchandise was the United States, such as * * * Persulfates, we have considered the sale by the supplier to the reseller as the starting price in our margin calculations.'' In direct contrast to Persulfates, LG claims that LG's customer was outside the United States, and LG's products were shipped to a bonded area, for further in-transit bonded shipment to a third country, and, so far as LG ever knew, the goods did not enter into the Customs territory of the United States. LG concludes that Persulfates thus squarely contradicts the Department's suggestion that the route of the shipments from LG to its third country customer renders irrelevant the question of whether LG knew that the goods would enter the United States.

      LG also argues that the Department should correct errors committed in its calculation of the dumping margin on the diverted sales. LG contends that the Department may not apply adverse FA ignoring the timely data submitted by LG concerning the unlawfully diverted shipments. LG contends that it submitted timely expense data concerning the Mexican sales, which the Department chose not to accept, instead opting for adverse FA to determine the selling expenses for the diverted sales. LG argues that the Department's action is an abuse of discretion and must be reversed in the final results.

      LG cites Olympic Adhesives, Inc. v. United States (``Olympic Adhesives''), where the Court ruled that if a company was sent and completely answered repeated questionnaires, and nothing in the record suggests that the company withheld information, the Department is not allowed to use best information available. LG additionally cites Ferro Union, Inc. v. United States and Borden, Inc. v. United States as particular examples where the Department must implement a narrower interpretation of when to use FA.

      LG contends that the Department's rejection of LG's reported expenses in conjunction with its sales to the Mexican customer cannot be justified and is clearly unlawful under the standard of Olympic Adhesives. LG contends that the Department has failed to show that LG withheld requested information, failed to provide information by the applicable deadlines or in the form and manner requested, significantly impeded this proceeding, or provided information that was not accurate or verifiable. Thus, the Department's use of adverse FA in this review does not meet even the minimum required statutory conditions for the use of non-adverse FA.

      Further, LG disputes the Department's two justifications for the use of adverse inferences with respect to the Mexican sales. First, LG disputes the reasoning that ``because LG did not report these sales as U.S. sales, we are not using the expenses.'' LG argues that the Department may not punish it simply for taking a position regarding the underlying sales--a position supported by Department and Court precedents--

      [[Page 69710]]

      with which the Department disagrees. LG additionally disputes the contention that the Department was unable to verify these sales. Because none of the expenses in question are transaction-specific, none are in any way different for the diverted sales than for the U.S. sales, and the Department verified each expense.

      In conclusion, LG argues that the Department may not treat as U.S. sales third country sales for which there is no evidence of entry into the United States. LG contends that in the fourth review the Department did not include all the sales to the third country customer, but only sales for which there was corroborating Customs documentation which showed that the merchandise was entered into the United States. LG states that whether the error was intentional or inadvertent, however, the Department should correct this error in the final results. The law is clear that the Department may not treat as sales to the United States sales that did not enter the United States.

      In rebuttal, Micron argues that the Department properly determined that LG knew or should have known that the unreported sales to its customer in Mexico were destined for the United States. According to the petitioner, the volume and pattern of the sales, the circumstances of the placement of orders and payment by the customer's U.S.-based parent, and the delivery of the subject merchandise by LG to the customer's agent in the United States, all substantiate the Department's finding that LG knew or should have known that the ultimate destination of the sales was the United States.

      Micron disputes the ``facts'' submitted by LG. According to Micron, the declarations of Mr. Simon and Mr. Lee of LG, regarding LG's Mexican customer, are not ``facts,'' but simply statements from interested parties which do not square with neutral observers who claim something different. Micron cites a Dun & Bradstreet (``D&B'') Report which characterizes the company as a very small operation.

      Micron asserts that it is not credible that Mr. Simon or Mr. Lee, if they had actually inspected the facilities of LG's Mexican customer, would have concluded that the company could have consumed internally all of the merchandise that it purchased from LG.

      Micron also submits that the quantity of DRAMs that the company was buying makes it impossible for it to have been as small an operation as D&B and LG officials reported. Thus, according to Micron, if it really were an OEM consuming the merchandise it purchased from LG in its own production, the company evidently refurbished more computers than is possible for such a small company. (And this estimation does not take into account the fact that the computers being re-furbished would already contain some pre-existing DRAMs.)

      Further, Micron argues that this is an unheard of quantity for a second-hand computer vendor, and would put it among the top tiers of new computer vendors. Micron contends that if the Mexican customer had really shipped that many computers in a year, it would have been a much better known name. Furthermore, the story of refurbishing old computers is inconsistent with the type of modules purchased. The overwhelmingly vast majority of the sales are of newer model DRAMs the type that cannot be used in the older computers that the company was refurbishing. The more advanced DRAMs are for use in newer generation computers. In short, according to Micron, the company was buying modules for use in the newest PCs.

      With regard to manufacturing, Micron states that, during the POR, the company purchased a large amount of DRAMs for manufacturing. Micron points out that in his declaration, Mr. Simon stated that the facility in Tijuana had two manufacturing lines, with three or four workers each. These, Micron contends, are also evidently the same lines on which computers and other products, according to Daniel Lee's declaration, were being refurbished. Micron alleges that this means that Mr. Simon and Mr. Lee believed that it was reasonable that a small amount of workers could use a large amount of DRAMs to manufacture computer applications, while at the same time producing a great many computers. According to their declarations, both Mr. Simon and Mr. Lee had worked for LG for many years, and would have visited many computer manufacturing facilities while on sales calls. Micron contends that they would have noticed immediately the disparity between the quantity of merchandise purchased and the size of the companies facilities. Micron concludes that the statements that they thought the facilities could handle the quantity are extremely self-serving, but are just not credible.

      Micron argues that if LG really did not know what the company was doing, it would be because nobody had actually visited the Tijuana facilities. The fact that they described the Tijuana facilities in terms similar to those in the D&B report indicate that they had either visited the facilities or had read the D&B report. In either case, according to Micron, it would have been readily apparent that the Mexican company could not do what it said it was doing.

      During the POR, Micron alleges that the Mexican customer acted as a broker for LG's products. Micron states that most brokers have to take the risk, when they purchase DRAMs, that the price they can command on resale may fall below their own purchase price. LG, however, allowed the Mexican customer to distribute its DRAMs without such a risk. When the customer eventually sold the LG DRAMs, from several days to several weeks after it had bought them, LG reinvoiced the customer at a new price. Until near the end of the POR, when there was a temporary rise in market prices, the new price was always lower than LG's other customers. According to Micron, its research indicates that the price for the Mexican customer was always lower than the open market, or ``spot'' market, at the time of the re-invoicing.

      Micron asserts that LG's Mexican customer buys the DRAMs, and immediately tries to sell them into the spot market, with LG's knowledge and assistance. When it has a buyer, it informs LG of the price, and based on the spot price, LG re-invoices the sale. Thus, Micron maintains that the facts indicate that LG always knew that it was dealing with a DRAM broker, not an OEM. The fact that LG felt compelled to come up with the thoroughly implausible story of this customer being a computer accessory manufacturer and refurbisher means that it had something to hide.

      Micron argues that LG continues to misinterpret the law regarding whether actual knowledge is required to impute sales. According to Micron, the standard for attributing U.S. sales to a foreign producer or exporter is not restricted to ``actual knowledge,'' but is whether the producer or exporter ``knew or had reason to know'' that such sales were destined for the United States. Micron argues that in attributing U.S. sales to a foreign producer or exporter, longstanding administrative practice and judicial rulings, consistent with the statute's legislative history, establish that the Department may find either direct evidence of knowledge, or impute such knowledge, provided in either event its finding rests on a reasonable factual foundation. While the Department can rely on direct evidence of actual knowledge, such evidence is rarely forthcoming. Therefore, Micron maintains that the Department is not restricted to an actual knowledge test, and may impute knowledge, as warranted, as was referenced in the Department's May 27, 1999,

      [[Page 69711]]

      memorandum in this review. The Department applied this standard here.

      Micron states that the statute's legislative history expressly supports the Department's ``knew or had reason to know'' standard. See section 772(a) of the Act, 19 U.S.C. 1677a(a). According to Micron, although the current statute does not directly address how the Department should determine attribution of sales ``for exportation'' to the United States, the legislative history to the term's predecessor provision, ``purchase price,'' does.

      Micron states that in the Trade Agreements Act of 1979, Congress modified the definition of ``purchase price'' in 19 U.S.C. 1677a(b) to provide statutory authority for the administrative practice of basing U.S. price on the transaction from a producer to an unrelated reseller if the producer knew that the product was destined for the United States. The statute did not indicate the degree of knowledge necessary to find that a producer knew the destination of the merchandise, but the SAA adopted with the Trade Agreements Act of 1979 (H.R. Doc. No. 153, 96th Cong., 1st Sess., at 411 (1979)) states: ``The definition {of purchase price} makes clear that if the producer knew or had reason to know the goods were for sale to an unrelated U.S. buyer, and the terms of the sale were fixed or determinable from events beyond the control of the parties as of the date of importation, the producer's sales price will be used as the ``purchase price'' to be compared with that producer's foreign market value.'' The Department has explained that its application of the knowledge standard is based upon the House Report language cited in the 1979 SAA.

      Moreover, Micron asserts that, when Congress amended the statute in 1994, changing ``purchase price'' to ``export price,'' it made clear that ``notwithstanding the change in terminology, no change is intended in the circumstances under which export price (formerly ``purchase price'') versus CEP (formerly ``exporters sales price'') are used.'' See 1994 SAA, H.R. Doc. 103-316, 103d Cong., 2d Sess. (1994), at 822- 23. Micron claims that Congress implicitly endorsed retention of the ``knew or had reason to know'' standard under the old law when it changed purchase price to export price, and the Department continues to apply that standard to attribute to a foreign producer or exporter sales destined for the United States.

      Micron argues that the Department's administrative practice and judicial precedent support the ``knew or had reason to know'' standard. Micron maintains that consistent with the legislative history, the Department's longstanding and current practice is to determine whether the foreign producer ``knew or had reason to know'' that the sales in question were destined for the United States. The Department makes its knowledge finding on a case-by-case basis after assessing all of the information on the record. Micron alleges that the courts have repeatedly upheld the Department's practice. See, e.g., Federal Mogul Corp. v. United States, 17 CIT 1015 (1993) (``If the ITA finds that respondents knew, or should have known, that sales to Japanese OEMs with U.S. affiliates were destined for the U.S. market, the ITA will disregard those sales in calculating FMV''); Yue Pak. Micron argues that, in Yue Pak, the CIT expressly acknowledged that ``Commerce interprets the phrase ``for exportation to the United States'' to mean that the reseller or manufacturer from whom the merchandise was purchased knew or should have known at the time of the sale that the merchandise was being exported for the United States,'' and stated that it has upheld this interpretation. Yue Pak, 20 CIT at 498. The Court of Appeals affirmed the CIT decision in Yue Pak, adopting the holding and reasoning of the Court below. 111 F.3d 141-42.

      Micron alleges that LG ignores this longstanding practice, however, contending that several of the Department's determinations and certain judicial rulings require evidence of ``actual'' knowledge. LG, Micron argues, misreads these cases. They do not repudiate use of the ``knew or should have known'' standard. Rather, as discussed below, those cases turned on whether there was evidence of knowledge, actual or constructive, with respect to the destination of the sales in question. For example, Micron contends that LG's assertion that the Court, in NSK, implicitly rejected a reason to know standard is simply erroneous. The point of the Court's holding in NSK, according to Micron, is that knowledge of the ultimate destination of the goods, whether actual or constructive, must exist with respect to particular sales. The type of required knowledge does not, as LG asserts, limit the evidentiary basis to proof of actual knowledge, or some admission by the producer. Micron states that the Department may reasonably impute knowledge concerning the ultimate destination of particular sales if the facts support such an inference, as they clearly did here. Unlike the situation in NSK, in this case, the Department looked at a very specific group of sales, and compiled an extensive record on the distinct facts and circumstances bearing on LG's reason to know that these particular sales to this particular customer were destined for the United States.

      Similarly, according to Micron, LG contends that the CIT's ``knew or had reason to know'' test is relevant only to the issue of knowledge of sales in the home market under section 773(a) of the Act, 19 U.S.C. 1677b(a), as opposed to knowledge of sales for export to the United States. LG Case Brief at 27. The INA case, Micron maintains, stands for no such proposition.

      Micron claims that in INA, the Court clarified the standard for determining whether sales of a respondent may be included in the home market database. The Court stated that the test was whether the respondent ``knew or should have known that the merchandise was not for home market consumption based upon the particular facts and circumstances surrounding the issues.'' The Court did not construe, and in fact, made clear it was not altering, the standard for imputed knowledge of U.S. sales. The Court merely noted that while knowledge (actual or imputed) of the U.S. destination must be established to treat an exporter's sales as sales to the United States, it was not necessary to find such knowledge of the ultimate destination in order to exclude sales for export from the home market database. The Court never suggested that imputed knowledge was only permissible in considering home market sales.

      According to Micron, LG's arguments regarding Yue Pak are incongruous because, as discussed above, NSK and INA no more directly address LG's contention that ``should have known'' is not a sufficient basis for attributing sales than does Yue Pak. Neither of these cases discredit, but instead clarify, the Department's constructive knowledge standard. Indeed, far from ``overruling'' Yue Pak, neither NSK nor INA even reference the Court's earlier decision in Yue Pak.

      Moreover, Micron claims that, LG inaccurately describes both the facts and the law under Yue Pak. According to Micron, the Department and the CIT considered extensive evidence that indicated knowledge by the PRC producers, but very little if any of the evidence could be considered the sort of direct evidence that would permit a finding that the PRC suppliers in question actually ``knew'' of the U.S. destination, such as when a producer is informed in advance of the U.S. destination, or otherwise admits its awareness. Rather, the bulk of the evidence was what might be considered

      [[Page 69712]]

      indirect, i.e., specific labeling instructions relaying DOT and OSHA requirements, special order purchase practices, and the percentage of shipments to the United States versus those to third countries. Such evidence, Micron asserts, is precisely the sort of evidence indicating that the producer had ``reason to know'' of the U.S. destination, and thus the CIT's affirmance of the Department's finding of knowledge is directly relevant here.

      Nor, according to Micron, does LG's citation to TRBs II support its theory of the knowledge standard. LG seizes on the Department's statement in that case that ``lacking evidence of actual knowledge that particular sales were destined for the United States, we cannot assume such knowledge, regardless of general knowledge that some merchandise was intended for exportation to the United States.'' However, Micron argues that the Department was merely noting that the proper evidentiary basis must exist in order to infer knowledge; it was not abandoning its longstanding knowledge standard. Indeed, the Department reaffirmed the ``knew or had reason to know'' formulation in the immediately following section of the decision, finding that respondent Premier's suppliers were unaware of the U.S. destination of their merchandise.

      Further, Micron alleges that other cases cited by LG similarly do not repudiate a constructive knowledge standard, but merely show that where there was no reasonably plausible evidence suggesting the producer had knowledge at the time of the sale that the particular sales were destined for the United States. Therefore, the Department need not consider, let alone impute, knowledge. Micron contends that such cases are a far cry from the situation here, where the record is replete with evidence establishing that the producer knew or had reason to know of the U.S. destination of the sales in question. In such cases, the Department can and does impute knowledge.

      For example, Micron alleges that in Tapered Roller Bearings from China, there was no indication that the goods ever entered the United States. In contrast, Micron argues, in the present case the record shows not only purchase orders issued directly by the U.S.-based purchaser to the producer's U.S.-based sales affiliate, but also delivery to the purchaser in the United States, entry of the goods for consumption in the United States, and payment by the purchaser from a U.S. bank. Thus, issuance of the purchase orders by a ``U.S. firm'' is only one piece of evidence among many.

      Similarly, Micron maintains that the Ukrainian Plate decision starkly differs from the instant case. There, the Department had no basis to entertain imputed knowledge, because the evidence in this regard was virtually nonexistent. The Department has vastly more information in support of its decision in the instant case.

      Micron alleges that in an attempt to side-step the collective impact of the multiple factors supporting the Department's preliminary decision here, LG addresses each factor in isolation, arguing that such factors as ``dealings with a U.S. company or a shipment route through the United States do not transform a third country sale into a U.S. sale.'' Aside from assuming the conclusion--that these were ``third country sales''--the decisions cited by LG can all be distinguished from the instant case.

      With respect to Magnesium, Micron contends that LG confuses the allegations by the petitioner with the findings made by the Department. There, the Department found that the producer did not know until after the time of sale that it was selling to a U.S. customer. Here, by contrast, the producer's U.S.-based sales outlet, LG, was very clearly dealing directly and repeatedly with the U.S.-based customer. Similarly, in Manganese Sulfate from China, it was not until after the date of sale that the shipping document showing the U.S. port as the destination of shipment was issued, and numerous other factors indicated lack of knowledge. And in Tapered Roller Bearings from China, purchase orders from a U.S. company were insufficient to impute knowledge because the shipments were to a third country and there was no other evidence that the producer was aware at the time of sale that the merchandise was destined for the United States.

      In this regard, Micron takes issue with LG's contention that after LG had arranged for delivery of the goods to its agent in the United States, they were transported away from the United States in bond, and in-bond entries are not considered to enter the Customs territory of the United States. Micron argues that, in fact, as the Department noted, LG shipped the DRAMs to its customer's agent in the United States, without requesting nor receiving assurance that the goods would be placed in Customs bond upon arrival and thereafter remain in bond until exported outside the United States. Moreover, as LG must acknowledge, the goods were in fact entered for consumption into the Customs territory of the United States.

      In the Persulfates determination, according to Micron, the ``knowledge'' (or lack thereof) of the ultimate destination was not relevant; it was sufficient that the producer had knowledge that the goods were being shipped to an unaffiliated purchaser in the United States, and that the purchaser entered the goods for consumption. In this regard, Micron maintains that Persulfates offered an alternative basis for attributing the sales in question to LG, as the fact that the Mexican customer entered the merchandise for consumption in the United States rendered the knowledge issue irrelevant.

      Micron believes the Department's application of adverse FA in the calculation of the margins for the Mexican sales is appropriate. Micron states that the Department should apply a total adverse FA rate to all of LG's U.S. direct and indirect sales.

      Micron maintains, however, that if a calculation of the rate for the Mexican sales is required, the Department acted in accordance with law in using adverse FA to determine LG's dumping margin for the preliminary results, and should use the same methodology for the final results. Micron argues that LG withheld requested information, failed to provide information in the form and manner requested, significantly impeded the proceeding, and failed to act to the best of its ability to comply with the Department's request. See 19 U.S.C. 1677e. According to Micron, LG deliberately withheld important information requested by the Department concerning U.S. sales, and attempted instead to characterize that information as sales to third-countries. Not only did LG fail to provide information of its U.S. sales in the form and manner requested by the Department, but LG's willful attempt to mislead the Department, to LG's benefit, significantly impeded the proceeding. Micron argues that LG's failure to submit requested data constituted ``noncompliance with an information request'' within the meaning of Olympic Adhesives. In addition, LG's failure to produce requested information when it knew that these allegedly third-country sales were in fact sales to the United States, constituted a failure to cooperate. Therefore, LG failed to act to the best of its ability by knowingly withholding information requested by the Department. As a result, the Department appropriately applied an adverse inference under Section 1677e(b) in selecting from the facts otherwise available.

      Micron states that to facilitate its analysis under Section 1677e, the

      [[Page 69713]]

      Department has developed several factors that it applies on a case-by- case basis. See SAA at 870 (``In employing adverse inferences, one factor the agencies will consider is the extent to which a party may benefit from its own lack of cooperation.''); Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France et al.; Final Results of Antidumping Duty Administrative Reviews, 62 FR 2081, 2088 (Jan. 15, 1997) (considering (1) the experience of the respondent in antidumping duty proceedings, (2) whether the respondent was in control of the data which Commerce was unable to verify or rely upon, and (3) the extent the respondent might have benefitted from its own lack of cooperation); see also Extruded Rubber Thread from Malaysia; Final Results of Antidumping Duty Administrative Review, 63 FR 12762 (Mar. 16, 1998) (using same criteria).

      Micron alleges that in applying these factors to this case, it is indisputable that the Department's use of total adverse FA to determine LG's dumping margin was warranted. Not only is LG an experienced respondent in the annual review processes, but LG was in control of the U.S. sales data requested, and due to its deceptive failure to report these sales, the Department was unable to verify such information. More important, however, LG stood to benefit from its lack of cooperation. Had the Department not known of LG's U.S. sales, its calculation of LG's dumping margin would be skewed in LG's favor. Micron contends that this is simply unacceptable. See SAA at 870.

      DOC Position. A full discussion of our final conclusion, which requires references to proprietary information, is included in the December 6, 1999, Memorandum from John Conniff to Holly Kuga regarding sales through a third country by LG contained in the official file for this case. Generally, however, we have found that the record evidence concerning unreported sales supports the conclusion that LG knew, or should have known, that at the time it sold the subject DRAMs, the merchandise was destined for consumption in the United States.

      With respect to knowledge, we do not agree with LG's contention that the Department may not assign a FA rate on the basis of the unreported sales since LG had no actual knowledge of the diversion of these sales. Numerous court decisions, including those by the U.S. Court of Appeals for the Federal Circuit, have held that the appropriate standard for making this decision is ``knew or should have known at the time of the sale that the merchandise was being exported for the United States.'' See Yue Pak. See also Peer Bearing Co. v. United States, 800 F. Supp. 959, 964 (CIT 1992), Certain Pasta From Italy: Termination of New Shipper Antidumping Duty Administrative Review, 62 FR 66602 (1997), and Manganese Sulfate. While the statute does not indicate the degree of knowledge necessary to find that the producer knew the destination of the merchandise, the courts have stated that even if a respondent denies knowledge of the destination of its sales, the Department may review all facets of a transaction, and based on extrinsic source data, determine that it is appropriate to impute knowledge in a given case. See INA 1997, 957 F. Supp. at 265.

      In the matter of these unreported sales, first we note that LG essentially dealt with a U.S. customer. When shipping the merchandise, LG took no steps itself to ensure that, when the merchandise was delivered to the United States, it was subsequently placed under Customs bond and transported to a third country, clearing Customs upon export from the United States. What the record shows is that LG sold an enormous amount of DRAMs to a very small company and turned the merchandise over to the customer in the United States. Consequently, in contrast to such cases as Ukraine Plate and Magnesium, LG knew for certain that it was shipping DRAMs into the United States.

      Moreover, this is not a situation where an exporter sells and ships a relatively small amount of subject merchandise to a third country and then, sometime much later, the customer reexports the merchandise to the United States. In this case, we are confronted with a staggering amount of merchandise that is being shipped by LG directly to the United States. The merchandise is subsequently being entered for consumption into the United States within days, if not hours, of it leaving the possession of LG.

      The relative size and nature of the purchaser's operations and the quantity of acquisitions it made are germane to this case in several respects. The amount of purchases this customer made are not modest. In fact, the entered value of these transactions was quite large. However, based on LG's description of the purchaser's operations, it is clear that this party was not equipped to absorb such a vast amount of DRAMs. In particular, LG should have known that the purchaser was buying more DRAMs than it reasonably could consume in the manufacture of modules or the refurbishment of computers and printers. Furthermore, the amounts the customer purchased were so enormous they had to appear inconsistent with the size of the third-country DRAM markets in question. Moreover, as Micron points out, this customer could be expected to sell the vast majority of its merchandise to the United States. Consequently, not only was it reasonable to assume that this firm would sell some or all the subject merchandise that it purchased, but that it would sell the merchandise to the United States.

      In summary, based on the nature and characteristics of these transactions, we conclude that LG knew, or should have known, that the merchandise was destined for the United States. Considering the above, and as more fully described in the above-mentioned agency memorandum, the Department has decided to include the unreported sales during the POR in the analysis conducted of LG's sales for these final review results. See the FA section of this notice for a discussion of the FA that were applied in the case of LG.

      Comment 2: LG's Knowledge of U.S. Sales: Germany. On September 13, 1999, the Department placed on the record a memorandum and accompanying exhibits regarding certain LGSG sales to a customer in Europe that subsequently shipped the LG DRAMs in question to its related entity in the United States. The documents consisted of an anonymous e-mail from a former LG employee, LG verification exhibits, U.S. Customs data, and a signed declaration concerning this transaction chain from a former LG salesman. On October 7, 1999, LG submitted information in response to the Department's September 13, 1999, memorandum. Other interested parties also filedrelevant factual information regarding this matter by letter dated October 7, 1999.

      LG questions the Department's placing these allegations on the record so late in the proceeding, and states that nothing in the September 13, 1999, memorandum or elsewhere in the record provides a lawful basis for the Department to treat sales of DRAMs by LG's German subsidiary to its European customer as ``U.S. sales'' of LG. To the contrary, LG argues that the record provides no evidence that any responsible official of LG knew, at the time of sale, that these particular shipments were ultimately destined for the United States.

      LG contends that the record before the Department provides overwhelming evidence that LG correctly considered these sales to Germany as third-country sales and accurately treated them as such in this proceeding. Furthermore,

      [[Page 69714]]

      LG claims that the evidence demonstrates that the transactions between LGSG and its European customer were motivated solely by legitimate business reasons and not by pricing differentials or the existence of an antidumping duty order.

      LG argues that, as a matter of law, because LG did not know that particular sales were destined for the United States, none of the sales can properly be treated as U.S. sales by LG. LG contends that it is well-established that an exporter may not report sales made to a third country as U.S. sales unless the exporter knew at the time of sale that particular sales were destined for the United States. LG believes that the NSK case is directly on point, because while LG had generalized knowledge that some of the DRAMs sold to its European customer might ultimately be shipped to the United States, LG did not know that any particular sales were destined for import into the United States.

      Similarly, LG argues that in INA, the CIT distinguished between the knowledge standard for treating sales to resellers as sales to the home market, and the standard for treating sales to export markets as sales to the United States. Thus, LG argues, only if the respondent had known which specific sales were destined for the United States could the Department have considered the sales to be U.S. sales under section 772(b) of the Act. LG asserts that in this case too, where LG did not have such knowledge, the sales to Germany cannot be considered U.S. sales.

      Likewise, LG maintains that in Tapered Roller Bearings and Parts Thereof from China, the Department followed the holdings of NSK and INA in finding that generalized knowledge by suppliers that some sales to a reseller were destined for the United States was not adequate to treat the suppliers' sales as U.S. sales. Thus, according to LG, the Department has stated explicitly and unambiguously that the sort of general knowledge that some merchandise was intended for exportation to the United States that LG possessed with regard to the subject sales is insufficient for the sales to be treated as U.S. sales by LG.

      LG contends that the record in this case establishes that LG did not know, and had no way of knowing, that any particular sales by LGSG to Europe were destined for the United States. Indeed, LG does not know even now which of its sales to its European customer were destined for the United States because it distributed the DRAMs that it bought from LG both to the United States and elsewhere, and did not inform LG as to the destination of the goods either before or after the time of sale. In these circumstances, where LG lacked ``actual knowledge that particular sales were destined for the United States,'' LG maintains that the law is clear that the sales may not be considered as U.S. sales of LG.

      According to LG, the Department has placed on the record evidence obtained from the U.S. Customs Service that purports to ``indicate * * * the likelihood that all of LG['s] sales [to the European customer] entered the U.S.'' LG believes that there are numerous deficiencies, however, in the evidence provided by the Department and the conclusions that the Department draws from that evidence. LG states that from the Customs data, quite the opposite is true--all of LG's sales to its European customer did not enter the United States.

      First, according to LG, the undisputed evidence on the record shows that during the fifth review period, LGSG sold a great quantity of DRAMs to its European customer. The Customs data produced by the Department, however, shows that fewer DRAMs were sold to the United States by its European customer during this period. Thus, LG concludes that even if all DRAMs that Customs claims came into the United States were manufactured by LG, there are still almost a majority of DRAMs that were sold by LGSG to its European customer that were not resold by the customer into the United States.

      Second, LG maintains that there is no information contained in the Department's Exhibit 3a concerning the identity of the manufacturer of the DRAMs imported into the United States. LG alleges that the Department has placed on the record only two instances of underlying invoices in which LG is identified as the manufacturer, and these two imports cover an insignificant amount of units. Thus, LG contends that there is no way to determine, much less conclude that it is ``likely,'' that LG was the manufacturer of all of the DRAMs imported by its European customer into the United States. For all the record shows, the remaining DRAMs imported into the United States could all have been manufactured by Samsung or other DRAM suppliers.

      Finally, LG claims that, for more than half of the transactions between LGSG and its European customer, the Department is unable to provide any evidence linking these sales to the specific Customs data regarding U.S. entries of Korean DRAMs. For the remainder of the transactions between LGSG and the European customer, which the Department has purported to link to particular U.S. imports by the customer, the Customs data fail to identify a manufacturer or even a product code. Thus, this record provides no evidence, other than two individual import transactions, that the customer shipped to the United States any DRAMs that LGSG had sold to it.

      According to LG, the European customer's U.S. affiliate purchased DRAMs in Europe as a method in order to take advantage of various countries' Outward Processing Relief (``OPR'') provisions. Although DRAMs later went to a duty-free status in Europe and there was no longer a need to use the OPR provisions, the supply chain had been established. LG states that, because of the reliability of this transaction chain and the ``historic'' ties between the European customer and its U.S. affiliate, the sales continued through this channel.

      LG argues that it makes no sense for the Department to conclude that LG made sales through Europe in order to avoid reporting them as U.S. sales when these transactions would have lowered its dumping margin. In the Department's recent decision in DRAMs from Taiwan, the Department stated with regard to two separate incidents that no adverse action is warranted when a respondent has erroneously reported or failed to report sales but correcting the error would lower the respondent's dumping margin. Thus, even if the Department were to conclude that LG's failure to report these sales was an error, there would still be no cause for the Department to take adverse action against LG.

      LG claims that an e-mail sent by a former LG employee to the Department accusing LG of dumping DRAMs into the United States through its sales to the European customer is unreliable and has no evidentiary value. LG asserts that the employee left the company under unfavorable circumstances. LG submitted in its letter on October 7, 1999, a record of this individual's employment which documented his problems with the company. LG believes in light of the circumstances of his termination, the e-mail is wholly unreliable as evidence against LG. Additionally, LG argues that the evidence of an accountant who had no involvement in sales lacks probative value, particularly when that evidence is evaluated in light of his obvious bias and when that evidence is measured against abundant, reliable evidence that entirely contradicts it.

      LG also questions the veracity of the declaration by Mark Vecchiarelli, the LG manager responsible for sales LG to the customer in question during the POR. LG has registered its strenuous objections to the Department's conduct

      [[Page 69715]]

      with regard to Mr. Vecchiarelli, both in performing a ``secret'' interview with him and in drafting multiple versions of his declaration.

      LG disputes Mr. Vecchiarelli's claim that he ``was responsible for servicing all of the semiconductor requirements of [the customer in question] on a worldwide basis'' and that he was ``responsible for the pricing and supply decisions for all sales worldwide to the

      [company] .'' LG claims that Mr. Vecchiarelli was responsible for all of LG's sales to the parent company worldwide, but he was not responsible for sales by other subsidiaries of LG to the branches of this company located outside the United States. His successor, Mr. Pizarev, confirmed this account in LG's October 7, 1999, submission to the Department.

      LG claims that during the time that Mr. Vecchiarelli worked for LG, it was Mr. Sung-Jung Woo of LGSG who was responsible for making sales from LGSG to Europe, not Mr. Vecchiarelli. LG also disputes Mr. Vecchiarelli's statement that he ``left LG on good terms for a more lucrative position and for the career advancement opportunities available at TranSwitch.'' LG claims that Mr. Vecchiarelli left his employment at LG to become the Western Area Sales Manager at Macronix America, a subsidiary of a well-known Taiwanese memory semiconductor producer and a direct competitor of LG. LG believes this omission obscures Mr. Vecchiarelli's credibility.

      Further, LG disputes Mr.Vecchiarelli's statement that he ``made sales to [the customer's] divisions located outside the United States and arranged for other LG entities to supply the semiconductors to these * * * divisions for ultimate delivery to [its] manufacturing facility in the United States.'' According to LG, Mr. Sung-Jung Woo of LGSG was responsible for making these sales, not Mr. Vecchiarelli. In addition, LG points out that Mr. Pizarev, who was trained by Mr. Vecchiarelli to be his permanent successor as the account manager, has attested to the fact that Mr. Vecchiarelli never mentioned that he sold DRAMS to this customer in the United States through other LG subsidiaries in third countries.

      Moreover, LG also claims a ``floor'' price, or minimum price for the sales of DRAMs in the United States, as compared to Europe, did not exist. This is documented by the fact that prices to the European customer in question were not in fact lower than LG's prices in the United States.

      In addition, LG argues that the customer in question claimed multiple uses for the discrete DRAMs it purchased, contradicting Mr. Vecchiarelli's statement that merchandise was ultimately destined for the United States, and that the parent ``did not subcontract, anywhere else in the world, the production of memory modules using the discrete DRAMs LG sold to [it].'' LG claims that Mr. Vecchiarelli was not in a position to know or supply the customer's global supply needs. According to LG, Mr. Vecchiarelli only had control of fulfilling the customer's supply requirements through LG.

      LG concludes its arguments by stating that even if every word in Mr. Vecchiarelli's declaration is truthful and accurate, nothing in Mr. Vecchiarelli's declaration indicates that LG knew that particular sales to its European customer were destined for the United States. While some specific orders from LGSG may have been shipped in their entirety to its U.S. affiliate, others clearly were not; some or all of the DRAMs in those other orders were sent elsewhere, to destinations outside the United States. Rather, LG maintains that the evidence on the record shows that while LG had generalized knowledge that some DRAMs sold to its European customer might end up in the United States, LG did not know that any particular sales were destined for the United States. Further, LG contends that the evidence on the record shows that there were no significant differentials between LG's prices in Europe and in the United States. LG also questions the credibility of the assertions made by the two ex-LG employees referred to in the materials released by the Department. For all of these reasons, LG states that it is clear that the Department in the final results should not treat LGSG's sales to the customer in question as U.S. sales of LG.

      Micron argues that LG engaged in multiple schemes to manipulate the calculation of its dumping margin by supplying the U.S. market with subject merchandise shipped through intermediaries and third countries. According to Micron, LG's attempts to explain away the unmistakable import of the record are unavailing. Two former employees of LG have come forward with direct evidence of an evasion scheme in which LG supplied the U.S. market by shipping DRAMs through its affiliate in Germany, knowing the DRAMs sold to the customer in question were destined for the customer's operations in the United States. Micron contends this deliberate evasion of the antidumping duty order has been fully substantiated by the sales data provided by LG at verification as well as the import records received by the Department from Customs. The egregious conduct demonstrated by respondent in this case demands that the Department apply total adverse FA to establish the dumping margin for LG.

      According to Micron, the information LG has submitted confirms that LG was well aware that the Korean-made DRAMs that it supplied through Europe were being shipped to the United States. Micron argues that the record evidence supports the finding that LG had more than its admitted ``general knowledge'' regarding the U.S. destination of the LG DRAMs sold to the European customer in question. Micron maintains that the cumulative evidence of record--including the sworn statement of Mr. Vecchiarelli and the Department's corroborating data showing exact correspondence between individual LGSG sales to the customer in question and individual import transactions in the Customs data-- indicate that LG had actual knowledge that particular sales to its European customer were for export to the United States.

      Thus, Micron claims that, as LG itself points out, during the fifth review period LG sold a great amount of DRAMs to the European customer in question. According to Micron, the available Customs data show that, during the same period, the European customer's U.S. affiliate imported a large quantity of DRAMs that LG sold to the European customer in question. Moreover, Micron states that the available data show that a significant portion of these transactions were back-to-back, with the sale from LGSG to the European customer coinciding with a corresponding shipment (units and value) from its European customer to the U.S. affiliate. Micron contends that the correspondence of the sales volume admittedly sold from LGSG to the European customer to the available Customs import data provides sufficient evidence of LG's actual knowledge that particular sales were destined for the United States.

      Micron disputes LG's claim that the Department lacks sufficient information to conclude that all of the entries of Korean-made DRAMs shown on the Customs import listing were in fact made by LG. First, Micron claims that the Customs listing of DRAM import transactions in Exhibit 3a to the September 22, 1999, Memorandum indicates that all of the transactions were entered showing Korea as the country of origin and the customer in question as the importer. Further, the Department's September 22, 1999, Memorandum indicates that the attached import transaction

      [[Page 69716]]

      documentation in Exhibit 3c are provided as ``two examples'' demonstrating that the entries of DRAMs at issue were in fact manufactured by LG.

      Second, Micron states that the back-to-back transactions listed in Exhibit 3b to the Department's September 22, 1999, Memorandum, identical as to quantity and value, further confirm that LG was the manufacturer of these DRAMs. Finally, Micron alleges that the customer in question never contended that the imports entered by its U.S. affiliate were manufactured by any party other than LG. Since the customer does not dispute that all imports consist of LG-made DRAMs, and the sample import documentation establishes that LG was the manufacturer, it is reasonable to conclude that all of the listed entries consist of LG DRAMs.

      Further, Micron alleges that the customer's manufacturing operations reinforce LG's knowledge of the U.S. destination of its sales. Micron argues that the statement of Mr. Vecchiarelli establishes that LG had actual knowledge of the U.S. destination of the discrete DRAMs that LG was supplying to the customer in question through third countries. Mr. Vecchiarelli describes the particular types of DRAMs that LG was selling to the customer in question, and the operations in which those DRAMs were being utilized. Micron contends that any vendor supplying a large multinational OEM with a large volume of product will not remain so ignorant of the OEM's operations as LG contends to be. See DRAMs from the Republic of Korea--Revision of Exhibit 3 of the Department's September 13, 1999 Memorandum from John Conniff to the File.

      Micron notes that LG places great reliance on the very generalized denials of the customer in question. According to Micron, the customer claimed that it uses discrete DRAMs in many manufacturing locations other than the United States, but makes the most generalized claim of alternative uses and fails to identify even one specific location where the purchased LG DRAMs were being used. Moreover, Micron maintains, the denial is set forth in a compound form that lumps the LG-supplied DRAMs with DRAMs purchased from all other vendors: ``DRAM sold to the European IPO by L.G. Semicon and other vendors went to a variety of

      [sites throughout Europe] .''

      Micron states that LG attempts to buttress its story by attributing to the arrangement some unsubstantiated, and internally inconsistent, business justifications. According to Micron, LG also places great emphasis on the second-hand statements of Mr. Sung-Jung Woo, an LG employee who had previously served as sales manager of LGSG. These statements are not provided directly by Mr. Woo, but instead through the declaration of Mr. Jae-Byung Kim, another LG employee.

      Micron maintains that the unexplained second-hand nature of the statements attributed to Mr. Woo casts significant doubt on their reliability. Micron contends that, since Mr. Woo continues to be employed by LG, there appears to be absolutely no reason why LG could not have provided a first-hand account by Mr. Woo. For that reason, the conclusory denials of LG's knowledge of the destination of the sales should be given little weight.

      Micron argues that the prices charged by LG to its European customer confirm LG's intent to evade the antidumping order. According to Micron, the prices charged by LG through its alternative sales through LG and LGSG confirm the critical facts contained in the statement of Mr. Vecchiarelli that LG maintained a ``floor price'' on its sales through LG to the United States, and that LG made sales through LGSG when the price needed to make the sales to its European customer was below that ``floor price''.

      Micron asserts that, in a market in which prices are continually declining, prices averaged over twelve months can be significantly skewed by the volumes sold at different times; and this was particularly true with sales by LGSG to the customer in question. Indeed, Micron states that when prices are examined on a daily basis, there is a clear pattern confirming Mr. Vecchiarelli's statement that LG was selling through LGSG in order to continue to supply its European customer's U.S. affiliate.

      According to Micron, LG also points to the statement of Mr. Vlad Pizarev, Mr. Vecchiarelli's successor, as indicating that pricing is the one function that is centralized worldwide. Micron states that LG emphasizes the statement: ``Prices were usually the same worldwide.'' Micron argues that, as noted, this statement very pointedly does not dispute, and the establishment of a single world-wide price for this customer only confirms, LG's need to supply this customer in the United States through an alternative route when the agreed-upon world-wide price is set.

      Micron argues that Mr. Vecchiarelli's statement is corroborated by other evidence of record and provides every indication of reliability. According to Micron, LG makes an attack on Mr. Vecchiarelli's integrity in an attempt to discredit his testimony. Those claims, Micron argues, are misplaced and should be rejected. First, the Department employees who spoke directly with Mr. Vecchiarelli had a first-hand basis on which to judge his credibility and reliability. LG acknowledges that Mr. Vecchiarelli left LG on good terms, and can proffer no substantial reason why Mr. Vecchiarelli should harbor any bias towards LG. Second, Mr. Vecchiarelli's apparent employment at Macronix America immediately after leaving LG's employment provides absolutely no basis for inferring any bias against LG. Mr. Vecchiarelli was no longer employed at Macronix at the time he provided his statement to the Department, so the basis for any potential ``bias'' had already been eliminated. Moreover, LG by its own actions has indicated that it did not consider Mr. Vecchiarelli's employment at Macronix to constitute a disqualifying bias against LG. As related in the Statement of Mr. Pizarev, Mr. Vecchiarelli continued to be engaged by LG as a consultant ``on a retainer from LG for a couple of months'' after he left LG to work at Macronix.

      Third, LG grossly mischaracterizes the supposed discrepancies in Mr. Vecchiarelli's statement. Thus, LG first contests his statement that he was responsible for the pricing and supply decisions for all sales worldwide to the customer in question. Yet LG's own submissions confirm this statement. As already discussed above, LG submitted the statement of another LG employee (Mr. Pizarev), who confirmed that the parent company's pricing is ``centralized'' worldwide. And the organization charts submitted by LG at verification quite plainly indicate that Mr. Vecchiarelli was in charge of worldwide sales to the customer in question. In fact, Mr. Vecchiarelli is shown at the top of the chart, with the title ``WW Act. Mgr''. Furthermore, this document from the Department's sales verification report clearly reviews the customer in question's DRAM product needs on a worldwide basis, with information on products manufactured at each location.

      Micron reiterates that nowhere does LG deny that LG maintained a ``floor price'' system, as Mr. Vecchiarelli described it in his statement.

      In sum, Micron contends that Mr. Vecchiarelli's statement describing LG's evasion scheme is credible, corroborated by the Customs documents as well as by LG's own sales documents, and confirmed in many respects both by LG's and the customer in question's admissions and by their failure to deny critical aspects of the arrangement.

      [[Page 69717]]

      Micron submits that LG's complaints regarding the Department's procedures should be summarily rejected. According to Micron, it did not respond to the summary arguments in LG's October 7 letter because they appeared to be nothing more than a preview of legal arguments to be presented in LG's case brief. Micron contends that it now appears that LG is resting on the arguments as summarily stated in the October 7 letter. Those arguments are entirely baseless and, like so many of LG's assertions in this proceeding, not worthy of serious consideration.

      Micron maintains that LG's allegations that the Department (1) failed to ``promptly'' place information on the record, (2) conducted ``secret'' interviews, and (3) failed to afford respondent with a ``meaningful'' opportunity to respond to allegations, thereby denying LG's ``fundamental right to due process,'' totally lack merit.

      First, according to Micron, LG ignores the fundamental nature of an antidumping proceeding. Second, LG's allegation with respect to the Department's ``secret interviews'' with Mr. Vecchiarelli goes against the statute, which affirmatively authorizes the conduct of ex parte meetings. Third, LG's ``strenuous'' objections to these meetings, and to the proffer of information by a former employee concerning fraudulent conduct by the employer, are totally baseless.

      Finally, the Department is afforded great discretion in conducting its proceedings. E.I. Dupont de Nemours & Co. v. United States, Slip Op. 98-7, 1998 WL 42598, at *11 (CIT Jan. 29, 1998) (``Commerce enjoys broad discretion in conducting investigations and reviews under the antidumping statute''). As the Court of International Trade has previously recognized:

      Commerce regularly balances its interest in conducting an efficient, uniform and expeditious administrative investigation against its equally compelling interest in conducting accurate fact- finding. Such a weighing of competing interests involves choices of administrative practice and procedure which Commerce, in its specialized role as administrator of antidumping investigations, is uniquely qualified to make.

      See Union Camp Corp. v. United States, 53 F. Supp. 2d 1310, 1328 (CIT 1999); see also NEC Corp. v. United States Dept. of Commerce, 978 F. Supp. 314, 327 (CIT 1997), aff'd, 151 F.3d 1361 (Fed. Cir. 1998), cert. denied, 119 S.Ct. 1029 (1999) (noting that the Assistant Secretary for Import Administration ``enjoy(s) a presumption of honesty and integrity which must be overcome''). In short, Micron contends that LG has failed to provide any legal foundation for its allegations concerning the Department's investigative procedures, and the Department should dismiss these claims as groundless.

      DOC Position: A full discussion of our final conclusion, which requires references to proprietary information, is included in the December 3, 1999, Memorandum from John Conniff to Holly Kuga regarding sales through a third country by LG contained in the official file for this case.

      In sum, an employee in a significant position of LG stated for the record that he set up a sales channel for one of LG's major customers to procure DRAMs for the United States through LG's subsidiary in Germany. LG has not submitted anything for the record of the instant review that would lead us to believe that the employee's responsibilities were any less than he described. LG itself acknowledges that it knew that ``some of the merchandise'' sold through Germany was destined for the United States. The record contains ample information to document the fact that the overwhelming majority of the merchandise sold through Germany did, in fact, ultimately enter the United States for consumption during the POR. Therefore, we believe the record evidence supports the conclusion that LG knew, or should have known, at the time it sold the subject DRAMs, that the merchandise was destined for consumption in the United States.

      Comment 3: Adjustments to LG's Reported Cost of Manufacturing. LG claims that the Department should not adjust its cost of manufacturing by including certain costs from LG's construction-in-progress (``CIP'') account.

      Micron argues that the Department properly adjusted LG's reported cost of manufacture for certain production expenses that LG had excluded from cost of manufacture and instead relegated to a CIP account.

      DOC Position: Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department should adjust LG's reported COM is moot.

      Comment 4: Adjustment to LG's Reported G&A Expense. In its preliminary results, the Department adjusted LG's reported G&A expense by excluding foreign currency transaction gains and losses related to accounts receivables. See Preliminary Results, 64 FR at 30,485; Analysis Memorandum at 4 & Attachments 7, 9, 10. LG alleges that the Department's calculations contain a significant error that should be corrected in the final results. Specifically, the Department inadvertently added three zeros to three of the figures in the tables contained in Attachments 9 and 10: accounts receivable, long-term accounts payable, and bank deposits. Thus, the Department should use a revised G&A ratio percent in the Final Results.

      No rebuttal briefs were filedwith regard to this issue.

      DOC Position. Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department should adjust LG's reported G&A expense is moot.

      Comment 5: The Department Should Use the Data Submitted by LG in Its March 26, 1999 Supplemental Response. In the preliminary results, the Department used the sales and cost data submitted by LG with its original October 8, 1998, questionnaire response. However, LG submitted revised cost and sales data with its March 26, 1999, supplemental response. LG argues that this data was timely submitted and was used as the basis for the Department's verification in April 1999, and the Department should, therefore, use LG's March 26, 1999, data in the final results.

      No rebuttal briefs were filedwith regard to this issue.

      DOC Position: Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department should use LG's March 26, 1999, data submission is moot.

      Comment 6: The Department Should Correct Programming Errors in the Calculation of G&A and Interest Expense for Modules LG argues that the Department made programming errors in its application of the revised G&A and interest expenses for memory modules.

      No rebuttal briefs were filedwith regard to this issue.

      DOC Position: Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department revises LG's G&A and interest expenses for memory modules is moot.

      Comment 7: The Department Should Correct a Programming Error in the Calculation of LG's COP and CV for DRAMs. LG claims that the Department should correct a programming error in the calculation of COP and CV for DRAMs.

      No rebuttal briefs were filedwith regard to this issue.

      [[Page 69718]]

      DOC Position: Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department corrects the programming error in the calculation of COP and CV for DRAMs is moot.

      Comment 8: The Department Should Correct a Programming Error that Significantly Overstates the Duty Assessment Rates Covering LG Imports. LG claims that, due to a computer programming error, the Department's duty assessment rates by importer are significantly overstated.

      No rebuttal briefs were filedwith regard to this issue.

      DOC Position: Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department has the duty assessment programming error is moot.

      Comment 9: The Department Should Calculate LG's CV Selling Expenses Based on Density. LG claims that the Department erroneously calculated a single weighted-average home market selling expense figure for CV- based on sales of all products. To correct this distortion in the dumping margin calculation, the Department should calculate CV selling expenses based on density.

      No rebuttal briefs were filedwith regard to this issue.

      DOC Position: Given that the Department is rejecting LG's reported sales and cost information to calculate LG's margin, and is applying total FA, the issue of whether the Department calculates CV selling expenses based on density is moot.

      Final Results of Review

      As a result of this review, we have determined that the following margins exist for the period May 1, 1997 through April 30, 1998:

      Weighted- Weighted- Manufacturer/Exporter

      average margin average per percentage megabit rate

      Hyundai Electronics Industries, Co.,

      10.44

      .03 Ltd.................................... LG Semicon Co., Ltd.....................

      10.44

      .03 G5 Corporation..........................

      10.44

      .03

      The Department shall determine, and the Customs Service shall assess, antidumping duties on all appropriate entries. The Department will issue appraisement instructions directly to the Customs Service. These final results of review shall be the basis for the assessment of antidumping duties on entries of merchandise covered by this review. For Hyundai, for duty-assessment purposes, we calculated an importer- specific assessment rate by aggregating the dumping margins calculated for all U.S. sales to each importer and dividing this amount by the total estimated entered value reported by Hyundai of those sales. Hyundai, in accordance with the Department's questionnaire, estimated the entered value of its sales by calculating the average of the entered value of each control number for the POR. For all other respondents, we based the importer-specific assessment rate on the facts available margin percentage.

      Furthermore, the following deposit requirements will be effective upon publication of this notice of final results of review for all shipments of DRAMs from Korea entered, or withdrawn from warehouse, for consumption on or after the publication date, as provided for by section 751(a) of the Act: (1) for the companies named above, the cash deposit rates will be the rates listed above; (2) for merchandise exported by manufacturers or exporters not covered in this review but covered in a previous segment of this proceeding, the cash deposit rate will continue to be the company-specific rate published in the most recent final results which covered that manufacturer or exporter; (3) if the exporter is not a firm covered in this review or in any previous segment of this proceeding, but the manufacturer is, the cash deposit rate will be that established for the manufacturer of the merchandise in these final results of review or in the most recent final results which covered that manufacturer; and (4) if neither the exporter nor the manufacturer is a firm covered in this review or in any previous segment of this proceeding, the cash deposit rate will be 4.55 percent, the all others rate established in the LTFV investigation. These deposit requirements shall remain in effect until publication of the final results of the next administrative review.

      This notice serves as a final reminder to importers of their responsibility under 19 CFR 351.402 (f) to file a certificate regarding the reimbursement of antidumping duties prior to liquidation of the relevant entries during this review period. Failure to comply with this requirement could result in the Secretary's presumption that reimbursement of antidumping duties occurred and the subsequent assessment of doubled antidumping duties.

      This notice also serves as the only reminder to parties subject to APO of their responsibility concerning the disposition of proprietary information disclosed under APO in accordance with section 351.305 (a) of the Department's regulations. Timely notification of return/ destruction of APO materials or conversion to judicial protective order is hereby requested. Failure to comply with the regulations and the terms of an APO is a sanctionable violation.

      We are issuing and publishing this in accordance with sections 751(a)(1) and 777(i)(1) of the Act.

      Dated: December 6, 1999. Richard W. Moreland, Acting Assistant Secretary for Import Administration.

      [FR Doc. 99-32399Filed12-13-99; 8:45 am]

      BILLING CODE 3510-DS-P