Council Regulation (EC) No 1480/2003 of 11 August 2003 imposing a definitive countervailing duty and collecting definitively the provisional duty imposed on imports of certain electronic microcircuits known as DRAMs (dynamic random access memories) originating in the Republic of Korea

• (87) As explained already above in the context of the assessment of the three previous investigated measures, Hynix had exhausted its possibilities of receiving bank loans as early as January 2001 due to its high exposure to its major creditor banks and its weak credit rating. By March 2001 its rating had been downgraded further and its exposure increased due to the GOK-directed lifting of the prudential credit limits as regards some banks. It had not serviced its syndicated loan granted in January 2001 and its liabilities kept cumulating further, despite all the measures taken to alleviate its situation. Under these circumstances, there is no indication that Hynix would have been able to get a loan from the market. Hynix did not raise such a possibility during the procedure, neither did it indicate that it even tried to get further loans from either the Korean or foreign markets.

• (88) The information on the record indicates that the banks did not appear to have had any commercial grounds to purchase KRW 1 trillion worth of convertible bonds (‘CBs’) from Hynix in June 2001. The evidence on the record indicates that the bonds were purchased in order to provide Hynix with cash to cover its existing liabilities. The banks acknowledged the inability of Hynix to service its debts already at the beginning of May 2001. In the Creditor Financial Institutions Council Resolution of 7 May 2001, it is stated that Hynix could not fulfil its obligations as regards the syndicated loan for the first quarter of 2001 and that it was highly likely that it would not be able to do so also for the second quarter. This was stated to be a default under Article 12 of the loan agreement. However, in order to prevent related problems like cross default, Article 11 of the loan agreement was not applied for the first and second quarter of 2001. The banks thus effectively exempted Hynix from the consequences of the default under its loan agreement. Nevertheless, they still granted new financing by purchasing the CBs in June 2001.

• (89) Several Hynix creditor banks also increased their loan loss provisions with respect to the Hynix debt ([^E0039]). Prior to participation in the June 2001 CB purchase, KorAm Bank, Hana Bank, Shinhan Bank and Kookmin Bank increased their loan loss reserves for Hynix by 25 %. By the third quarter of 2001, these banks had classified 60 % or more of their loans to Hynix as non-performing. It is recalled that three of these banks failed to cooperate with the investigation (see recital 18 above).

• (90) The banks purchased the CBs on 20 June 2001. It is noted that between mid-June and 20 June 2001, the Hynix stock price had sunk considerably. Despite this, the banks purchased the CBs even though it must have been evident to them at that time that they would not recover their money ([^E0040]). Within one month the banks had incurred huge losses on the bonds. The Hynix stock price had declined sharply and the banks directed certain portions of the acceptance price of Hynix CBs as losses in settling their accounts at the end of June 2001.

• (91) The information received after the publication of the provisional measures and not rebutted by the parties indicates that contacts between the creditor banks and public officials took place in March to May 2001. According to this information, the 10 March 2001 meeting referred to in recital 82 was convened in order to get the banks to support the Hyundai Group of companies ([^E0041]). When some banks indicated reluctance in providing support to Hynix since they were concerned about the downgrading of external credibility and violation of credit limits, the GOK financial officials still persuaded the banks to do it ([^E0042]). On 24 April 2001 it was reported that Hyundai Creditors' Association decided not to purchase KRW 1 trillion convertible bonds (‘CBs’) requested by Hynix. Hynix had no liquidity to pay off its maturing liabilities and the proceeds of the CB issuance, KRW 1 trillion, were to be used for servicing its existing debt. The presidents of the creditor banks, including KEB, KDB and Chohung Bank had a meeting on 23 April 2001 in which most participants were opposed to accepting CBs in the amount of KRW 1 trillion ([^E0043]). On the same day, the CEO of Citibank, Citibank being the financial adviser of Hynix at the time and the organiser of the May measures, formally requested financial support for Hynix from the GOK, and the FSC Chairman expressed a positive position on it. Against this background, it was reported that financial support to Hynix, including acceptance of CBs, had become more likely ([^E0044]). It is recalled that Citibank has also failed to cooperate with the investigation (see recital 18).

• (92) It was reported that KorAm Bank, one of the banks not cooperating with the investigation, refused to take over its share of the KRW 1 trillion of CBs on the basis that Hynix failed to deliver a memorandum pledging to make its best effort to reduce its debts. It is observed that there were reports indicating that as a response to this, the FSS stated that it would not forgive the bank if it would not participate, adding that it will take stern measures against the bank, such as disapproving new financial instruments and subjecting the bank to a tighter audit. Following this warning, KorAm reversed its decision and participated in the purchase of CBs ([^E0045]).

• (93) Considering that in the absence of cooperation by the parties the final conclusions need to be based on best facts available, further information, even if extracted from secondary sources, should be observed in this context. It is noted that one of the measures originally investigated was an alleged SGICO (Seoul Guarantee Insurance Corporation) guarantee for KRW 600 billion of Hynix CBs to be purchased by investment trust companies. It was confirmed during the investigation that this bond purchase never took place and therefore no guarantee was provided ([^E0046]). The reasons for the failed transaction, however, were not explained. It is observed that there are reports indicating that the banks at the time required that in order for them to purchase the CBs, the investment trust companies also needed to do so since the banks did not want to bear all the financial burden themselves. The investment trust companies refused to buy Hynix bonds since they claimed it was common knowledge that the bonds were insolvent bonds and that they were not able to sell them at the time when they were struggling to recover the confidence of the market. According to the reports, the investment trust companies stated that they would not be able to avoid criticism that they were investing their customers' money in an insolvent company, which could lead to lawsuits against them. If the investment trusts were forced by the financial authorities to accept the new corporate bonds in the amount of several billions of won under these circumstances, the whole investment trust circle may become insolvent ([^E0047]). On the basis of the above it appears that the banks were participating in all the discussions on the May 2001 measures since early March 2001 and that they were well aware of the arguments of the investment trusts and the reasons these refused to subscribe the bonds. The banks themselves had also decided not to accept the CB purchase some weeks earlier and had acknowledged the inability of Hynix to service its existing debt.

• (94) Following the disclosure of this information to the interested parties, the parties admitted that an FSS official was present in one meeting and that subsequent contacts between FSC/FSS and the banks took place. It was explained that the FSS official was only present as an observer to act as a witness for the creditors' prior commitments of funding, and not to influence the creditor banks or their decision to extend further credits to Hyundai companies. It was also explained that the ‘follow-up phone calls’ with the FSS/FSC officials and the creditor banks were only conducted in the exercise of the normal prudential supervisory role of the FSC/FSS. However, considering that the parties have been withholding important information until confronted with it and impeding the investigation by instructing other parties not to cooperate, it is difficult to find these explanations fully convincing. In addition, information on the record indicates that the official attending the meeting of 10 March 2001 was the FSC Vice Chairman, i.e. a high-ranking official. In these circumstances, pursuant to Article 28 of the basic Regulation which allows the conclusions to be based on best facts available in cases of non-cooperation by the parties, it is concluded that the banks were not freely and independently deciding on the issue of the CB purchase on the basis of commercial considerations, but were directed to buy the bonds by the GOK. The information from the secondary sources referred to above supports this conclusion.

• (95) In its comments on the final disclosure the GOK argued that account should be taken of the fact that private investors had bought Hynix GDRs worth USD 1,25 billion in June 2001 and that therefore the banks' decision to buy Hynix CBs was in line with commercial considerations. It is noted that these investors must have been influenced by the anticipated continuous availability of the GOK-directed KDB programme to refinance Hynix's maturing liabilities. Moreover, these investors did not have the same insight in the situation of Hynix as the creditor banks. Therefore, it is not unreasonable to conclude that the decision of the banks to invest should have been different from that of the GDR investors.

• (96) On the basis of the best facts available it is therefore concluded that the decision of the banks to purchase KRW 1 trillion of Hynix CBs was not taken on the basis of commercial considerations, but was taken due to direction by the GOK pursuing public policy goals. The amount of the CBs purchased is therefore a financial contribution by a government within the meaning of Articles 2(1)(a)(iv) and (i) of the basic Regulation. This purchase conferred a benefit to Hynix in the amount of KRW 1 trillion, the purchase price of the bonds, since comparable funds were not available to it in the market. Hynix's situation had deteriorated further from January 2001 and it had no possibility to receive an equivalent loan from the market. For these reasons, the amount of the CB purchase of KRW 1 trillion is considered as a subsidy in the meaning of Article 2 of the basic Regulation. Since this is an ad hoc subsidy provided to only one company, it is specific pursuant to Article 3 of the basic Regulation and therefore countervailable.

• (97) According to Article 5 of the basic Regulation, the amount of countervailable subsidies shall be calculated in terms of the benefit conferred on the recipient. The CBs purchased by the creditor banks had an interest rate and were to be paid back at their maturity by Hynix. They are therefore comparable to a loan. According to Article 6(b) of the basic Regulation the first point to examine is whether a comparable commercial loan was available in the market and if so, the benefit would be the difference between the interest rates applied. As explained in recital 87 there were no comparable loans available to Hynix in May 2001. For this reason, the benefit conferred by the bond purchase needs to be determined on another basis.

• (98) According to Section E(b)(v) of the Guidelines, the amount of the loan forgiven or defaulted on will be treated as a grant. This means that if at the time of the granting of a loan it is already evident for the granting authority that it will not recover its money, the loan should be considered as a grant. In the present case, as explained under recitals 88 to 90, it was obvious for the banks at the time of the purchasing of the bonds that they would not recover the money they had spent on purchasing the Hynix's CBs. Therefore the benefit to Hynix and the amount of countervailable subsidy is the purchase price of the bonds, KRW 1 trillion. Using the calculation method explained in recitals 67 and 108 of the provisional Regulation, the subsidy amounts to 5,4 %.

THE OCTOBER 2001 RESCUE PACKAGE, CONSISTING OF DEBT TO EQUITY SWAP, THE SECOND ROLL-OVER OF DEBT AND THE PROVISION OF A NEW LOAN OF KRW 658 BILLION

• (99) Following the publication of the provisional Regulation, Hynix, the GOK, KEB, Woori Bank, Chohung Bank (‘CHB’), Citibank Seoul and National Agricultural Cooperation Federation (‘NACF’) submitted comments concerning the provisional findings as regards the October 2001 rescue package.

• (100) The comments of the parties are listed and assessed below. However, since the calculation of the amount of the countervailable subsidy is an important question as regards the October 2001 measures, it is considered useful to explain how and following which principles the assessment of the amount of subsidy is conducted, in particular considering that the October 2001 measures were granted to a company in financial difficulties.

Introduction

• (101) In the provisional Regulation, the October 2001 measures, namely the debt-to-equity swap, the new loan of KRW 658 billion and the extension of maturities and roll-over of debt, were considered as grants. This was based on Section E(b)(v) of the Guidelines, which provides that a loan will be considered as a grant if it is forgiven or defaulted on. Recitals 104 to 107 of the provisional Regulation provisionally concluded that due to the financial situation of Hynix, no corresponding financing was available to it in the market and that the banks knew at the time of the October 2001 measures that they would not recover their money invested in/lent to Hynix.

• (102) The first point to examine when assessing the calculation of the subsidy in cases of provision of financing to a company in difficulty is the credit rating of the company at the time of the measures. The rating is indicative of the financial situation of the company and its ability to attract investments or receive financing at the given time since the potential investors/lenders base their decisions above all on the rating given to the company by an independent international rating agency.

• (103) It is noted that the rating given to Hynix in October 2001 by Standard & Poor's was SD — ‘selective default’. This is the second-worst rating possible, the worst being ‘default’, which according to Standard & Poor's criteria is not prospective, as the other ratings, but used only when the default has actually occurred. SD is assigned when an issuer can be expected to default selectively, that is, continue to pay certain issues or classes of obligations while not paying others. Consequently, the rating of Hynix at the time of the measures was very alarming to any investor/lender and did not support a decision to grant further financing to it under commercial considerations.

• (104) The second point to examine would be the history of the company in servicing its existing debts and the rating given to it by the lenders concerned. If a company with even a moderate rating continued to service its existing debt to some of its lenders, it would not be completely unreasonable for these lenders to take this into account in their internal rating of the company concerned and also consider this fact positively when assessing whether further lending to the company concerned could be considered. In the present case, however, it is noted that Hynix had not serviced its existing debts in 2001, and had even defaulted on the interest payments of the syndicated loan granted in January 2001. As explained in recital 88, the banks announced the default already in May 2001, but decided not to take the appropriate measures to address the situation. Several banks raised their loan-loss provisions as regards Hynix already in May 2001 and again in October 2001 before the measures, indicating that the further financing was already considered a loss. The internal ratings of the banks concerned reflected this assessment. The six banks concerned by the October 2001 measures rated Hynix internally from ‘precautionary’ to ‘doubtful’. Such a situation does not support a decision to grant further financing to the company concerned.

• (105) The third point to examine would be the general financial situation of the company concerned. If the company is experiencing temporary liquidity problems but has healthy fundamentals, it is justified for potential lenders to take this into account when taking their credit decisions.

• (106) It is noted that the overall financial situation of Hynix had been critical for a long period. At the end of 2000, Hynix had accumulated more than USD 9,46 billion of liabilities. This was almost twice its net worth and more than four times the value of market capitalisation of the company ([^E0048]). In October 2001, Hynix's debts were six times its equity. This situation was also well known by the banks concerned. As explained in recital 44, KEIC already stated in November 2000 that Hynix was in a state of technical insolvency. Similar statements can be found in the internal reports of the banks concerned which refer to Hynix having negative capital and excessive lending.

• (107) The financial ratios of Hynix according to the so-called Altman Z-score model, designed specifically to predict the likely failure of Korean companies, were in 1999 to 2001 far worse than other Korean companies that actually did fail. The Hynix scores were 700 % higher than the threshold level of a company in severe financial distress ([^E0049]). Hynix's current and quick ratios were indicative of a company with totally inadequate liquidity to cover obligations ([^E0050]). Hynix's net profit margin, return on assets and return on equity were negative in nearly every year since 1997, with net margins reaching minus 24,3 % in 2000 and a minus 93,83 % in 2001. In any year from 1997 to 1999, the company's cash flow from operations was not enough to cover even one sixth of its debt or total liabilities.

• (108) The 2001 financial accounts also discounted the possibility of future financial success of Hynix, stating that the entire normalisation of the company's operation required the continuous support from creditor banks until the selling price of semiconductor products sufficiently recovered. These facts indicate that both the history and the future of Hynix was characterised by its dependency of financial assistance from its creditors. Hynix had reached a point where it was unable to finance its operations by the returns generated by its activities.

• (109) The next point in the assessment of the basis of calculation of the subsidy would be to see whether other loans from commercial sources had been granted to the company concerned, and if so, under which conditions. If such loans were granted, the terms and conditions of such loans could serve as a benchmark for measuring the benefit conferred to Hynix by the measures concerned, and thereby also the amount of subsidy.

• (110) In the present case, it is noted that Hynix could not obtain any commercial loans at any time in 2000 and 2001. Hynix's borrowing was exclusively concentrated on GOK-provided, GOK-directed and GOK-guaranteed loans. The largest creditor of Hynix at the time of the measures was KDB, which held 44 % of the total of Hynix loans in December 2001 ([^E0051]).

• (111) Prior to 1998, Hynix had received loans from French and Japanese banks, including the Bank of Tokyo, Société Générale and Credit Lyonnaise. Since then, however, none of these banks has participated in the restructuring of Hynix. The outstanding loans from foreign banks, including those from Citibank Seoul, accounted for only 5 % of the total Hynix loans in 2000 to 2001 ([^E0052]). It is noted that the assessment of the role of Citibank is explained in recitals 130 to 133.

• (112) The lack of participation of foreign banks suggests that further investment in Hynix was not considered to be rational and economically justified. It is noted that the outstanding amount of KRW 37,5 billion of the syndicated loan from foreign borrowers granted in 1996 and led by Société Générale was declared for default and cross default in 2001.

Financial contribution by governmentComments of the parties

• (113) Hynix and the GOK both argued that the October 2001 package did not constitute a subsidy since the financing provided was not granted by the GOK. Hynix argued that as regards Citibank and KEB, no evidence was provided showing that the GOK had used its influence to force the banks to accept the October 2001 measures. As regards Woori Bank, CHB and NACF, Hynix argued that the GOK shareholding should not be confused with government control of credit decisions. It was further argued that the credit decisions were based on commercial considerations by each of the banks concerned.

• (114) KEB argued that the GOK had played no role in the restructuring programmes in relation to KEB's commercial decisions. It argued that the GOK has no special right either contractually or under Korean law that would allow it either to block or influence KEB's commercial decisions. According to KEB, the same applies also to the FSS. In relation to the FSS's management recommendations for the KEB business operations it was argued that these actions fall within the scope of its ordinary activities pursuant to the Banking Act and the banking Regulations and were not to be understood as meaning any kind of interference or involvement on the part of the FSS or any other financial supervisory authority.

• (115) Woori Bank argued that, despite its position as the largest shareholder, the GOK was prohibited from exercising any influence over the Woori bank's decision-making process.

• (116) CHB argued that the GOK has no direct or indirect influence on its management activities. Even if Korea Deposit Insurance Corporation (‘KDIC’) has a substantial shareholding in CHB, CHB allegedly remained a commercial bank and its decisions were taken on the basis of purely commercial considerations. It argued that the Memorandum of Understanding (‘MOU’) referred to in the provisional Regulation only concerned the various restructuring measures imposed by the GOK in response to the 1997 financial crisis.

• (117) Hynix and CHB also argued that other banks (Industrial Bank of Korea (‘IBK’), Seoul Bank, Kwangju Bank and Kyongam Bank) where the GOK also had a substantial shareholding did not participate in the October 2001 package, and this demonstrated that the GOK had not interfered in the business activities of any of the banks involved.

• (118) NACF argued that it was formed for the purpose of improving the economic, social and cultural status of farmers and quality of their lives. KDIC as a preferred investor was not allowed by law to participate in NACF decision-making. As such, its mandate did not include implementation of public policy goals.

• (119) Citibank Seoul denied being in any way affiliated with the GOK and explained that it participated in the investigated measures solely in pursuit of its commercial interests.

Assessment of the Comments

• (120) It is noted that it has been provisionally concluded on the basis of the principles set out in recitals 10 to 14 above that KDB is a public body. As regards the other banks, these were not found to be public bodies but were found to be directed by the GOK to carry out the measures in question.

• (121) In terms of demonstrating direction, the fact that the GOK was the only or the major shareholder in some of the banks is an important point that needs to be taken into account in the assessment of the situation. The amount of a shareholding by the GOK, or other shareholders, determines the extent to which they can influence the decision-making process set out in company law and the articles of association of the company concerned. In this respect it is important to refer to the Korean Prime Ministerial Decree No 408 adopted in November 2001, which addresses management responsibilities in companies where the government has a shareholding. Though Article 6 of the Decree emphasises the importance of allowing such companies to manage themselves, it crucially allows the State to fulfil its full rights as a shareholder.

• (122) The information on the record also indicates that the State as a majority shareholder can appoint directors of financial institutions. As such, the GOK regularly nominates or carries influence in the selection of a CEO and other members of a bank's management ([^E0053]). Through its voting power, the GOK is therefore in a position to influence the operations of the KEB, Woori and the CHB, where it is the largest shareholder. For example, information on the record indicates that the State influenced the decision to change KEB's CEO following poor performance with Hynix and certain credit-card companies ([^E0054]).

• (123) Moreover, it should be noted that the provisional Regulation did not rely solely on the GOK's shareholdings to prove direction. Recitals 88 to 98 of the provisional Regulation explained extensively as regards each bank the reasons for which it was found that the GOK had exercised direction with regard to the October 2001 measures. In addition to the explanations provided in the provisional Regulation, these findings are now further reinforced by additional information, some of which has been received after the publication of the provisional measures.

• (124) As explained under recitals 32, 42 and 43, the records contain evidence that the GOK gave direct instructions to financial institutions concerning some of the measures under investigation. This information also indicates that the FSS was involved in the measures. The record also indicates that the GOK coerced these institutions to participate in some measures by threatening to impede their business operations ([^E0055]). It is noted that the Commission had on several occasions asked both the GOK and banks concerned about such involvement, yet all parties have consistently denied that there was any State involvement in relation to the measures in question.

• (125) On 3 August 2001, briefly before the October 2001 measures, the Korean Deputy Prime Minister stated in the seminar meeting of the Newspaper and Broadcast Editors Association that in the event that the creditor group was unable to resolve the Hynix issue, the GOK would come forward to make a quick decision. If the creditor group could not make a decision on additional financing to Hynix, the financial authorities should decide.

• (126) As for the KEB argument that the GOK has no legal or contractual right to influence its commercial decisions, it is noted that the findings of the provisional Regulation on the GOK direction were not based on such contractual or legal rights. The findings on this point sought to demonstrate how the GOK was able to use its power to direct KEB in a more subtle way, for example through its position as a major shareholder. Moreover, the preliminary conclusions on this point are reinforced by new evidence received after the publication of the provisional Regulation where GOK ministers intervened to instruct KEB in relation to some measures investigated — see recitals 32, 42 and 43.

• (127) As regards the FSS and FSC, evidence obtained since the provisional measures also demonstrates that the FSS was involved in the decisions on measures providing financing to Hynix, e.g. the letter from Korean ministers to KEB referred to in recital 126 above. As regards the FSS/FSC's role of carrying out its supervisory activities in relation to the ongoing restructuring of KEB, the evidence on the record shows that the FSC approved the extension of KEB lending limits, a fact which ran counter to the restructuring plan which demanded a reduction of its excessive credit exposure to Hynix. Moreover, in the KEB financial accounts a considerable amount of Hynix debt-to-equity swap was not included in the loss calculation of the troubled loans. Consequently, the comments of KEB on the provisional Regulation have not provided any new substantial evidence that would alter the assessment set out in recitals 90 to 91 of the provisional Regulation. These conclusions are therefore confirmed.

• (128) As regards NACF, recitals 96 to 98 of the provisional Regulation explained the grounds on which it was provisionally concluded that NACF was directed by the GOK in participating in the October 2001 measures. Even though the form of NACF as a cooperative is fully recognised, its cooperative goals to improve the economic, social and cultural status and quality of life of farmers are also economic policy objectives to support agriculture. It is further noted that NACF is jointly managed by the Ministry of Finance and Economy and the Ministry of Agriculture and Forestry. It has specialised functions as an agricultural policy bank, such as implementation of Korea's agricultural policy and provision of GOK funds for the agricultural sector. The new Agricultural Cooperative Law of August 1999 provides for GOK grants/support when necessary (Article 9), allows the GOK to take an equity stake (Articles 147 to 151) and enables NACF to issue Agricultural Finance Bonds and requires the GOK to guarantee it (Articles 153 and 156). The credit rating given by NACF to Hynix at the time of the measures, namely 8 ([^E0056]), did not justify granting further financing to Hynix. The comments of NACF on the provisional Regulation have not provided any new substantial evidence that would alter the assessment set out in recitals 96 to 98 of the provisional Regulation. These conclusions are therefore confirmed.

• (129) As regards the Woori Bank and the CHB ([^E0057]), recitals 88 and 89 of the provisional Regulation explained the reasons on the basis of which it was provisionally concluded that Woori Bank and CHB were directed by the GOK to participate in the October 2001 measures ([^E0058]). These conclusions on direction are reinforced by the information received after the imposition of provisional measures and which is referred to in recitals 91 to 94, 124 and 125. Woori Bank and CHB have not provided any new evidence in their comments which would alter the Commission's conclusion as set out in recitals 88 and 89 of the provisional Regulation. These conclusions are therefore confirmed.

• (130) As regards Citibank, as explained in recitals 93 and 95 of the provisional Regulation, it failed to cooperate fully with the investigation. In these circumstances, findings with regard to its involvement in this financial package to Hynix have been established in accordance with Article 28 of the basic Regulation, i.e. on the basis of facts available.

• (131) In terms of its role as a lender in this investigation it should be noted that the non-cooperation of Citibank prevented any reliable information being obtained on the precise functions and practices of Citibank in relation to the financial package given to Hynix. Thus, Citibank's assertions that it was not acting under GOK direction could not be verified via a proper response to a questionnaire and a subsequent verification visit. The same applies to the lending practices of Citibank. It should also be noted that the failure of Citibank to cooperate significantly impeded the investigation, not just in terms of identifying Citibank's precise functions in relation to its role as a lender, but also in respect to its role as the financial adviser to Hynix. Due to the non-cooperation the true status of Citibank as regards Hynix and the nature and intensity of its overall relationship to the GOK remain unknown. The importance of cooperation and verification visits is demonstrated by the fact that as regards some of the other banks involved in the case, it was proper cooperation and verification that was instrumental in allowing definitive conclusions to be drawn as regards the relations between them, Hynix and the GOK. What the investigation has, however, shown, due to information given by other parties, is that Citibank was a central figure in discussions between the GOK, Hynix and other parties involved in this case. Indeed, Hynix stated that Citibank should not cooperate in the investigation because ‘such disclosure may divulge information relating to Hynix's cost, finance or accounting’. It is recalled that Hynix was majority-owned by the GOK via the banks at the time of this intervention.

• (132) In these circumstances, the final findings had to be based on the facts available with regard to whether Citibank as a lender acted under GOK direction and whether the benefit to Hynix equated to the full financial contribution made by Citibank to Hynix. In this respect, it is uncontested that unlike the other participants in these measures, Citibank's first intervention as a creditor bank of Hynix was in January 2001 when the financial situation of Hynix was already sufficiently bad to deter any other new bank from getting financially involved. Citibank's own rating for Hynix in October 2001 was D ‘doubtful’. Nevertheless, Citibank provided financing to Hynix. Citibank was requested to explain their general lending policy and whether it was normal to provide financing to doubtful companies. As already explained, Citibank did not provide any details which would have explained their motivation in participating in the measures investigated.

• (133) As for the GOK direction, information on the record indicates that there were considerable links between the GOK, Hynix and Citibank. For the GOK and Citibank these went beyond the loans investigated. Such links can be interpreted in two ways. Firstly, they can be interpreted as indicating that Citibank may well have had commercial reasons for providing financing to Hynix, or as alleged by one of the parties, they could be interpreted as placing Citibank in a vulnerable position with regards to direction by the GOK. Due to the non-cooperation it was not possible to establish whether Citibank acted on coercion or whether this was in accordance with their normal business practice. In this respect, it is recalled that the reason for non-cooperation was to prevent access to Hynix's cost, finance and accounting data. In the absence of any other explanation, this can reasonably be taken as an indication that this data in Citibank's possession contain information revealing that there were no commercial reasons to provide the financing in question, and that the financing was provided due to GOK direction. This reason is also the one suggested in the complaint. In addition, as explained in recital 94 of the provisional Regulation, Citibank has had an unusually close and symbiotic relationship with the GOK since 1967, when it was authorised to operate in Korea. This close relationship between the GOK and Citibank is witnessed in the role played by Citibank in assisting the GOK to extricate itself from the Korean financial crisis of 1997. Citibank led and successfully completed Korea's bank debt restructuring for a total of USD 21,75 billion in 1998. Moreover, Citibank helped the GOK and government-related institutions to access capital markets during the Korean financial crisis by successfully sponsoring a USD 4 billion global bond offering. All these facts confirm that Citibank has a very close relationship to the GOK. On the basis of these facts, and of the refusal of Citibank to grant access to the information in its possession, and failing any other verifiable evidence being available, the conclusion to be drawn in accordance with Article 28(6) of the basic Regulation is that the GOK was involved and directed Citibank to provide the financing in question.

• (134) As for the argument that not all banks where the GOK had a substantial shareholding participated in the October 2001 measures and that this disproved that GOK direction was taking place, it is noted that Kwangju Bank and Kyongam Bank were about to be merged into Woori Financial Holding at the time of the measures. As Woori Bank was already one of the major creditor banks in the October 2001 restructuring, it is understandable that it was not necessary for Kwangju and Kyongam Banks to participate in the measures separately. The same situation applies to Seoul Bank, which the GOK was planning to sell to Hana Bank following the recommendations of the IMF. In anticipation of this sale, a rise in speculative loans on the books of the Seoul Bank was not seen to be beneficial.

• (135) As regards IBK, it is noted that IBK is a specialised bank whose mandate is to promote the development of small and medium-sized enterprises. Its operations are underwritten by the GOK. By extending further lending to Hynix, IBK would have exceeded its mandate in a very apparent way which would have been extremely difficult to justify. In fact, by granting the D/A facility and participating in the CB purchase as regards Hynix, IBK had already breached its mandate and further exposure to a large company in difficulties would not have been in line with its policy goals.

Benefit conferred by the measures

• (136) KEB, Woori Bank, CHB, NACF and Citibank Seoul all argued that they participated in the October 2001 package since they wanted to maximise their recovery rate for the loans already granted to Hynix. They considered that the value of Hynix as a going concern was higher than the immediate liquidation value. They argued that their participation in the October 2001 package was determined on the basis of their credit evaluation of Hynix, supported by analysis from outside consultants. It should be noted, however, that Hynix has refused to provide a copy of the analysis determining the liquidation value on the basis that it was confidential. Consequently, it has not been possible to assess its merits and, therefore, final findings in this regard have been determined in accordance with Article 28 of the basic Regulation.

• (137) It is also noted that the position of the decision of these banks to participate in the measures is not consistent with their own evaluation of the financial situation of Hynix, and in particular with the credit rating they gave it. The banks in question rated Hynix at the time of the measures between ‘precautionary’ and ‘doubtful’. No satisfactory explanation has been provided as to why in spite of the very critical situation of Hynix they still decided to continue with financial support.

• (138) It is noted that independent banks operating under market principles are required by their shareholders to evaluate the repayment probability of the loans they make and to take their decisions on whether to grant the loans or not on this basis. Such banks are required to follow objective lending criteria and must answer to shareholders when losses occur, as was the case with the Hynix loans. With respect to banks participating in the October 2001 measures, this can therefore only be explained by the fact that the decision makers were shielded from these consequences because of the high level of GOK ownership and factual control of the lending decisions. It should be further noted that the argument defending the merits of continuous lending to a heavily indebted company is particularly difficult to justify considering that the banks were at the same time swapping debt to equity, providing new financing and also raising the loss provisions for Hynix loans. This was based on their experience during 2001, when Hynix did not service its debts, as explained in recital 88. This indicates clearly that the banks, although aware that they would not recover their funding, provided it. This is not in line with lending practices based on commercial considerations.

• (139) Hynix and the GOK argued that Hynix's credit rating of ‘selective default’ was only a consequence of the restructuring programme and should not have been taken into account by the banks.

• (140) This argument does not appear to have any basis in fact. It is not reasonable to insist that a firm's credit rating is not the determining factor for potential investors. Moreover, the rating itself is more important than the reasons for it. Any potential investor planning to invest in a company examines its rating. In addition, following the logic argued by Hynix, the rating of Hynix should have been raised after the October 2001 restructuring measures were carried out. It never was, however, and remained at ‘selective default’. This still remains the case today. Clearly the rating agencies considered the whole situation of Hynix and investors operating under market conditions would have, and should have, taken into account the credit rating that applied to Hynix at the time of the October 2001 measures.

Calculation of the subsidyComments of the parties and assessment of the comments

• (141) As regards the calculation of the subsidy, Hynix raised five points in its comments. Firstly, Hynix argued that the debt-to-equity swap should not be included in the subsidy calculation. In addition, the amount of debt-to-equity swap of KRW 2,994 trillion included the loans of all 15 banks, not only the loans of the six banks found to have provided a subsidy. Therefore, the share of the banks not providing a subsidy should be deducted from the total amount of the debt-to-equity swap. This point was also raised by the GOK and Shinhan Bank.

• (142) After the verification of this point, it was concluded that the share of those banks not participating in the Option 1 measures should indeed be deducted. According to the information in the records, the amount to be deducted was KRW 511 billion. The total amount of the debt-to-equity swap considered as subsidy was corrected accordingly to KRW 2,483 trillion.

• (143) The second point raised by Hynix is that the value of extension of maturities and reduction of interest rates, KRW 1,586 trillion, is not correct, since it comprised alleged debt forgiveness and interest rate reduction of banks which were not found to provide a subsidy. The share of these banks should therefore be deducted from the total amount. This point was also raised by Shinhan Bank.

• (144) It is noted that an error has occurred on this point of calculation of the subsidy. According to information on the record the amount of loans subject to maturity extensions was KRW 1,825 trillion.

• (145) Thirdly, Hynix argued that the allocation method of the subsidy was incorrect. Since the October 2001 package only referred to the last three months of 2001, this should be taken into account when allocating the benefit over time.

• (146) The subsidy granted to Hynix in the context of the October 2001 package was a non-recurring subsidy. The amount of countervailable subsidy was calculated pursuant to Article 7(3) of the basic Regulation by allocating the amount of subsidy over the period of time which reflects to normal life of the assets in the industry concerned. According to point F(a)(ii) of the Guidelines, the allocation is done using the straight line method, which results in equal portions of a subsidy falling on each year, one of which being the investigation period.

• (147) The method advocated by Hynix is not reasonable, as it would artificially reduce the benefit conferred by the subsidy. If the amount of subsidy was allocated starting from October 2001 onwards, the duty would be in force for a longer period of time than when it is calculated under the normal calculation method laid down in the Guidelines, i.e. the beginning of the investigation period. A method referred to by Hynix would also lead to a situation in which each and every non-recurring subsidy granted during the investigation period would have its own individual countervailing duty period. This would be impossible to implement. In addition, if the numerator — the subsidy — covered only three months, the denominator — the recipient's turnover — would be reduced to cover the corresponding period. In this case, given the low level of sales in value terms in the last quarter of 2001, this could well increase the ad valorem rate of the subsidy. For these reasons, there are no grounds for changing the allocation method used in the provisional Regulation. Therefore, the conclusions set out in recital 108 of the provisional Regulation on the allocation of subsidy over time pursuant to the Guidelines are hereby confirmed.

• (148) The fourth argument of Hynix was that the interest rate added to the amount of subsidy was wrongly calculated. The interest rate should only have been added as regards the three last months of the year 2001, October to December, since this was the time when the benefit was available to Hynix.

• (149) The reasoning rejecting the argument of Hynix concerning the allocation of the subsidy over years in recital 147 applies also here. The linear allocation method distributes the benefit of all non-recurring subsidies received during the investigation period equally over the number of years in the allocation period. The annual commercial interest rate is added to this amount pursuant to point F(a)(ii) of the Guidelines, in order to have an equal denominator for the interest rate for all cases. Since the subsidy is allocated equally over the years, the interest rate follows the same method and an equal interest rate is calculated for all portions of subsidy resulting from the allocation. Any issue relating to timing of the grant during the investigation period is compensated by the fact that the subsidy is being allocated over the entire year's sales in order to determine the subsidy amount. For these reasons, there are no grounds for changing the calculation method for the interest rate used in the provisional Regulation. Therefore, the conclusions set out in recital 108 of the provisional Regulation on the calculation of the interest rate pursuant to the Guidelines are hereby confirmed.

• (150) As regards the amount of the countervailable subsidies, Hynix argued that the subsidy amounts should be allocated on the consolidated turnover of Hynix instead of the turnover of Hynix Semiconductor Inc., since the subsidies benefited Hynix and all its subsidiaries.

• (151) The practice of the European Community in anti-subsidy investigations is to allocate the subsidies over the turnover of the company investigated. The company is defined as a separate accounting entity. This is important, since the investigation targets subsidies granted in the exporting country benefiting a specific product or products. Only the information relevant for this product in the exporting country is investigated and verified by the investigating authorities. The European Community does not consider the consolidated turnover an appropriate basis, since it does not relate to the product concerned and often includes subsidiaries situated outside the exporting country which have not been investigated. It is noted that in the present case, all the Hynix subsidiaries included in the consolidated turnover are situated outside the Republic of Korea.

• (152) For these reasons, the subsidy amount shall be allocated on the turnover of Hynix and the conclusion on this point in recitals 67 and 108 of the provisional Regulation is hereby confirmed.

Final findings on the calculation of the subsidy of the October 2001 measures

New loan

• (153) In recital 105 of the provisional Regulation, the loan of KRW 658 billion was provisionally considered to be a grant pursuant to Section E(b)(v) of the Guidelines, since the providers of the loan did not expect to recover the loan at the time of granting it. Hynix's credit rating in October 2001 was ‘selective default’, it had a history of not having serviced its debt during 2001 and its liabilities exceeded six times its equity. In accordance with Article 6(b) of the basic Regulation, no alternative sources of financing were available in the market and Hynix had never even evoked such a possibility during the procedure. Under these circumstances it must have been evident to the banks concerned that they would not recover their money. The benefit to Hynix and the amount of countervailable subsidy is therefore the amount of the loan, KRW 658 billion. The conclusion of the provisional Regulation in recital 105 as regards the amount of the countervailable subsidy is therefore confirmed.

Extension of maturities and interest rate cuts of remaining loans

• (154) As regards assessing the benefit and the amount of subsidy granted by the roll-over of debt and the maturity extensions, it is firstly noted that according to the information on the record, roll-overs of debt are considered as new loans in Korea. The assessment of the situation should therefore be similar to that of the case of a new loan. In such cases, the financial situation of the company and its payments history would be the decisive criteria. If it was obvious to the bank at the time of the maturity extension that it would not recover its loan, regardless of the maturity extension, the loan subject to a maturity extension would be considered as a grant, as would a new loan under the same circumstances.

• (155) In the present case, the maturities of loans were extended by three years. As explained in recital 107, the situation of Hynix in October 2001 was so alarming that none of the lending banks had a reason to believe that they would ever recover their money. Due to this very critical situation of Hynix, considering the maturity extensions as grants appears to be the appropriate method of calculating the amount of subsidy.

• (156) Another possibility to calculate the amount of subsidy would be to compare the interest rate the company actually pays for the extended credit period to the rate it is supposed to pay in case it is granted an extension of maturity. It is noted that normally the interest rate is increased if the maturity is extended. This reflects the increased risk of the lender and the fact that it needs to be compensated for granting the extension. However, the reason why this method is not appropriate in the present case is the fact that for companies in severe difficulties, there is no reliable market benchmark interest rate: the lender is free to charge any interest rate it wants in case it decides to extend a maturity for such a borrower instead of executing its mature claim. Even infinite interest rates can be used.

• (157) In the present case the interest rate was lowered to 6 % for all loans subject to maturity extensions. This was extremely low, considering in particular that the syndicated loan granted to Hynix in January 2001 when it had the rating of BBB-, had an interest rate close to 13 %. Considering the fact that in October 2001 Hynix was rated SD and that it is normal practice for commercial banks to increase the interest rate in case of maturity extension, it can be concluded that a normal commercial lender could have very well used an infinite interest rate if it ever considered extending a maturity under such circumstances. Hynix would not have been able to sustain the cost of such interest rate. Since all the evidence in the record indicated that Hynix had no alternative sources of financing in the market than the October 2001 measures, the amount of loans subject to maturity extension, KRW 1,825 trillion, are considered as a grant pursuant to Article 6(b) of the basic Regulation and Section E(b)(v) of the Guidelines. The benefit to Hynix and the amount of countervailable subsidy is therefore the amount of the loans subject to maturity extensions, KRW 1,825 trillion.

• (158) It is noted that the accuracy of this conclusion is confirmed by the events that took place subsequently outside the investigation period, in 2002. In June 2002 CHB set aside loss reserves equivalent to 80 % of its exposure to Hynix and planned to provide fully for its exposure to Hynix by the end of year 2002 ([^E0059]). Other Hynix creditor banks also increased their loan loss reserves from 80 % to 100 % of outstanding loans to Hynix, clearly indicating that Hynix loans were unrecoverable ([^E0060]). The further bailout of Hynix which was carried out in December 2002, only a little more than a year after the October 2001 measures, confirmed these predictions. A further KRW 1,9 trillion of debt was swapped into equity in December 2002. It is noted that the amount of debt subject to maturity extension in October 2001 almost equals to this amount, being KRW 1,825 trillion.

Debt-to-equity swap

• (159) In recital 105 of the provisional Regulation the debt-to-equity swap was concluded to be considered as a grant pursuant to Section E(b)(v) of the Guidelines, since no market investor would have invested in Hynix shares at the time of the measure. In addition, the GOK-directed banks had also forgiven the same amount of outstanding debt to Hynix.

• (160) According to Article 5 of the basic Regulation, the amount of countervailable subsidies shall be calculated in terms of the benefit to the recipient which is found to exist during the investigation period of subsidisation. It is noted that the debt-to-equity swap was immediately booked in Hynix's 2001 financial accounts, wiping out a considerable amount of its debt. It is therefore reasonable to conclude that the benefit to Hynix equated to the full amount by which its debt was reduced. Its financial situation was immediately improved by that amount, a fact which can have a number of beneficial consequences, e.g. the removal of the cost of the servicing of the debt and using the saved money in investments instead. The same applies even if the intervention of the banks is viewed as a type of equity infusion. The benefit to Hynix is the full amount of that capital.

• (161) Apart from the question of the benefit conferred to Hynix by the measure, the situation from the point of view of the banks who transformed their loans into equity in the transaction should also be analysed. According to Section E(f)(i) and (ii) of the Guidelines, government provision of equity capital is not considered as conferring a benefit, unless the investment decision can be regarded as inconsistent with the usual investment practice of private investors in the exporting country concerned. The criterion to assess the benefit is whether a private investor would have put money into the company in the same situation in which the government provided equity.

• (162) It is noted that Hynix's rating was at selective default and its financial situation was critical. Furthermore, the DRAM prices had plummeted from June 2001 onwards and in October 2001 they were at their lowest point ever. The stock price of Hynix had also fallen 72 % from June to September. This development just before the October 2001 measures indicated clearly that investment in Hynix shares was extremely risky and likely to result in considerable losses. Under these circumstances, it can be concluded that no normal market investor would have invested in Hynix shares in October 2001.

• (163) According to Section E(f)(vii) of the Guidelines, the equity provided can be considered as a grant where it can be determined that the government had no intention of receiving any return on its investment and was in effect giving a disguised grant to the company in question. In the present case the GOK-directed banks forgave a considerable amount of their outstanding loans to Hynix, a measure considered as a grant pursuant to Section E(b)(v) of the Guidelines. This measure conferred a benefit to Hynix which equals to the amount of loans forgiven. The fact that the banks received equity in the transaction does not change this assessment, since considering both the market situation and the situation of Hynix at the time of the measures, no other investor would have bought Hynix shares and the banks could not expect to receive any return to their investment under those circumstances.

• (164) For the reasons explained above, the amount of debt-to-equity swap provided by the GOK-directed banks is considered as grant pursuant to Articles 6(a) and (b) of the basic Regulation and Sections E(f)(v) and (vii) of the Guidelines. The benefit conferred to Hynix by this measure, and the amount of countervailable subsidy, is the amount of the debt-to-equity swap provided by the GOK-directed banks, KRW 2,483 trillion.

• (165) It is noted that the accuracy of this conclusion is confirmed by the events that took place subsequently outside the investigation period, in 2002. In December 2002 the Hynix creditor banks carried out the third bailout of the company in two years. The measures included roll-over of KRW 3,2 trillion of debts and debt-to-equity swap of KRW 1,9 trillion. Hynix's total liabilities had increased to more than USD 7,9 trillion despite the two earlier bailouts. The capital write-down plan combined every 21 shares into one share. This was done to mitigate the capital erosion of approximately KRW 22 trillion that had resulted from accumulated deficit and discount on capital stocks ([^E0061]). The banks pointed out that they had already set aside loan loss reserves equivalent to 80 % to 100 % of the loans extended to Hynix ([^E0062]). This situation, and in particular the magnitude of the capital write-down of 21:1, clearly indicates that Hynix equity can be considered worthless and the investment in it as loss.

• (166) The calculation of the amount of subsidy has been revised on the basis of the comments of the parties and the new information that has given rise for changing the provisional findings as regards some investigated measures. In order to avoid double-counting, those subsidies which were included in the October 2001 measures have been deducted from them. The net subsidy amount of the October 2001 measures is therefore 19,4 % ad valorem. Considering also the conclusions in recitals 47, 80 and 98, the total amount of the subsidy is the following:

DEFINITION OF THE COMMUNITY INDUSTRY

• (167) Following the imposition of provisional measures no comments were received on the definition of the Community industry as set out in recitals 110 and 111 of the provisional Regulation. It is therefore confirmed that the two cooperating Community producers representing 100 % of the total Community production of the product concerned during the IP constitute the Community industry within the meaning of Article 9(1) of the basic Regulation.

INJURYPRELIMINARY REMARKS

• (168) As set out in recital 115 of the provisional Regulation, for reasons of confidentiality, indices are used where necessary to show the evolution of trends. Following the publication of the provisional Regulation Hynix claimed that whilst the use of indices to describe its data is justified for confidentiality reasons, the mere disclosure in the provisional Regulation of indices in the analysis of the situation of the Community industry did not allow Hynix to assess with a sufficient degree of certainty the situation of the Community industry. Hynix further argued that there were no reasons of confidentiality not to disclose to it the actual figures of the Community industry, since the Community industry comprises two companies. It is noted that Hynix's claim was of a general nature, without specifying which data it would like to review in order to allow it to assess the situation of the Community industry.

• (169) However, Article 29(5) of the basic Regulation prohibits disclosure of information for which confidential treatment has been requested and good cause has been shown for such treatment. Furthermore, since the Community DRAM market is highly concentrated with very few producers controlling the market, it would be easy for Hynix to deconstruct any cumulative figures disclosed to it so as to acquire an accurate understanding of confidential business information of the two Community producers composing the Community industry. Such information would be of great commercial value to Hynix and knowledge of it by third parties could harm the Community industry. Finally, the use of indices to show the development of the injury indicators over time is appropriate in a market in which there are only two producers. These indices allow Hynix, on the basis of its own Community market knowledge and the information placed in the file available for inspection by interested parties to analyse accurately the information provided and determine with a sufficient degree of certainty the situation of the Community industry. Accordingly, it is considered that Hynix's claim in this respect is not justified and is therefore rejected.

• (170) No further comments were received on the preliminary remarks as set out in recitals 112 to 115 of the provisional Regulation. Those preliminary remarks are hereby confirmed.

COMMUNITY CONSUMPTION

• (171) The total figure for Community consumption of the product concerned corresponds to the total imports plus all sales in the Community produced by the Community industry.

• (172) In the absence of any new information, the provisional findings concerning Community consumption as described in recital 117 of the provisional Regulation are confirmed. Throughout the period under consideration Community consumption of DRAMs increased by 316 % as follows:

IMPORTS OF THE PRODUCT CONCERNED INTO THE COMMUNITYVolume of imports from Korea

• (173) Following the publication of the provisional Regulation, Hynix claimed that its import volume of the product concerned expressed in '000 Mbits increased at a much lower pace than was described in recital 118 of the provisional Regulation. Hynix argued that imports and market share (see below) should be based on the total sales of its subsidiaries in the Community (resales) of the product concerned to unrelated customers in the Community during the period under consideration and not on the volume of its imports of DRAMs produced in Korea. It is noted that these resales as reported by Hynix were significantly higher (up to 60 %) than the actual imports also reported by Hynix during the first two years of the period under consideration. No satisfactory explanation was given for this big discrepancy. It is considered that imports should comprise all sales of DRAMs produced by Hynix in Korea and sold to its subsidiaries in the Community, plus some minor sales to unrelated customers in the Community via a third country.

• (174) Hynix, supported by the GOK, also argued that, in establishing the trend of its imports into the Community over the period under consideration, the imports of the product concerned into the Community from LG Semicon Co., Ltd that occurred before the date of the merger of this company with Hyundai Electronics Industries Co., Ltd (Hynix since 29 March 2001) should have been included in its import data. Hynix and the GOK provided certain ad hoc data in this respect, i.e. two figures concerning imports in Mbits from LG Semicon Co., Ltd in 1998 and 1999. Otherwise, it was claimed that imports form Hynix should be considered for injury purposes only following the date of its merger with LG Semicon Co., Ltd in 1999.

• (175) However, LG Semicon Co., Ltd was a distinct legal entity before Hyundai Electronics Industries Co., Ltd acquired it from LG Electronics Inc. and its related companies on 7 July 1999. After the acquisition, LG Semicon Co., Ltd was renamed Hyundai Micro Electronics Co., Ltd and this company merged with Hyundai Electronics Industries Co., Ltd on 13 October 1999. The acquisition of LG Semicon Co., Ltd was therefore a new investment for Hyundai Electronics Industries Co., Ltd without any retroactive effects on its position on the market. In other words, Hyundai Electronics Industries Co., Ltd, predecessor of Hynix, which was the only company found to have benefited from subsidies, increased its production capacity. Furthermore, no party, including Hynix, has reported any imports from LG Semicon Co., Ltd for 1998 and 1999 in its questionnaire response or any other submission and as such were not verified by the Commission. LG Semicon Co., Ltd data cannot constitute part of Hynix's subsidised imports. For the reasons set out above, the claim made by Hynix and the GOK is rejected.

• (176) The volume of imports from Korea as described in recital 118 of the provisional Regulation was established on the basis of the aggregation of duly verified import data provided by Samsung and Hynix in their respective questionnaire responses. The volume of imports from Korea thus established increased by 331 % during the period under consideration. It is confirmed that imports from Hynix increased faster, i.e. by 361 % during the period under consideration.

• (177) Even if import data of LG Semicon Co., Ltd for 1998 and 1999 as provided ad hoc by Hynix and the GOK, and which could not be verified, were taken into account, they should be added to the volume of imports from Korea and not to the volume of imports from Hynix. The volume of imports from Korea would then have increased by 219 % during the period under consideration.

Prices of imports from Korea and Hynix

• (178) Import prices were calculated using data on import quantities and values provided in the questionnaire responses of the exporting producers in Korea. Since no comments were received on import prices following the publication of the provisional Regulation, the provisional findings concerning import prices as described in recital 119 of the provisional Regulation are confirmed. Throughout the period under consideration import prices developed as follows:

• (179) With regard to the provisional findings on price undercutting as described in recital 122 of the provisional Regulation, Hynix and the GOK claimed that the reference to the fact that price undercutting took place for some transactions is irrelevant since price undercutting, in accordance with usual practice, is determined on a weighted average basis. Hynix further argued that in order to conclude whether its prices undercut a number of specific transactions of the Community industry, either a transaction-by-transaction comparison should be made or a daily comparison taking also into account, where possible, the time within the day.

• (180) However, there is no provision in the basic Regulation which stipulates that price undercutting should be calculated on a weighted average or any other specific basis. According to Article 8(3) of the basic Regulation it should be considered, when assessing the effect of subsidised imports on prices, whether price undercutting has been ‘significant’. It does not set out any requirement relating to the calculation of the margin of undercutting, nor does it provide for a particular methodology to be followed in this respect. Moreover, according to the same Article, price undercutting is not alone determinative in an injury determination, rather it forms part of the overall assessment of injury to the Community industry and is conducted so as to provide guidance in the context of the assessment of injury and causation.

• (181) Furthermore, it is noted that the DRAM market is very transparent and characterised by substantial price competition. Indeed fixed costs are very high. Suppliers need to develop sufficient economies of scale and strive to maintain market shares. Price undercutting will occur in specific competitive situations with specific customers. But, if one supplier offers a price to a Community customer that undercuts the price of a DRAM sold by a Community producer, competitive pressure quickly eliminates such undercutting. Thus, in such markets, big customers can force competing suppliers to meet any lower price offering. It is therefore difficult to establish price undercutting over a given period. Such price depressing situations prevented the Community producers from increasing prices, something which would otherwise have occurred.