tutorial video tutorial video Discover the CiteMap in 3 minutes
Source(s) of the information:
Source(s) of the information:

Lawyers, other representatives, expert(s), tribunal’s secretary

Report of the Panel

TABLE OF CASES CITED IN THIS REPORT

Short TitleFull Case Title and Citation
Brazil – Aircraft (Article 21.5 – Canada) Panel Report, Brazil – Export Financing Programme for Aircraft, Recourse by Canada to Article 21.5 of the DSU ("Brazil – Aircraft (Article 21.5 – Canada) "), WT/DS46/RW, adopted 4 August 2000, as modified by the Appellate Body Report, WT/DS46/AB/RW, DSR 2000:IX, 4093.
Brazil – Aircraft (Article 21.5 – Canada II Panel Report, Brazil – Export Financing Programme for Aircraft, Second Recourse by Canada to Article 21.5 of the DSU ("Brazil – Aircraft (Article 21.5 – Canada II) "), WT/DS46/RW/2, adopted 23 August 2001, DSR 2001:XI, 5481.
Canada – Aircraft Credits and Guarantees Panel Report, Canada – Export Credits and Loan Guarantees for Regional Aircraft ("Canada – Aircraft Credits and Guarantees"), WT/DS222/R and Corr.1, adopted 19 February 2002.
Canada – Aircraft (Article 21.5 – Brazil) Appellate Body Report, Canada – Measures Affecting the Export of Civilian Aircraft, Recourse by Brazil to Article 21.5 of the DSU ("Canada – Aircraft (Article 21.5 – Brazil) "), WT/DS70/AB/RW, adopted 4 August 2000, DSR 2000:IX, 4299.
EC – Sugar Exports (Australia) GATT Panel Report, European Communities – Refunds on Exports of Sugar – Complaint by Australia, adopted 6 November 1979, BISD 26S/290.
EC – Sugar Exports (Brazil) GATT Panel Report, European Communities – Refunds on Exports of Sugar – Complaint by Brazil, adopted 10 November 1980, BISD 27S/69.
EEC – Wheat Flour Subsidies GATT Panel Report, European Economic Community – Subsidies on Export of Wheat Flour ("EEC – Wheat Flour Subsidies"), 21 March 1983, unadopted, SCM/42.
Indonesia – Autos Panel Report, Indonesia – Certain Measures Affecting the Automobile Industry ("Indonesia – Autos"), WT/DS54/R, WT/DS55/R, WT/DS59/R, WT/DS64/R and Corr.1, 2, 3, 4, adopted 23 July 1998, DSR 1998:VI, 2201.
Japan – Alcoholic Beverages II Appellate Body Report, Japan – Taxes on Alcoholic Beverages ("Japan – Alcoholic Beverages II"), WT/DS8/AB/R, WT/DS10/AB/R, WT/DS11/AB/R, adopted 1 November 1996, DSR 1996:I, 97.
US – Corrosion-Resistant Steel Sunset Review Appellate Body Report, United States – Sunset Review of Anti-Dumping Duties on Corrosion-Resistant Carbon Steel Flat Products from Japan ("US – Corrosion-Resistant Steel Sunset Review"), WT/DS244/AB/R, adopted 9 January 2004.
US – Countervailing Measures on Certain EC Products Appellate Body Report, United States – Countervailing Measures Concerning Certain Products from the European Communities ("US – Countervailing Measures on Certain EC Products"), WT/DS212/AB/R, adopted 8 January 2003.
US ‑ Export Restraints Panel Report, United States – Measures Treating Exports Restraints as Subsidies ("US – Export Restraints"), WT/DS194/R and Corr.2, adopted 23 August 2001, DSR 2001:XI, 5767.
US – FSC Appellate Body Report, United States – Tax Treatment for "Foreign Sales Corporations" ("US – FSC"), WT/DS108/AB/R, adopted 20 March 2000, DSR 2000:III, 1619.
US – Norwegian Salmon CVD GATT Panel Report, Imposition of Countervailing Duties on Imports of Fresh and Chilled Atlantic Salmon from Norway ("US – Norwegian Salmon CVD"), adopted 28 April 1994, BISD 41S/II/576.
US – Section 211 Appropriations Act Appellate Body Report, United States – Section 211 Omnibus Appropriations Act of 1998 ("US – Section 211 Appropriations Act"), WT/DS176/AB/R, adopted 1 February 2002.
US – Upland Cotton Panel Report, United States – Subsidies on Upland Cotton ("US – Upland Cotton"), WT/DS267/R, and Corr.1, 8 September 2004.
US – Lead and Bismuth II Panel Report, United States – Imposition of Countervailing Duties on Certain Hot-Rolled Lead and Bismuth Carbon Steel Products Originating in the United Kingdom ("US – Lead and Bismuth II"), WT/DS138/R and Corr.2, adopted 7 June 2000, as upheld by the Appellate Body Report, WT/DS138/AB/R, DSR 2000:VI, 2623.
US – Lead and Bismuth II Appellate Body Report, United States – Imposition of Countervailing Duties on Certain Hot-Rolled Lead and Bismuth Carbon Steel Products Originating in the United Kingdom ("US – Lead and Bismuth II"), WT/DS138/AB/R, adopted 7 June 2000, DSR 2000:V, 2595.

I. introduction

A. complaint of the european communities

1.1.
On 21 October 2002, the European Communities requested consultations with Korea pursuant to Article 4 of the Dispute Settlement Understanding ("the DSU"), Article XXIII:1 of the General Agreement on Tariffs and Trade 1994 ("GATT 1994"), and Articles 4, 7 and 30 of the Agreement on Subsidies and Countervailing Measures ("the SCM Agreement"), with regard to measures affecting trade in commercial vessels.1
1.2.
The European Communities and Korea held the requested consultations on 22 November and 13 December 2002, and 7 May 2003, but failed to reach a mutually satisfactory resolution of the matter.
1.3.
On 11 June 2003, the European Communities requested the establishment of a panel to examine the matter.2
1.4.
On 10 July 2003, the European Communities requested that the above request be placed on the agenda of the meeting of the Dispute Settlement Body ("the DSB") scheduled for 21 July 2003. The European Communities further requested that, at the same meeting, the DSB initiate the procedures provided for in Annex V of the SCM Agreement pursuant to paragraph 2 of that Annex.

B. establishment and composition of the panel

1.5.
At its meeting of 21 July 2003, the DSB established a panel in accordance with Article 6 of the DSU and pursuant to the request made by the European Communities in document WT/DS273/2.
1.6.
At that meeting, the parties to the dispute also agreed that the Panel should have standard terms of reference. The terms of reference are, therefore, the following:

"To examine, in the light of the relevant provisions of the covered agreements cited by the European Communities in document WT/DS273/2, the matter referred to the DSB by the European Communities in that document, and to make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in those agreements."

1.7.
On 11 August 2003, the European Communities requested the Director-General to determine the composition of the Panel, pursuant to paragraph 7 of Article 8 of the DSU. This paragraph provides:

"If there is no agreement on the panelists within 20 days after the date of the establishment of a Panel, at the request of either party, the Director-General, in consultation with the Chairman of the DSB and the Chairman of the relevant Council or Committee, shall determine the composition of the panel by appointing the panelists whom the Director-General considers most appropriate in accordance with any relevant special or additional rules or procedures of the covered agreement or covered agreements which are at issue in the dispute, after consulting with the parties to the dispute. The Chairman of the DSB shall inform the Members of the composition of the panel thus formed no later than 10 days after the date the Chairmen receives such a request."

1.8.
On 20 August 2003, the Director-General accordingly composed the Panel as follows:

Chairman:Mr. Said El-Naggar

Members: Mr. Gilles Gauthier

Ms. Ana Novik Assael

1.9.
China, Japan, Mexico, Norway, Chinese Taipei and the United States reserved their third-party rights.3
1.10.
On 11 April 2004, Mr. El-Naggar, Chairman of the Panel, passed away. On 6 May 2004, the parties jointly requested that the Director-General appoint a new Chairman to the Panel. On 11 May 2004, the Director-General appointed Mr. Julio Lacarte-Muró as the new Chairman of the Panel.4

C. information gathering procedure under annex v of the scm agreement

1.11.
In its 10 July 2003 communication to the DSB requesting initiation of the information-gathering procedure under Annex V of the SCM Agreement, the European Communities stated that in order to facilitate the DSB's task of designating a representative pursuant to paragraph 4 of Annex V, it had proposed names and consulted with Korea. The European Communities indicated that it and Korea had not reached agreement in this respect, and thus requested that the DSB designate a representative to facilitate the information-gathering procedure.5 At its meeting of 21 July 2003, the DSB designated Mr. András Szepesi as its representative for this purpose.
1.12.
The date of 21 July 2003 was considered by the parties to be the date on which the matter was "referred to the DSB" in the sense of paragraph 5 of Annex V, such that the 60-day period established in that paragraph for completion of the information-gathering process would have ended on 19 September 2003. The parties agreed that the complaining party's first submission should be due six weeks after the end of the Annex V procedure, and the responding party's first submission six weeks after that, so that all information developed through the Annex V procedure could be used in the preparation of these submissions. The Panel established its timetable accordingly.
1.13.
At a late stage in the 60-day period the 19 September 2003 date was modified, with the agreement of the parties, to allow the parties additional time to translate certain voluminous documentation into one of the three WTO working languages. The amended deadline for completion of the Annex V procedure was 10 November 2003. The Panel was immediately informed of these modifications and the causes thereof. The Panel revised its timetable accordingly.
1.14.
On 10 November 2003, the Designated Representative submitted his report to the Panel. This report is set forth in Attachment 1.

D. additional procedures for the protection of business confidential information

1.15.
As indicated in the report of the Designated Representative the Panel, at the request of the parties, adopted additional procedures for the protection of business confidential information. These procedures ("the BCI procedures"), are set forth in Attachment 2.

E. panel proceedings

1.16.
The Panel met with the parties on 9-10 March 2004, and 17-18 June 2004. The Panel met with third parties on 9 March 2004.
1.17.
The Panel submitted its Interim Report to the parties on 24 November 2004. The Panel submitted its final report to the parties on 22 December 2004.

II. FACTUAL ASPECTS

2.1.
This dispute concerns the following measures, alleged by the European Communities to constitute prohibited subsidies and/or actionable subsidies in the sense of Parts II and III of the SCM Agreement:

· The Act Establishing the Export-Import Bank of Korea ("KEXIM"), any implementing decrees and other regulations, alleged to specifically allow and enable KEXIM to provide Korean exporters of capital goods with financing at preferential rates.

· The pre-shipment loan ("PSL") and advance payment refund guarantee ("APRG") schemes established by KEXIM.

· The individual granting of pre-shipment loans and advance payment refund guarantees by KEXIM to Korean shipyards, including Samho Heavy Industries ("Samho-HI" or "SHI"), Daedong Shipbuilding Co. ("Daedong"), Daewoo Heavy Industry ("DHI"), Daewoo Shipbuilding and Marine Engineering ("Daewoo-SME", or "DSME"), Hyundai Heavy Industries ("Hyundai-HI", or "HHI"), Hyundai Mipo ("MIPO"), Samsung Heavy Industries ("Samsung") and Hanjin Heavy Industries & Construction Co ("Hanjin").

· Corporate restructuring measures including debt forgiveness, debt and interest relief and debt-to-equity swaps, affecting Daewoo-SME, Samho-HI, and Daedong) .

· The Special Tax Treatment Control Law ("STTCL"), in particular the special taxation on in-kind contribution (Article 38) and the special taxation on spin-off (Article 45-2) scheme.

III. parties' requests for findings and recommendations

A. the european communities

3.1.
The European Communities asks the Panel to find that Korea has granted subsidies inconsistent with its obligations under the SCM Agreement, because:

"Korea, through the KEXIM Act, KEXIM Decree and Interest Rate Guidelines provides prohibited subsidies, inconsistent with Article 3.1 and 3.2 of the SCM Agreement";

"Korea, through the establishment and maintenance of the APRG and preshipment loan programmes provides prohibited subsidies, inconsistent with Article 3.1 and 3.2 of the SCM Agreement";

"Korea, through individual grants of APRGs and preshipment loans provided prohibited subsidies, inconsistent with Article 3.1 and 3.2 of the SCM Agreement";

"Korea, by providing subsidies to Daewoo-SME/Daewoo-HI, Samho-HI/Halla-HI, and STX/Daedong through (i) workout plans and restructuring plans; (ii) tax concessions provided to Daewoo-HI/Daewoo-SME; and (iii) the grant of KEXIM APRGs and pre-shipment loans, has caused serious prejudice to the interests of the European Communities in violation of Articles 5(c) and 6.3(c) of the SCM Agreement."

3.2.
The European Communities considers that the above violations of the SCM Agreement have nullified and impaired benefits accruing to it under the Marrakesh Agreement Establishing the World Trade Organization ("WTO Agreement") and accordingly asks the Panel to recommend that Korea withdraw these subsidies or remove the adverse effects of the actionable subsidies in accordance with Articles 4.7 and 7.8 of the SCM Agreement.

B. korea

3.3.
Korea requested the Panel to issue a number of preliminary rulings. The Panel's reasoning and conclusions in respect of Korea's requests for preliminary rulings are set forth in section VII.A, infra.
3.4.
Korea also requests the Panel to dismiss all of the claims of the European Communities.

IV. ARGUMENTS OF THE PARTIES

4.1.
The arguments of the parties are set out in their submissions to the Panel. The parties' executive summaries of their submissions are attached as Annexes to this report (see List of Annexes, page viii), and constitute an integral part of this Report.

V. ARGUMENTS OF THE THIRD PARTIES

5.1.
The arguments of third parties China, Japan, Norway, Chinese Taipei and the United States, are set out in their submissions to the Panel and are attached to this Report as Annexes (see List of Annexes, page viii), and constitute an integral part of this Report. Mexico made no submissions to the Panel.

VI. Interim Review

6.1.
On 24 November 2004, we submitted the interim report to the parties. Korea submitted a written request for interim review of certain aspects of the interim report. The EC did not request interim review. The EC also submitted written comments on Korea's request. Neither party requested an interim review meeting.
6.2.
Korea's communication concerning interim review contained two parts, a cover letter containing general comments, and an annex containing specific comments on certain identified paragraphs of the interim report. We address these two parts of Korea's comments separately, in sections VI.A and VI.B, respectively.
6.3.
We also have made certain technical revisions and corrections to the report.

A. Korea's general comments in its cover letter

6.4.
The cover letter to Korea's submission requesting review of aspects of the Panel's interim report sets forth "some reservations regarding some procedural issues and certain aspects of the prohibited subsidies analysis of the Panel". We note at the outset, however, that these comments are quite general, lacking specific references to particular paragraphs or sections of the interim report, and failing to indicate any changes that Korea requests the Panel make to the report. We therefore doubt whether these comments meet the requirement of Article 15.2 that a party's request for interim review identify "precise aspects" of the report. In any event, as discussed below, we see no need to make any changes to the report on the basis of the comments in Korea's cover letter.
6.5.
Korea states that the Panel misallocated the burden of proof "in places", shifting the burden from the EC to Korea, but identifies no specific place in which this alleged misallocation of burden occurred. We believe that the appropriate burden of proof was maintained in respect of both parties throughout the dispute. Korea also argues that the EC was permitted to introduce new factual information at a late stage of the proceedings, not as rebuttals but to establish wholly new points. We disagree with Korea's implication that it lost due process rights, as we recall that at various stages of the proceedings, the Panel requested certain information from both parties, and the parties requested certain information from each other, and that after each of these submissions, each party was given full opportunity to comment on the new factual information submitted by the other party. Concerning Korea's general objection to the benchmarks offered by the EC for APRG transactions, and the Panel's analysis of country risk spread, we believe that our findings adequately explain our overall approach and reasoning.6 As for Korea's general disagreement with our approach to defining a public body in the sense of Article 1.1(a) (1), as discussed in our findings, in our view Korea's approach confuses the concepts of financial contribution and benefit. Concerning actionable subsidies, Korea states that it disagrees with certain of our conclusions, but provides no specifics whatsoever. Finally, concerning Korea's statement that the EC "was permitted without remark to manifestly abuse the Annex V process", we recall that Korea originally raised this issue in its first written submission, and that we ruled on it on 12 March 2004.7

B. Korea's specific comments on certain identified paragraphs of the interim report

Footnote 75 (footnote 77 of the Final Report)

6.6.
Korea requested that we change the reference to "Exhibit EC – 21" to refer to "Exhibit EC – 26". We have made the change requested by Korea.

Paragraphs 7.136 and 7.137

6.7.
Korea argues that Exhibits KOREA – 58 and 59 were provided in response to Question 69 from the Panel, which sought "an example (with supporting documentation) of two instances in which different Korean shipyards were not able to select the APRG provider itself". Korea asserts that the Panel did not ask Korea to rebut the specific APRG benchmarks offered by the EC. Korea supposes that the Panel adopted this approach because it was looking at the issue from a systemic general point of view. Korea asserts that if it showed that there were questions about the choice of source of APRGs, it would be up to the EC to show who made that choice, and establish that the benchmarks proposed by the EC were appropriate. Korea asserts that had the Panel demanded that Korea rebut the specific instances offered by the EC, the Panel should not have asked for examples in a manner that indicated to Korea that the Panel was taking a general perspective. Korea submits that, by doing so, the Panel prohibited Korea from offering evidence that it might have been able to generate to rebut specific instances rather than just providing examples as the Panel requested.
6.8.
Korea also disagrees with the Panel's statement that the examples in Exhibit KOREA – 83 "are APRGs provided to shipyards that fall outside the scope of the EC's prohibited export subsidies claims". Korea asserts that there were no limitations on the scope of the EC's claims. Korea submits that the information in Exhibit KOREA – 83 is therefore relevant, probative and directly responsive to Question 69 from the Panel.
6.9.
The EC submits that Korea's comment is based on the erroneous view that Korea needed only to rebut specific evidence advanced by the European Communities where specifically "demanded" to do so by the Panel and that, as a result, the Panel "prohibits Korea from offering evidence that it might have been able to generate to rebut specific instances rather than just providing examples as the Panel requested."
6.10.
Regarding Exhibit KOREA – 83, the EC asserts that Korea misrepresents or misunderstands the Panel’s reasoning. The EC argues that the Panel is dealing in this part of the Report with the EC's complaint that individual APRG transactions constituted export subsidies. The EC asserts that it is therefore irrelevant that the Panel accepted the admissibility of the EC's per se claim against KEXIM’s export subsidy regimes. The EC suggests that the Panel could take account of Korea’s point and avoid others misunderstanding the issue in the same way by amending the penultimate sentence to paragraph 7,137 to read:

The evidence contained in Exhibit Korea-83 relates to APRGs provided to shipyards that fall outside the scope of the EC's these prohibited export subsidy claims.

6.11.
First, we note that para. 213 of Korea's First Written Submission states that:

Shipyards do not always select the APRG provider by itself. Sometimes, they are compelled to make use of the financial institutions, domestic or foreign, designated by the ship owners for issuing the APRGs regardless of whether the premium rates by such institutions are higher than those offered by other financial institutions. (emphasis supplied)

6.12.
Korea's own submission therefore makes it clear that Korea was not approaching this issue from a "general perspective". Instead, Korea argued that APRG providers were "sometimes" designated by buyers. Korea's argument could not, therefore, have operated as a general defence, or response, to the benchmarks proposed by the EC. Rather, the onus was on Korea, as the party alleging a fact, to prove that the providers of the specific APRGs identified by the EC were designated by the relevant buyers of the ships in question.
6.13.
Second, it was up to Korea to decide how it wished to respond to the claims, arguments and evidence of the EC. It was not up to the Panel to instruct Korea in this regard. Nor did the Panel "prohibit" Korea from offering any evidence that it chose. Korea had numerous opportunities to present whatever evidence it wished in response to the claims, arguments and evidence presented by the EC. The Panel addressed Question 69 to Korea in order to clarify Korea's argument regarding buyer designation. It did not do so in order to dictate how Korea should respond to the claims, arguments and evidence presented by the EC. The fact that the Panel phrased Question 69 in a particular manner did not prevent Korea from deciding for itself how it wished to respond to the claims, arguments and evidence presented by the EC.
6.14.
Regarding Exhibit KOREA – 83, we note that our findings regarding individual APRG transactions were necessarily limited to those transactions specifically identified by the EC. We had no basis to make findings in respect of other individual APRG transactions for which no evidence had been submitted by the EC. We have made the change suggested by the EC in order to clarify this matter.

Paragraph 7.167

6.15.
Korea asserts that the value of Yangdo Dambo is (at least) 50 per cent of the value of Physical Collateral such as government bonds. Korea submits that it clearly showed that differences existed as regards different types of collaterals and that the value of the Yangdo Dambo exceeds that of cash deposits for foreign APRGs which was 10 to 30 per cent of the amount covered by the APRGs in terms of collateral value.
6.16.
The EC asserts that Korea's comment is addressed in para. 7,169 and footnote 95 of the Interim Report.
6.17.
We maintain the view we expressed in footnote 95 that "at a certain point the value of a smaller portion of credit coverage by a stronger form of collateral will equal, or exceed, the value of a larger portion of credit coverage by a weaker form of collateral". We also recall our statement in para. 7,167 that, in order for Korea's argument to prevail, "Korea would need to demonstrate that the collateral value of the Yangdo Dambo exceeds that of the cash deposits". Korea has failed to show this, since Korea has failed to establish the collateral value of the cash deposits, i.e., it has failed to establish how much the credit spread was adjusted in order to reflect the value of the 10 – 30 per cent cash deposits.

Paragraph 7.234

6.18.
Korea asserts that, contrary to the Panel's finding in para. 7,234, it did provide sufficient information to rebut the EC's argument regarding the [BCI: Omitted from public version] guarantee.Korea refers in this regard to its response to Question 17 from the EC, read in light of the information provided by the EC in Exhibit EC – 118.
6.19.
The EC asserts that the evidence referred to by Korea is not sufficient to rebut the EC's argument.
6.20.
Korea's reply to Question 17 from the EC refers to a "payment guarantee". It does not provide the value of that guarantee. Nor does it indicate the terms of that guarantee. Nor is this information to be found in Exhibit EC – 118. Accordingly, there is no basis for us to change our finding in para. 7,234.

Paragraph 7.237

6.21.
Korea asserts that there is a difference in the value of the types of collateral referred to in para. 7,237 (i.e., factory and Yangdo Dambo). Korea asserts that the value of the factory is greater than the value of the Yangdo Dambo.
6.22.
Korea's comment does not address the issue raised in para. 7,237 of the Interim Report. The fact that the different types of collateral may have different values is not denied in para. 7,237. Instead, we indicate that we have no information regarding what those values might be. Korea's comment does not address this issue, since it does not identify any evidence submitted during the Panel proceedings regarding the values of the types of collateral at issue. In any event, we note that para. 7,237 also provides additional reasons (unrelated to the value of the relevant collaterals) for rejecting the STX/Daedong corporate bond data submitted by Korea.

Paragraph 7.240

6.23.
Korea asserts that it provided information regarding the interest rates for the Hyundai/Mipo corporate bonds at para. 238 of its First Written Submission.
6.24.
We do not disagree with Korea. However, Korea's comment does not address the issues raised in para. 7,240 of the Interim Report. In particular, Korea's comment does not point to record evidence indicating whether or not the relevant bonds were guaranteed or collateralized.

Paragraph 7.243

6.25.
Korea asserts that the reference to "collateral" in Exhibit KOREA – 22 was an error. Korea identifies record evidence indicating that the HHI corporate bonds were not collateralized.
6.26.
The EC does not respond to Korea's comment. In the absence of any objection by the EC, and on the basis of the explanation provided by Korea, we shall amend para. 7,243 of the Report to remove references to the issue of collateralization. However, we note that para. 7,243 will still reflect the fact that the bond rates cannot be used as market benchmarks because of differences in maturity.

Footnote 154 (footnote 156 of the Final Report)

6.27.
Korea asserts that the Panel failed to acknowledge that joint and several guarantees do have certain values. Korea asserts that the fact that joint and several guarantees have certain values is clear from paragraphs 15 and 16 of Exhibit KOREA – 90.
6.28.
The Panel did not state that joint and several guarantees do not have any value. Rather, in footnote 154 the Panel quoted Korea's explicit statement that KEXIM treated such guarantees as if they had no value. Korea has not denied this. The Panel also stated that "[i]n any event, since Korea has failed to quantify the alleged value of such collateral, there is no basis for us to make any adjustment". This statement remains valid, since Korea has failed to identify any evidence regarding the value of the joint and several guarantees at issue. Paragraphs 15 and 16 of Exhibit KOREA – 90 do not indicate the value of such guarantees.

Footnote 174 (footnote 176 of the Final Report)

6.29.
We have corrected a clerical error identified by Korea.

Paragraphs 7.290 and 7.291

6.30.
Korea asserts that it did establish that the value of collateral was reflected in KEXIM's interest rate calculations. Korea also asserts that it identified Attachment 1 to the KEXIM Interest Rate Guidelines as the statutory basis for the application of different credit risk spreads depending on the types of security interests provided.
6.31.
We note that Attachment 1 to the KEXIM Interest Rate Guidelines does not contain any reference to [BCI: Omitted from public version]. Nor was any other evidence presented by Korea regarding the amount of any interest rate adjustment made in respect of such collateral. Accordingly, there is no need for us to amend our statement (para. 7,292) that "[e]ven if an additional adjustment were necessary, therefore, there is no basis for the Panel to determine what exactly that adjustment should be".

Footnote 182 (footnote 184 of the Final Report)

6.32.
Korea asserts that the references to "Samho" should be replaced by "STX/Daedong".
6.33.
We disagree with Korea's comment, as the footnote is dealing with the EC's treatment of Samho PSLs in Figure 17 of the EC's first Written Submission. This is what Korea is referring to in para. 235 of its First Written Submission. We have slightly amended footnote 184 of the Final Report to clarify this point.

VII. FINDINGS

A. korean requests for preliminary rulings

1. 29 August 2003 request for suspension of the Annex V information-gathering procedure

7.1.
On 29 August 2003, Korea requested the suspension of the Annex V information-gathering procedure. On 3 September 2003, the Panel sent the following communication to the parties regarding this matter:

I am writing to you in respect of Korea's request for preliminary rulings dated 29 August 2003. This letter concerns only Korea's request regarding the suspension or adaptation of the Annex V procedure pending issuance of the preliminary rulings sought by Korea. It does not address the substance of any of the requested preliminary rulings.

In Section II.D of its request, Korea submits that the Annex V procedure should be immediately suspended until such time as the Panel has issued preliminary rulings on the issues raised by Korea in its request. In the alternative, Korea asks (Section I, para. 6) that replies to the Facilitator's questionnaires should only be submitted to the Panel (through the Facilitator), but not to the parties, until such time as the Panel has had the opportunity to make the preliminary rulings requested by Korea. In its comments dated 2 September 2003, the European Communities submits that there is no basis to suspend or otherwise interrupt the Annex V procedure.

First, the Panel notes that, in accordance with paragraph 2 of Annex V, the Annex V procedure in the present case was initiated by the Dispute Settlement Body ("DSB"). There is no provision in Annex V which envisages the suspension of that procedure, either by the DSB itself or by any other body. In the absence of any provision explicitly authorising the Panel to suspend a procedure initiated by the DSB, the Panel has no authority to grant Korea's request for suspension of the Annex V procedure.

Second, we understand paragraph 4 of Annex V to mean that the Annex V procedure is under the control of the Facilitator. In light of paragraph 4, we do not see any scope for intervention by the Panel in procedural issues relating to Annex V. We further note that Korea already asked the Facilitator to suspend or adapt the Annex V procedure in the third paragraph of a letter dated 8 August 2003, and that Korea's request was rejected by the Facilitator in a letter to the parties dated 11 August 2003.

For the above reasons, we reject Korea's request that the Panel should suspend or adapt the Annex V procedure.

2. Other preliminary rulings requested by Korea on 29 August 2003

3. Exclusion of certain Annex V information

7.3.
At para. 68 of its first written submission, Korea asserts that the EC abused the Annex V procedure, by using that procedure to obtain information for (Part II) claims in respect of which the Annex V procedure does not apply. In particular, Korea asserts that the EC used the Annex V procedure (reserved for Part III claims) to obtain information regarding non-shipbuilding sectors, even though its Part III claims were limited to the shipbuilding sector. Korea asserts that the EC did so in order to obtain information to support its Part II claims, which do extend beyond the shipbuilding sector, but which fall outside the scope of the Annex V process.
7.4.
Korea submits that the Panel should exclude from its consideration under Part II of the SCM Agreement any evidence or information obtained under the Annex V procedure. Korea also submits that the Panel could decide to exclude any and all evidence obtained in the context of the Annex V process from the evidence considered by the Panel in reaching its decision as to the EC’s claims under both Part II and Part III of the SCM Agreement.19
7.5.
On 12 March 2004, the Panel issued the following ruling regarding this matter:

1. Korea asserts that the EC abused the Annex V procedure by using it to obtain information for (Part II) claims in respect of which the Annex V procedure does not apply. In particular, Korea asserts that the EC used the Annex V procedure (which it claims is reserved for Part III claims) to obtain information regarding non-shipbuilding sectors, even though its Part III claims were limited to the shipbuilding sector. Korea asserts that the EC did so in order to obtain information to support its Part II claims, which do extend beyond the shipbuilding sector, but which in Korea's opinion fall outside the scope of the Annex V process.

2. Korea submits that the Panel should exclude from its consideration under Part II of the SCM Agreement any evidence or information obtained under the Annex V procedure. Korea also submits that the Panel could decide to exclude any and all evidence obtained in the context of the Annex V process from the evidence considered by the Panel in reaching its decision as to the EC’s claims under both Part II and Part III of the SCM Agreement.

3. We note that the information at the heart of Korea's preliminary objection relates to the individual APRG and PSL transactions identified in paragraphs 170 and 172 of the EC's first written submission. In requesting transaction-specific APRG and PSL information from Korea, the Facilitator carefully limited the scope of the request to APRG and PSL transactions relating to "companies (involved (directly or indirectly) in trade in commercial vessels) ". In conformity with that request, the transaction-specific information provided by Korea did not extend beyond the commercial vessels sector. As a result, the transaction-specific information at issue does not extend beyond the commercial vessel sector. Furthermore, the relevant information concerns the existence of subsidization, and was relied on by the EC in respect of its Part III claims. We note that paragraph 2 of Annex V envisages the gathering of such information as necessary "to establish the existence and amount of subsidization". In addition, paragraph 5 of Annex V states that the designated representative's report to the Panel should include "data concerning the amount of the subsidy in question". In our view, therefore, the information relied on by the EC in support of its Part III claims regarding the existence of subsidization was properly gathered under the Annex V procedure. The EC therefore did not abuse the Annex V procedure in seeking that information.

4. We must now consider whether or not the EC was entitled to use that information for the additional purpose of supporting its Part II claims. In particular, the question is whether information properly gathered under the Annex V mechanism regarding the existence of alleged subsidization, which was properly relied on by the EC in support of its Part III serious prejudice claims against certain alleged subsidies, could also be used in the context of Part II claims concerning the same alleged subsidies.

5. In the context of the EC's Part III claims, we must determine whether or not the relevant APRG and PSL transactions constitute subsidies. In doing so, we are bound by the provisions of Article 1 of the SCM Agreement. At paragraphs 170 and 172 of the EC's first written submission, the EC is requesting us to perform the same analysis of subsidization20 in respect of the same measures in the context of its Part II claims. We see nothing in Annex V that would require us to ignore our Part III analysis of subsidization when reviewing the EC's Part II claims which concern allegations of the same subsidization in respect of the same measures. Nor indeed do we see any requirement in the SCM Agreement to perform this analysis more than once for any given measure alleged to be a subsidy.

6. In any event, even if we were precluded from relying on the relevant Annex V information when determining the existence of subsidization in the context of the EC's Part II claims, at the very least any finding of the existence of subsidization in respect of the EC's Part III claims would compel us to "seek" that very same information (i.e., regarding the same alleged subsidies) from Korea under Article 13.1 of the DSU. In other words, the very same information would in any event be brought before the Panel, but only much later in the proceedings, after the Panel had made findings regarding the existence of subsidization in respect of the EC's Part III claims. We recall that there is no textual basis for requiring us to rule that the relevant Annex V information is not directly admissible in respect of the EC's Part II claims. We therefore see no basis for ruling that the relevant information should only enter the record indirectly, and much later in the proceedings, via Article 13.1 of the DSU.

7. For the above reasons, we decline to rule that the EC was precluded from using information that was properly gathered under the Annex V mechanism regarding the existence of alleged subsidization, and properly relied on by the EC in respect of its Part III serious prejudice claims against certain alleged subsidies, in support of additional Part II claims concerning the same alleged subsidies.

4. Prejudice to Korea's earlier preliminary ruling request

7.6.
At para. 76 of its first written submission, Korea submits that "it would appear reasonable for the Panel to reconsider" its 19 September 2003 preliminary ruling concerning the alleged impermissible ambiguity of the EC's serious prejudice claims. Korea suggests that reconsideration of the ruling would be necessary subject to two conditions: (1) that the EC's statement (in its 5 September 2003 submission) that it had "never" intended its serious prejudice claims to cover a wider product scope than commercial vessels is inaccurate; and (2) that this may have had a bearing on the Panel's decision to decline Korea's preliminary ruling request on this point.
7.7.
Although Korea does not explicitly say so, we understand that Korea is referring to the Panel's preliminary ruling regarding the conformity of the EC's request for establishment with Article 6.2 of the DSU. The relevant ruling is found at para. 31 of the abovementioned communication dated 19 September 2003. It is apparent from the text of that communication that the Panel's ruling was not based on the EC's statement that it had "never" intended its serious prejudice claims to cover a wider product scope than commercial vessels. The Panel only took into account the EC's assertion that "its actionable subsidy claim relates to commercial vessels". In other words, the Panel focused on what was included in the EC's serious prejudice claim, and not whether or not the EC had ever demonstrated an interest in developing a serious prejudice claim outside of the shipbuilding industry. Since the second condition identified by Korea is not fulfilled, there is no need for the Panel to reconsider its earlier ruling.

5. Withdrawal of Article 5(a) claim

7.8.
Korea asserts that the EC has failed to pursue its claim under Article 5(a) of the SCM Agreement, and that it has not presented a prima facie case of any alleged injury in its first submission. Korea therefore requests that the Panel formally find that this claim has been effectively withdrawn.
7.9.
This issue is addressed by the Panel at para. 7,521 infra.

6. Price suppression and price depression / Article 6.3(c)

(a) Exclusion of price undercutting claim

7.10.
Korea notes that the EC's Article 6.3(c) claim rests solely on price suppression and price depression, and not price undercutting. Korea submits that the EC has failed to pursue its claim on price undercutting, and that it has not presented a prima facie case of any such price undercutting in its first submission. Korea therefore requests that the Panel formally find that this claim has been effectively withdrawn.
7.11.
This issue is addressed by the Panel at para. 7,521 infra.

(b) Scope of serious prejudice claim

7.12.
Korea asserts that the product scope of the EC’s serious prejudice claims must be restricted to commercial vessels pursuant to the EC’s latest assertion of 5 September.
7.13.
We understand that Korea's assertion regarding the scope of the EC's serious prejudice claim relates to Korea's original request for a preliminary ruling in respect of Article 6.2 of the DSU. Korea appears to be asking the Panel (although there is no explicit request) to rule that the scope of the EC's serious prejudice claims be restricted to commercial vessels.
7.14.
At para. 31 of the abovementioned 19 September 2003 communication, the Panel stated inter alia:

Since the European Communities has indicated that there will be no actionable subsidy claims in respect of products other than commercial vessels, there is simply no need for us to consider whether or not the request for establishment was sufficiently specific in respect of other products.

7.15.
It is clear, therefore, that the Panel already indicated its understanding that the scope of the EC's serious prejudice claim is limited to commercial vessels.

7. The extent of the record

7.16.
At para. 82 of its first written submission, Korea disputes the EC's argument that the responses provided in the context of the Annex V process should be considered the "record" on the basis of which the panel should reach its conclusions in this case.
7.17.
Since Korea has not requested a preliminary ruling on this issue, there is no need for the Panel to consider this matter further.

8. Continued relevance of Korea's 29 August 2003 request for preliminary rulings

7.18.
At para. 84 of its first written submission, Korea submits that a number of issues raised in its 29 August 2003 request for preliminary rulings remain applicable. In our view, these issues have already been addressed in our communications dated 19 September 2003 and 12 March 2004. We therefore do not consider it necessary to revisit the issues raised in Korea's submission of 29 August 2003.

B. alleged prohibited export subsidies

7.19.
The EC claims that Korea has provided and continues to provide its shipbuilding industry prohibited export subsidies, contrary to Articles 3.1(a) and 3.2 of the SCM Agreement.
7.20.
First, the EC claims that such subsidies were and are provided pursuant to the Export-Import Bank of Korea ("KEXIM") Act, the KEXIM Decree, and the KEXIM Interest Rate Guidelines (the "KEXIM legal regime", or "KLR"). The EC asserts that these measures "as such" (i.e., by their very existence, irrespective of their application in a given case) violate Articles 3.1(a) and 3.2 of the SCM Agreement.
7.21.
Second, the EC claims that Korea provides and has provided prohibited export subsidies pursuant to the KEXIM Advance Payment Refund Guarantee ("APRG") and Pre-shipment Loan ("PSL") programmes. Again, the EC claims that these measures "as such" violate Articles 3.1(a) and 3.2 of the SCM Agreement.
7.22.
Third, the EC claims that individual APRGs and PSLs (i.e., the application of the APRG and PSL programmes in individual cases) violate Articles 3.1(a) and 3.2 of the SCM Agreement.

1. Kexim Legal Regime

7.23.
Before turning to the substance of the EC's claims, we observe that a measure is generally to be treated as a prohibited export subsidy if it is a subsidy (as defined by Article 1 of the SCM Agreement) that is "contingent … upon export performance" (in the meaning of Article 3.1(a) of the SCM Agreement). While the EC's claims do not raise many issues regarding the notion of export contingency, the parties have made extensive arguments on the question of whether or not the measures at issue constitute subsidies. A subsidy exists if there is a "financial contribution" by a government or public body (or a private body entrusted or directed by the government) that confers a "benefit".
7.24.
We shall begin our analysis of the EC's claim against the KLR by determining whether measures taken under the KLR constitute financial contributions, and whether KEXIM is a public body. If so, we shall then determine whether or not the KLR confers a benefit. We shall then consider whether or not measures taken under the KLR are contingent on export performance.

(a) Does the KLR Provide for Financial Contributions?

(i) Arguments of the parties

7.25.
The EC submits that the KEXIM Act provides for direct transfers of funds within the meaning of Article 1.1(a) (1) (i) of the SCM Agreement. The EC asserts that Article 18 of the KEXIM Act provides for loans and loan guarantees, among other types of financing. The EC states that loans and loan guarantees are identified as types of direct transfer of funds in Article 1.1(a) (1) (i) of the SCM Agreement.
7.26.
Korea submits that the KLR does not provide for "financial contributions" in the meaning of Article 1.1(a) (1) (i) of the SCM Agreement because KEXIM does not engage in "government practice". Korea asserts that even if a body is a public body, it does not make a financial contribution if it is not involved in a "government practice". Korea states that the term "government practice" means the exercise of government authority, e.g., regulatory powers and taxation authority. According to Korea, this is confirmed by the context of Article 1.1(a) (1) (iv) of the SCM Agreement,which provides that there is a "government practice" only when a body carries out a function "which would normally be vested in the government" and "the practice, in no real sense, differs from practices normally followed by governments". Korea argues that the panel in US – Export Restraints looked at the negotiating history of the term "normally vested in the government" in Article 1.1(a) (1) (iv) of the SCM Agreement and concluded that the term "was a general reference to the delegation to private parties of the particular government functions of taxation and expenditure of revenue and not a reference to government market intervention in the general sense, or the effects thereof."21
7.27.
Korea asserts that KEXIM is set up for the specific purpose of meeting needs of an industrial or commercial nature, i.e., activities involving the extension of financing facilities on markets where it competes with other public or private operators based on market-oriented principles. Korea argues that in extending financing facilities such as APRGs or PSLs, KEXIM operates in a traditional banking capacity, performing functions normally performed by banks – not by governments.

(ii) Evaluation by the Panel

7.29.
In our view, the phrase "government practice" in Article 1.1(a) (1) (i) is simply a grammatical construction, or series of words, chosen because sub-paragraph (i) of Article 1.1(a) (1) could not have been drafted in the direct form.22 As such, it refers to cases ("practice") where governments or public bodies provide direct or potential direct transfers of funds. The phrase "government practice" is therefore used to denote the author of the action, rather than the nature of the action. "Government practice" therefore covers all acts of governments or public bodies, irrespective of whether or not they involve the exercise of regulatory powers or taxation authority. If the phrase "government practice" fulfils the filtering role advocated by Korea, this phrase would presumably also have been included in sub-paragraphs (ii) and (iii) of Article 1.1(a) (1). In particular, we would have expected it to be included in sub-paragraph (iii), such that only the provision of goods and services pursuant to the exercise of regulatory powers or taxation authority would be covered by that provision.23 However, sub-paragraph (iii) is not drafted in this way.
7.30.
We note Korea's argument that Article 1.1(a) (1) (iv) of the SCM Agreement refers to functions "normally (...) vested in the government" and "practice [that] in no real sense, differs from practices normally followed by governments". We note that this language was addressed by the panel in US – Export Restraints. That panel referred to the report of the Group of Experts on the Calculation of the Amount of a Subsidy, which in turn referred to a 1960 panel report.24 Like that panel, we too "find very significant the Group of Experts' interpretation that the 1960 Panel's reference to 'practice... in no real sense different from those normally followed by governments' was a general reference to the delegation to private parties of the particular government functions of taxation and expenditure of revenue,...".25 The Panel notes that the reference to functions "normally vested in the government" textually mirrors the reference to "practices normally followed by governments". Accordingly, the Panel considers that the reference to functions "normally vested in the government" should also be understood to mean functions of taxation and revenue expenditure. Thus, a function may be said to be "normally vested in the government" if that function involves the levy of taxation or the expenditure of revenue. Accordingly, since loans and loan guarantees involve revenue expenditure, they may be treated as functions "normally vested in the government", whether or not they are provided pursuant to the exercise of regulatory powers or taxation authority.26
7.31.
In light of the above, and since Korea does not dispute the EC's assertion that loans and loan guarantees fall within the scope of Article 1.1(a) (1) (i), we find that the KLR provides for "financial contributions" in the meaning of Article 1.1(a) (1) of the SCM Agreement.

(b) Is KEXIM a Public Body?

(i) Arguments of the parties

7.32.
The EC submits that KEXIM is a public body because (i) it is created and operates on the basis of a public statute giving the Government of Korea ("GOK") control over its decision-making, (ii) it pursues a public policy objective, and (iii) it benefits from access to state resources.
7.33.
Concerning governmental control over decision-making, the EC submits that as of December 2002 KEXIM was owned 51.6 per cent by the GOK, 42.8 per cent by the Bank of Korea, and 5.6 per cent by the Korea Development Bank ("KDB"), both of which latter are wholly subscribed by the Government of Korea. The EC further asserts that KEXIM’s key management is appointed and dismissed by the GOK, and its operations and budget are subject to the approval and control of the GOK. The EC also submits that KEXIM's annual Operation Plans are formulated under the control of the GOK.
7.34.
Concerning public policy objective, the EC asserts that pursuant to Article 1 of the KEXIM Act, KEXIM was created to fulfil the publicpurpose of promoting "the sound development of the national economy and economic cooperation with foreign countries."27 According to the EC, KEXIM itself acknowledges that it is a public body that acts in the interests of the country by serving as "an official export credit agency providing comprehensive export credit and project finance to support Korean exporters and investors" and facilitating "the development of the national economy and enhanc[ing] economic cooperation with foreign companies as a financial catalyst."28 The EC further notes that KEXIM’s annual reports specifically refer to KEXIM as "a special governmentalfinancial institution",29 and as "an agent of the Government".30 Similarly, the EC asserts that KEXIM’s website describes KEXIM as a "special government financial institution under the guardian authority of MOFE [i.e., the Ministry of Finance and Economy]"31, and as "a government institution [that] supports the Government’s policies on international trade and overseas investment."32 According to the EC, other KEXIM materials confirm that KEXIM "is a special government financial institution whose purpose is to promote the development of the Korean economy and economic cooperation with foreign countries... [and] expand appropriate financing activities to conform with government policies."33
7.35.
Concerning access to state resources, the EC submits that the GOK is required to guarantee any net loss incurred by KEXIM. In this regard, the EC asserts that GOK, and government-owned banks, injected over 1.6 trillion Korean Won ("KRW") between 1998 and 1999 into KEXIM, and at least an additional KRW 270 billion since January 2000. The EC asserts that an unlimited guarantee for losses and massive capital injection provide evidence of government influence and control over KEXIM.
7.36.
In the alternative, the EC submits that KEXIM is a private body "entrusted" or "directed" by the Korean Government within the meaning of Article 1.1(a) (1) (iv) of the SCM Agreement.
7.37.
Korea asserts that KEXIM is not a public body, as it is a body that carries on a business equivalent to that of a private operator. Korea submits that an organization is a public body only when it acts in an official capacity, or is engaged in governmental functions. Korea submits that the term "public" in Article 1.1(a) (1) of the SCM Agreement should be defined as "[a]cting in an official capacity on behalf of the people as a whole; as a public prosecutor".34
7.38.
According to Korea, the pursuit of a public policy objective does not confer on a body the status of a public body when such body is set up for the specific purpose of meeting needs of an industrial or commercial nature through the supply of goods or services on markets which are open to other public or private operators under fully competitive conditions. Korea submits that a general public policy purpose reflected in sectoral focuses is characteristic of many privately owned companies, particularly in the financial sector. By way of example, Korea asserts that investment trusts are limited in their activities in many countries, home mortgage lending is often a separate specialty, and often it is required that merchant banking be legally separate from retail banking. Korea asserts that it is a matter of focusing expertise, protecting consumers (corporate as well as natural), and protecting the integrity of the overall financial markets from errors caused by financial institutions venturing into substantive areas where they have insufficient expertise.
7.39.
Korea refers to the International Law Commission's Articles on State Responsibility in support of its position. According to Korea, Article 5 of the Articles on State Responsibility35 provides for a two-step analysis that helps clarify whether an entity is a public body. First, Korea submits that, pursuant to Article 5, the entity will be a public body if it "is empowered by the law of the State to exercise elements of the governmental authority." Korea asserts that this is a simple and logical test, based on the substance of what an entity is required to do rather than on questions of form such as whether a statute is a "public statute" or not. Regarding the second step, Korea submits that the acts in question will be considered acts of State only if such entities are acting pursuant to such authority in the particular instance. Thus, Korea asserts that it is not the case that an entity is a public body for all purposes simply because it might have been given authority to act for the State in some matters. Korea asserts that one must still determine that the acts in question were undertaken pursuant to the specific grant of governmental authority. According to Korea, if financing is offered as part of a commercial program by a para-statal entity, it is presumptively non-governmental and therefore should not be considered a financial contribution.
7.40.
Korea submits that in extending financing facilities, KEXIM is not acting in an official or governmental capacity. Korea asserts that KEXIM has no authority to regulate, and manufacturers and borrowers are free to seek financing from other financial institutions. According to Korea, the role played by KEXIM in its financing activities is the same as that of other financial institutions offering private financing. Korea argues that, even though KEXIM may have been established in the general interest of the public for the promotion of the growth of the national economy,36 it supplies financial services in markets that are open to other public or private operators under full competitive conditions in accordance with Article 18 of the KEXIM Act.
7.41.
In addition, Korea asserts that Article 26 of the KEXIM Act provides that KEXIM must operate to cover its expenses and fees so as to include a profit element. As regards guarantees such as the APRGs, Korea argues that this means that the premium rates must cover the long-term operating costs and losses of the programmes.37 According to Korea, even if a body disposes of governmental resources, it does not necessarily mean that the receiving body is a public body. Instead, Korea asserts that it could well be that the provision of governmental resources simply constitutes a subsidy to that body which still is a private body. Korea asserts that GOK injected capital into KEXIM not to cover KEXIM's losses, but to avoid negative credit ratings and maintain a sound Bank of International Settlements ("BIS") adequacy ratio. Korea rejects the EC's argument that KEXIM is not required to repay capital contributions made by the GOK, even during years in which KEXIM achieves a profit, since Article 36(2) of the KEXIM Act provides for the payment of dividends to the KEXIM shareholders, including the GOK, even if part of the profits will first be paid out to (non-GOK) preferential shareholders. Korea asserts that this preferential treatment was intended to help persuade commercial financial institutions and other entities to participate in capital contributions into KEXIM.
7.42.
Regarding government control, Korea argues that the daily operations of KEXIM are under the ultimate responsibility of, and thus decided by, the Board of Directors, without any form of control by the Government. Korea asserts that although KEXIM is to submit for approval by the Ministry of Finance and Economy the annual Operation Plans, which include the schedules/plans in broad perspectives as to administering loan provisions as well funding requirements therefor, the annual Operation Plan does not pertain to any terms or conditions prescribed for APRGs, nor does the Government require such terms or conditions for APRGs to be included.
7.43.
Korea also refers to paragraph 5(c) (i) of the GATSAnnex on Financial Services, which provides that the term "public entity" does not include "an entity principally engaged in supplying financial services on commercial terms". Korea argues that this definition is context for the interpretation of "public body" under Article 1.1 of the SCM Agreement in the present case, since KEXIM is engaged in the provision of financial services. According to Korea, since KEXIM is demonstrated to be an entity principally engaged in supplying financial services on commercial terms, the GATS Annex on Financial Services indicates that KEXIM should not be treated as a "public body".

(ii) Evaluation by the Panel

7.46.
Korea denies that its argument fails to properly distinguish between the concepts of public body and benefit. Korea asserts that the issue of lending on a commercial basis is, at the outset, a general one. According to Korea, therefore, if there is a general practice of lending on a commercial basis, then the entity is not a public body. We understand Korea to mean that one would assess whether or not an entity were lending on a commercial basis generally, rather than conducting the type of transaction-specific analysis that might be required for a "benefit" determination. We are not convinced by this argument, since it immediately raises the issue of how one would determine whether or not an entity were engaging in a "general practice of lending on a commercial basis".40 Would this only be the case if 100 per cent of total lending were on a commercial basis, or would 80 per cent suffice? And how would one determine that the lending is on a "commercial basis" without looking at the sort of factors envisaged in a "benefit" analysis? In our view, it is precisely because of the uncertainty surrounding such issues that it is important to maintain a clear distinction between the concepts of benefit and financial contribution / public body.
7.47.
We have the same concerns regarding Korea's reliance on paragraph 5(c) (i) of the GATS Annex on Financial Services.41 That is to say, Korea again fails to distinguish between the concepts of "public body" and "benefit". By defining "public body" on the basis of whether or not an entity operates on commercial terms, Korea is introducing considerations of benefit into the analysis of the private / public status of an entity. Furthermore, we question the relevance of the GATS Annex on Financial Services to an interpretation of Article 1.1(a) (1) of the SCM Agreement.42
7.49.
The SCM Agreement envisages a more straightforward approach, based on a clear distinction between public and private bodies. On the basis of this clear distinction, one may establish with relative certainty whether or not an entity is a public body whose financial contributions fall within the scope of the SCM Agreement. Only then need one address the more complex issue of whether or not a benefit is conferred (on the basis of a market benchmark). Korea's approach would blur the clear distinction between public and private bodies, and introduce complex considerations of benefit into the initial filtering process.
7.51.
Although there is some flexibility for the KEXIM President to change the Operation Program pursuant to resolutions passed by the Board of Directors, the circumstances in which the KEXIM President may do so are exhaustively set forth in the Operation Program, and therefore subject to Government control. In addition, we recall that the KEXIM President is a Government appointee.
7.52.
The Operation Programs provide detailed direction on the allocation of KEXIM financing between different financing activities and, thereafter, between different sectors. Thus, the Operation Programs generally identify three types of "financing supply" to be provided by KEXIM, and state the proportion of total financing for which each type of "financing supply" should account. Furthermore, the Operation Programs go so far as to stipulate how much of one specific type of "financing supply" should be directed towards specifically identified activities and industrial sectors.
7.53.
Korea asserts that the Operation Program does not pertain to any terms or conditions prescribed for APRGs and PSLs. While this may be true, GOK nevertheless enjoys extensive control over the parameters within which KEXIM must operate. Thus, if the Operation Program were to require KEXIM to cease providing financing in the shipbuilding sector, it would appear that KEXIM would be required to do so, even if that sector were the most profitable one in which KEXIM operated. The ability of the Government to issue such instructions is established by virtue of certain "basic directions" set forth in the Operation Programs.45 For example, the "basic directions" of the 1999 Operation Program required KEXIM to "support the export of capital goods such as ships, industrial plant, machinery, etc., which creates high net export earnings and industrial backward-forward effect".46 Such "basic directions" necessarily have an impact on the day-to-day operations of KEXIM, since they stipulate the areas in which KEXIM should focus its day-to-day operations (irrespective of purely commercial considerations).
7.54.
We consider that the "public" nature of KEXIM is further confirmed by KEXIM's own perception of itself as a "special governmental financial institution". In addition, we note that Korea describes KEXIM as an "export credit agency" (see Korea's reply to Question 52 from the Panel). This phrase is generally reserved for official export credit agencies, and not for private providers of export financing or insurance. Since the term "agency" suggests a relationship of agent and principal, one could reasonably assume that the relevant principal on whose behalf KEXIM acts as agent is the Government of Korea.
7.56.
Since we find that KEXIM is a "public body", there is no need to consider the EC's alternative argument that KEXIM is a private body entrusted or directed by the government.

(c) Subsidization

7.57.
The EC's claim that the KLR confers a benefit is based on a number of provisions of the KLR. We shall examine the parties' arguments concerning these provisions below. Before doing so, however, we must first decide whether or not to apply what is known as the traditional mandatory / discretionary distinction in reviewing the KLR provisions at issue.

(i) Application of the Traditional Mandatory/Discretionary Distinction

Arguments of the parties

7.58.
According to the traditional mandatory / discretionary distinction, only measures mandating WTO-inconsistent conduct could be condemned as such (i.e., on their face, rather than as applied in particular cases). Under this traditional distinction, therefore, the KLR could only be challenged as such if it mandated inter alia subsidization. The EC considers that this traditional distinction is no longer applicable, and claims that its application was excluded by the Appellate Body in
US – Corrosion Resistant Steel Sunset Review. According to the EC, therefore, it is no longer necessary that legislation must mandate export subsidization in order for it to be condemned under Article 3.1(a) of the SCM Agreement. The EC asserts that it is enough that legislation specifically envisages export subsidization in order for it to be condemned. The EC also argues that Article 3.2 of the SCM Agreement confirms that Members may not maintain the discretionary power to provide export subsidies. The EC relies on the panel report in Brazil – Aircraft to support an argument that a legal framework that provides for the provision of future export subsidies may be subject to an "as such" attack.
7.59.
Korea argues that the traditional mandatory / discretionary distinction remains applicable, and has not been overruled by the Appellate Body.

Evaluation by the Panel

(ii) Benefit

7.68.
The relevant arguments of the parties in respect of whether the KEXIM legal regime mandates (export) subsidies concern the KEXIM Act non-competition clause, the alleged absence of any obligation on KEXIM to take market conditions into account, the availability of government funding, the Market Adjustment Rate, the KEXIM Act provision concerning international competitiveness, and KEXIM documentation.

Non-competition clause

7.69.
Article 24 of the KEXIM Act provides:

[KEXIM] shall not compete with other financial institutions in performing the operations provided for in Article 18.

7.70.
Article 25(2) of the KEXIM Act provides in relevant part:

[KEXIM] may lend funds, discount drafts or notes, or guarantee obligations under paragraph (1) of Article 18 only when the term of repayment, payment or discharge is six (6) months or more but twenty five (25) years or less.

- Arguments of the parties

7.71.
The EC asserts that the non-competition clause means that KEXIM is specifically directed to perform functions and provide financing in situations in which no commercial bank would act, and therefore to make available loans and guarantees in financial circumstances that the market would not support.
7.72.
Korea states that in fact KEXIM is permitted to compete with commercial financial institutions, as confirmed by an earlier amendment of Article 18 of the KEXIM Act, whereby an obligation on KEXIM not to engage in operations "normally conductible by other financial institutions" was removed. Korea asserts that KEXIM is currently "in competition with commercial banks in all areas of financial services, except the long-term export credits with deferred payment terms that are regulated by the OECD Arrangement".57Korea argues that this is consistent with Article 25(2) of the KEXIM Act, whereby the maturity of KEXIM loans should be between six months and 25 years, and therefore of a maturity offered by commercial institutions. Korea argues that Article 24 "must be read together"58 with Article 25(2) to appreciate the limited scope of this prohibition. Korea further states that Article 24 should have been repealed, and that in fact "KEXIM has been contemplating proposing the repeal or amendment of Article 24".59

- Evaluation of the Panel

7.73.
We are not persuaded by Korea's arguments regarding the alleged interaction between Articles 24 and 25(2) of the KEXIM Act. First, we note that there are no cross-references between these provisions. Second, the fact that the limits under Article 25(2) on the maturity of KEXIM financing coincide with the maturity of commercial financing does not necessarily mean that KEXIM competes with commercial institutions, since it is still possible for KEXIM to comply with both provisions and offer financing with a maturity of between six months and 25 years in respect of which there is no competition from commercial financial institutions.
7.74.
Furthermore, even if Korea may be correct in stating that KEXIM does compete with private financial institutions in practice, we note that the EC's argument regarding Article 24 is made in the context of an "as such" claim against the KEXIM Act. Thus, because Article 24 continues to impose a legal obligation on KEXIM not to compete with other financial institutions, the fact that Article 24 may not be respected in practice is not relevant. The EC's "as such" claim concerns the KEXIM legal regime on its face, and not as applied in practice.
7.75.
That being said, we are not persuaded that the language of Article 24 of the KEXIM Act is sufficiently clear to conclude that it necessarily requires KEXIM to confer a benefit, i.e., offer terms that are more favourable than those available to the recipient on the market. In particular, does the non-competition clause mean that KEXIM is only required to provide financing when market operators are unable to do so? Or does it mean that KEXIM is not permitted to take business away from market operators? We consider that the terms of Article 24 are far too imprecise to draw the specific conclusion that it requires KEXIM to act in a below-market manner.

No obligation to take market conditions into account

- Arguments of the parties

7.76.
The EC claims that the KEXIM legal regime confers a "benefit" as such because the KEXIM Act imposes no obligation on KEXIM to take market conditions into account when disbursing funds.
7.77.
Korea submits that KEXIM is required to operate on a market-oriented basis. In particular, Korea asserts that the KEXIM legal regime requires KEXIM to appropriately assess the credit risks of the borrower, to apply interest rates or guarantee premia commensurate to the credit rating of the borrower/applicant, to take into account market situations when setting up interest rates/premium, to properly manage risks associated with the KEXIM business, and to ensure soundness of management.60

- Evaluation by the Panel

7.78.
We do not consider that a legal instrument may be found to mandate subsidization simply because it neither prohibits subsidization nor requires market conditions to be taken into account. The fact that a legal instrument is silent on subsidization should not lead to a conclusion that the resultant discretion will of necessity be exercised in a manner that results in subsidization. As stated by the Appellate Body in US – Section 211 Appropriations Act, "where discretionary authority is vested in the executive branch of a WTO Member, it cannot be assumed that the WTO Member will fail to implement its obligations under the WTO Agreement in good faith".61

Government funding

- Arguments of the parties

7.79.
The EC asserts that KEXIM need not act on market terms or with proper regard to risk, as the Government of Korea provides virtually unlimited funds to KEXIM. In this regard, the EC notes that Article 19 of the KEXIM Act provides that KEXIM "may borrow funds from the Government [and] the Bank of Korea...". The EC also argues that Article 36(2) of the KEXIM Act indicates that KEXIM is not required to pay... capital contributions made by the Government of Korea, even during years in which KEXIM achieves a profit. In this regard, the EC notes that Article 36(2) provides that KEXIM shall distribute its profits "on a preferential basis" "to capital contributors other than the Government." According to the EC, this provides further support for the understanding that KEXIM need not act in the same manner as a commercial bank, as it has access to funds of an important shareholder that does not demand to be treated in the same manner as shareholders operating pursuant to market considerations. The EC submits that KEXIM therefore receives a subsidy that it can pass on to its customers.
7.80.
The EC further asserts that Article 37 of the KEXIM Act provides that any net loss incurred by KEXIM that cannot be covered by its reserves shall be covered by funds from the Government of Korea. The EC argues that KEXIM need not therefore act in the same manner as a commercial bank, as it has no risk of insolvency. According to the EC, combining (a) the guarantee by Article 37 that the Government "shall provide funds to cover such net loss" and (b) the specific exclusion of the Government in Article 36(2) from the capital contributors that shall benefit from KEXIM’s net profit, underscores the fact that KEXIM need not act on market terms. The EC acknowledges that KEXIM must first attempt to cover net losses with its reserves, but argues that the ultimate guarantee of losses by the Government reduces the incentive for KEXIM to maintain a sufficient reserve and allows KEXIM to act otherwise than would a body subject to market forces.
7.81.
Korea asserts that there is no logical inference from Articles 19, 36(2) and 37 of the KEXIM Act that KEXIM need not act on market terms or with proper regard to risk. Korea argues that although these provisions indicate that the Government may provide funds, this is not the same as stating that KEXIM’s financing facilities need not be market-oriented. According to Korea, KEXIM is explicitly required by law to operate on a market-oriented basis.
7.82.
Korea asserts that the Government's capital contributions into KEXIM were necessary to allow KEXIM to maintain a good credit rating as well as a sound Bank of International Settlements ("BIS") adequacy ratio. Korea also asserts that Article 36 of the KEXIM Act was intended to encourage other entities to participate in capital contributions into KEXIM. Korea asserts that it is not uncommon in private corporations that the major shareholders receive less dividends and take more risks than other minor shareholders.
7.83.
Korea submits that the source of KEXIM's funds is not legally relevant, as the legal standard for subsidization is "benefit" to the recipient, rather than cost to the government. Korea submits that whatever the source of KEXIM’s funds, there is no indication in particular that any of these funds are used in loans or guarantees which bestow a benefit onto their recipient or were envisaged as such.

- Evaluation by the Panel

7.84.
Given that it is now well established that the legal standard for "benefit" is benefit to the recipient, and not cost to the government, the source of KEXIM's funds is irrelevant to the issue of whether or not the KEXIM legal regime mandates (export) subsidization. The fact that KEXIM may receive subsidized government funding does not mean that it will inevitably provide subsidized financing to its customers. It is possible that KEXIM might charge market rates and increase its profit margin instead.
7.85.
Furthermore, Articles 19, 36(2) and 37 of the KEXIM Act are relied on by the EC in support of an argument that KEXIM "need not" act on market principles. However, the EC has neither argued nor demonstrated that these provisions prevent KEXIM from acting on market principles. In other words, the EC has not argued that these provisions mandate subsidization.

Market Adjustment Rate

- Arguments of the parties

7.86.
According to the EC, the KEXIM Interest Rate Guidelines clarify that market conditions are not taken into account even in situations where KEXIM has specifically determined that its rates would be better than those available in the market. According to the EC, Articles 17(2) and 25(6) of the KEXIM Interest Rate Guidelines explicitly prevent, under certain situations, full market-based adjustments of KEXIM’s interest rates even when it has been determined that the rates are below those available in the market. According to the EC, the KEXIM Guidelines establish an explicit cap on the "Market Adjustment Rate" that would otherwise be applied to bring KEXIM’s rates into accord with the market’s rates.
7.87.
The EC notes that a new translation of Articles 17(2) and 25(6) of the KEXIM Interest Rate Guidelines recently submitted by Korea reads [BCI: Omitted from public version].62 According to the EC, Korea’s modified translation indicates that the Market Adjustment Rate is only a downward adjustment. The EC argues that if the Market Adjustment Rate can only be used to adjust the cost-based rate downwards, this provision will often lead to interest rates that are below the market rate. The EC asserts that in any event, these provisions do not, by their terms, ensure that KEXIM provides loans at market rates, since the Market Adjustment Rate is an adjustment to the cost based rate that takes account not only of lower offers by other banks but also of the "business relationship with the borrower and the distinctive features of the transactions, etc."63
7.88.
Korea submits that the Market Adjustment Rate in Articles 17(2) and 25(6) of the Interest Rate Guidelines operates on a market-oriented basis to take into account the PSL interest rates and the APRG premia offered by other financial institutions, the track record and relationship of the applicant with KEXIM and other considerations including the particulars of the project concerned.
7.89.
Korea asserts that a "market rate" exists in the form of "range" or "band", not a single rate. Korea explains that the Market Adjustment Rate is one of the spreads (discounts or premia) that are to be applied upward or downward to the base rate in addition to other spreads such as "credit risk spread" and "target margin". According to Korea, this Market Adjustment Rate may be applied by the KEXIM loan managers when determining the rates for a specific applicant in a specific individual transaction. Korea asserts that the following factors are to be taken into account when determining whether and to what extent to apply this Market Adjustment Rate: the APRG premia offered by other financial institutions; the track record and relationship of the applicant with KEXIM; and other considerations including particulars of the project concerned.
7.90.
Korea also submits that while the Market Adjustment Rate allows the KEXIM loan managers to react to the market, the loan manager is nevertheless prohibited, when he applies the Market Adjustment Rate for ‘downward’ adjustment of an ARPG fee, from applying it beyond a certain limit [BCI: Omitted from public version]. In contrast, there is no limitation when the loan manager applies it for upward adjustment. Korea asserts that in this sense, the Market Adjustment Rate is a "floor", not a "cap". Korea submits that the Market Adjustment Rate does not cause the final fee rate to be set below the market rates.

- Evaluation by the Panel

7.91.
According to Korea, Article 17 of the KEXIM Guidelines for Interest Rates and Fees Amended provides:

[BCI: Omitted from public version.]

7.92.
According to Korea, Article 25 of the KEXIM Guidelines for Interest Rates and Fees Amended provides:

[BCI: Omitted from public version.]

[...]

[BCI: Omitted from public version.]

7.93.
The above translation of Articles 17 and 25 was attached to Korea's reply to Question 57 from the Panel, and amends the translation initially provided by Korea during these proceedings. The EC has not contested the accuracy of the amended translation.
7.94.
We do not consider that the EC's argument that Articles 17(2) (Market Adjustment Rate in respect of loan interest rates) and 25(6) (Market Adjustment Rate in respect of guarantee premia) only allow a downward adjustment is correct. In our view, the fact that these provisions do not explicitly refer to upward adjustments does not mean that they should be interpreted to mean that upward adjustments are precluded. Indeed, we note that Korea has provided evidence of situations in which upward adjustments have been made in practice (see Korea's reply to Question 58 from the Panel). Furthermore, we note that Articles 17(2) and 25(6) do not impose limits on the amount of upward adjustment. The possibility of upward adjustment, and the absence of any limit on the amount of such upward adjustment, means that KEXIM is not automatically locked into providing below-market financing.
7.95.
In addition, we note the EC's argument that the above Market Adjustment provisions do not "ensure" that KEXIM provides services at market rates. As noted above, we do not consider that the absence of an obligation on KEXIM to apply market rates permits a finding that the KEXIM legal regime mandates below-market rates, and therefore subsidization. We also recall that in Canada – Aircraft – Article 21.5 the Appellate Body expressed reservations regarding the application of an "ensure" standard, noting that such a standard could "be very difficult, if not impossible, to satisfy".64 We further recall that the panel only applied an "ensure" standard because it was proposed by the parties in that case.65
7.96.
For the above reasons, we reject the EC's arguments concerning Articles 17(2) and 25(6) of the KEXIM Interest Rate Guidelines.

KEXIM On-Line Road Show

- Arguments of the parties

7.97.
The EC submits that KEXIM has itself acknowledged its role as providing financial contributions to Korean exporters in cases and on terms that would not be provided by commercial banks. In this regard, the EC notes that KEXIM’s 2003 "On-Line Road Show" stated that one of the core missions of KEXIM’s business was to serve "a complementary but pioneering role and function for the national economy, which would be hard for commercial banks to shoulder."66 The EC asserts that KEXIM has therefore acknowledged its role as providing financial contributions to Korean exporters in cases and on terms that would not be provided by commercial banks.
7.98.
Korea asserts that the "On-Line Road Show" contains a description of the specialized role and function being performed by KEXIM as an export credit agency. Korea submits that export credit agencies, such as KEXIM, generally provide specialized trade-related financing involving longer-term project-related loans (e.g., mid- and long-term export loans), special payment terms (e.g., deferred or specially structured payments) or specialized collateralization methods. Korea submits that it is important to remember the context of the establishment of KEXIM, i.e., the fact that Korea was a developing country with inadequately formed capital markets, among other things. According to Korea, it is quite typical in such situations for specialist banks to be set up to provide such pioneering expertise. Korea submits that the "On-Line Road Show" is irrelevant to the question of below-market financing by KEXIM.

- Evaluation by the Panel

7.99.
We note that the EC refers to the On-Line Road Show as evidence that "KEXIM has acknowledged its role as providing financial contributions to Korean exporters in cases and on terms that would not be provided by commercial banks".67 The On-Line Road Show describes KEXIM's alleged practice, rather than its legal obligations under the KEXIM legal regime. Since the On-Line Road Show relates to practice, rather than the KEXIM legal regime per se, it has no bearing on our findings regarding the EC's claim against the KEXIM legal regime "as such".

Maintenance of international competitiveness

- Arguments of the parties

7.100.
The EC asserts that Article 26 of the KEXIM Act demonstrates that KEXIM values the "international competitiveness" of Korean export-oriented industries over its own financial condition, a situation that increases KEXIM’s ability to provide support on terms better than those available in the market.
7.101.
Article 26 provides:

Except where inevitable for maintaining the international competitiveness to facilitate the export, or for promoting the overseas investment or overseas exploitation of natural resources, the interest rates, discount rates and fee rates applicable to loans, discounts and guarantees... shall be so set as to cover the operating expenses, commissions for undertaking of delegated operations, interest on borrowed funds, and depreciation of assets which [KEXIM] incurs.

7.102.
The EC asserts that although Article 26 of the KEXIM Act states that interest rates and fees imposed with respect to loans and guarantees are to be set so as to cover the operating costs of KEXIM, there is no requirement that KEXIM establish rates that comport with market rates. The EC notes that Article 26 permits KEXIM to avoid the requirement to cover its operating costs where "inevitable for maintaining the international competitiveness to facilitate... export....".
7.103.
Korea submits that Article 26 has no purpose other than to provide that all fees and rates must cover "at least" the costs when KEXIM provides financing. Korea asserts that Article 26 does not prohibit KEXIM from earning profits and, instead, effectively requires it to carry on profitable operations. Korea argues that other relevant provisions of the KEXIM Decree, such as Articles 17-3 through 17-13 (providing parameters for sound and profitable management of KEXIM), also effectively require KEXIM to carry on its business for profit. Korea further asserts that the Interest Rate Guidelines of KEXIM provide that KEXIM interest rates and fees are always aligned with market rates.
7.104.
Korea submits that the phrase "inevitable for maintaining the international competitiveness to facilitate... export" was included in Article 26 of the KEXIM Act in order to allow KEXIM the option to provide financing at below-cost levels in exceptional situations when KEXIM faces severe ‘rates’ competition from foreign financial institutions, as in the context of "matching" under the OECD Arrangement. Korea submits that because "matching" would be exceptional, Article 26 uses the term "inevitable", which means that under normal or ordinary circumstances this exception must not be applied. Korea notes that this "exception" under Article 26 has never been applied in practice thus far. Further, Korea asserts that KEXIM has interpreted this matching mechanism in such a restrictive manner that it can be applied only for matching of "country risk premium", not the total interest rate applied by the competing export credit agencies. Korea also submits that even if KEXIM’s interest rates had in exceptional circumstances gone below its operating expenses (which Korea claims they have never done), this has nothing to do with the finding of a benefit or a subsidy. According to Korea, as long as Article 26 permits KEXIM to match the low interest rates applied by other competing export credit agencies, KEXIM will always end up applying the market benchmark (i.e., the prevailing conditions in the market), whether or not the KEXIM rate is below or above its operating expenses.
7.105.
The EC notes Korea's argument that this provision allows KEXIM to finance at below-cost when "matching" under the OECD Arrangement. However, the EC considers that this explanation does not justify the provision. First, the EC asserts that there is absolutely no mention of either the OECD Arrangement or "matching" in this provision. Although Article 43 of the KEXIM Interest Rate Guidelines does refer to matching, the EC submits that there is nothing in these Guidelines that indicates that this Article should be read together with Article 26 of the KEXIM Act. Second, the EC notes that the panel in Canada –Aircraft Credits and Guarantees concluded that "matching" under the OECD Arrangement does not provide a valid affirmative defence for measures that violate the terms of the SCM Agreement.68

- Evaluation by the Panel

7.106.
Leaving aside the question of whether or not "matching" under the OECD Arrangement is in conformity with the SCM Agreement, we would simply note that the EC again applies the incorrect legal standard in pursuing its arguments regarding Article 26 of the KEXIM Act. First, the EC states that there is no requirement that KEXIM rates comport with market rates. As noted above, however, the fact that there is no requirement to act on market terms does not mean ipso facto that KEXIM is required, or mandated, to provide below-market terms. Second, the EC asserts that Article 26 permits KEXIM to avoid the requirement to cover its operating costs in certain circumstances, and increases KEXIM’s ability to provide support on terms better than those available in the market. Again, however, the EC does not argue that Article 26 requires KEXIM to provide below-market financing. Providing a public body with the legal and financial ability to subsidizeis not the same as requiring it to do so.

Conclusion

7.107.
For the above reasons, the EC has failed to establish a prima facie case that the KEXIM legal regime mandates subsidization. Although certain provisions of the KLR might indicate that it was intended as a means of providing subsidies, a conclusion that the KLR could be applied in a manner that confers a benefit would not be a sufficient basis to conclude that the KLR as such is mandatory legislation susceptible of inconsistency with Article 3.1(a) of the SCM Agreement.69

(d) Export contingency

(i) Arguments of the parties

7.108.
The EC submits that, pursuant to Article 18 of the KEXIM Act, financial contributions by KEXIM are "[f]or the purpose of facilitating exports of products" and, therefore, contingent on export within the meaning of Article 3.1(a) of the SCM Agreement.
7.109.
Korea has not taken a position on whether or not the KLR is export contingent.

(ii) Evaluation by the Panel

7.110.
In light of our finding that the EC has failed to establish a prima facie case that the KLR mandates subsidization, we do not consider it necessary to determine whether or not then KLR is "contingent... upon export performance" within the meaning of Article 3.1(a) of the SCM Agreement.

(e) Conclusion

7.111.
Given our finding that the EC failed to establish a prima facie case that the KEXIM legal regime mandates subsidization, we reject the EC's claim that the KEXIM legal regime "as such" is inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement.

2. APRG programme

7.112.
An APRG provides foreign buyers with a guarantee that they will be refunded any advance payments made to an exporter, including any accrued interest on the advance payments, in case the Korean company defaults under the relevant export contract. In exchange, the Korean exporter pays a premium consisting of (1) a minimum base rate, and (2) additional spreads (e.g., credit and market risk spreads).

(a) Arguments of the Parties

7.113.
The EC asserts that the KEXIM APRG programme "as such" provides for the grant of subsidies that are contingent on export, contrary to Articles 3.1(a) and 3.2 of the SCM Agreement. The EC submits that the APRG programme provides for financial contributions because (i) KEXIM is a public body, and (ii) APRGs constitute a "potential direct transfer of funds" pursuant to Article 1.1(a) (1) (i) of the SCM Agreement. The EC submits that the APRG programme confers a benefit because, although KEXIM levies a premium when granting APRGs, the premium fails to reflect the degree of creditworthiness, or lack thereof, of the Korean exporters. According to the EC, KEXIM issues guarantees without proper consideration of the risk involved in the transaction – in many cases, granting guarantees to financially troubled companies that would not have been able to obtain a guarantee from a commercial bank. The EC submits that the APRG programme is de jure export contingent because KEXIM APRGs are provided for the specific purpose of guaranteeing the down payments for Korean goods intended for export.
7.114.
Korea asserts that the EC's description of the measure is incomplete and inadequate to support its claim regarding the KEXIM APRG programme. According to Korea, whilst a definition is given of what the guarantees denominated as APRGs entail for the manufacturers of the capital goods and their purchasers, the EC fails to identify what precisely constitutes the so-called "APRG programme". Korea submits that the measure at issue is therefore un-defined and cannot as such be the subject of a detailed factual or legal analysis.
7.115.
Korea also denies that the APRG programme constitutes a prohibited export subsidy. Korea asserts that the APRG programme does not constitute a financial contribution covered by the SCM Agreement because KEXIM is not a "public body". Korea also denies that the APRG programme confers a "benefit". Korea asserts that the EC must provide a benchmark to define whether the APRG program "as such" yields premium rates that confer a benefit. Korea considers that the EC has failed to meet this burden, since it only refers to certain individual APRGs extended to shipyards alone rather than to the basic conditions of the program as such irrespective of the sector of industry. In the alternative, Korea submits that the APRG programme benefits from a safe haven pursuant to item (j) of the Illustrative List of Export Subsidies.

(b) Evaluation by the Panel

7.116.
We shall first address Korea's arguments regarding the identification of the measure. We shall then turn to the substance of the EC's claim against the APRG programme.
7.117.
Regarding the identification of the measure at issue, the Panel sought clarification from the EC regarding the extent to which its claim against the APRG programme differed from its claim against the KEXIM legal regime. In response to Question 138 from the Panel, the EC stated that the APRG and PSL programmes are linked to the KEXIM legal regime, in the sense that they are a consequence of the KEXIM legal regime. According to the EC, the APRG and PSL programmes are also distinguishable from the KEXIM legal regime since, although the KEXIM legal regime envisages the provision by KEXIM of financial services, these financial services do not necessarily need to be the APRG and PSL programmes, as KEXIM could provide financial assistance to exporters in other forms.
7.118.
Furthermore, in its first written submission, the EC stated that the APRG programme was introduced immediately after the establishment of KEXIM, and has been administered since that time pursuant to Article 18 of the KEXIM Act.70 The EC also submits that KEXIM is authorised to issue "performance guarantees related to contracts for export" pursuant to Article 23(1) of the KEXIM Operating Manual.71
7.119.
In our view, the EC has done enough to identify an APRG programme "as such". We consider that the legal basis for that programme is set forth in Article 18 of the KEXIM Act and Article 23(1) of the KEXIM Operating Manual. We shall conduct our analysis of subsidization and export contingency on the basis of those provisions.
7.120.
We recall that a measure is only a subsidy covered by the SCM Agreement if it is a "financial contribution" by a government or public body that confers a "benefit". We have already found that KEXIM is a "public body" in the meaning of Article 1.1(a) (1) of the SCM Agreement.72 Regarding the "financial contribution" element, we accept the EC's argument that the APRG programme, pursuant to Article 23(1) of the KEXIM Operating Manual, provides for a "potential direct transfer of funds" within the meaning of Article 1.1(a) (1) (i) of the SCM Agreement. We therefore find that the APRG programme constitutes a "financial contribution" covered by the SCM Agreement.
7.121.
As to whether the APRG programme "as such" confers a "benefit", however, we recall that we are applying the traditional mandatory / discretionary approach. The issue before us, therefore, is whether or not the APRG programme mandates the conferral of a benefit by requiring the provision of APRGs on terms more favourable than Korean shipyards could obtain on the market. We do not consider that the EC has established a prima facie case to this effect.73 Neither Article 18 of the KEXIM Act nor Article 23(1) of the KEXIM Operating Manual even refer to the terms on which KEXIM shall offer APRGs, let alone require below-market guarantees. The EC has not identified any other provisions regulating the terms of APRGs.74 Accordingly, we reject the EC's claim that the APRG programme "as such" constitutes a subsidy. For this reason, there is no need for us to examine the EC's claim that the APRG programme is de jure export contingent, nor Korea's reliance on item (j) of the Illustrative List.

3. PSL programme

7.122.
PSLs are loans made to Korean companies in connection with export contracts for the purpose of assisting Korean exporters to finance production.

(a) Arguments of the Parties

7.123.
The EC asserts that the KEXIM PSL programme "as such" provides for the grant of subsidies that are contingent on export, contrary to Articles 3.1(a) and 3.2 of the SCM Agreement. The EC submits that the PSL programme provides for financial contributions covered by the SCM Agreement because (i) KEXIM is a public body, and (ii) PSLs constitute a "direct transfer of funds" pursuant to Article 1.1(a) (1) (i) of the SCM Agreement. The EC submits that the PSL programme confers a benefit because PSLs are provided at preferential interest rates that place the Korean exporters in a more advantageous position than if they were to obtain such financing on market terms. The EC submits that KEXIM's website describes the PSL programme as designed "to encourage the export of capital goods such as... ships... involving larger credits and longer repayment terms than what suppliers or commercial banks would provide."75 According to the EC, this shows that the very purpose of KEXIM’s pre-shipment loans is to provide financing to shipbuilders on better terms than they could receive in the market. The EC submits that the PSL programme is de jure export contingent because KEXIM PSLs are provided for the specific purpose of guaranteeing the down payments for Korean goods intended for export.
7.124.
Korea submits that the EC has failed to identify what precisely constitutes the so-called "KEXIM pre-shipment loan program". Korea therefore considers that the measure at issue is
un-defined and cannot "as such" be the subject of a detailed factual or legal analysis. Korea further asserts that the EC fails to give any support other than by way of general statements -- without evidentiary support or, in some cases, outright inaccurately -- as regards the benchmark on the basis of which it claims that the pre-shipment loan program confers a benefit. For Korea, the EC fails to make a prima facie case that the PSL program "as such" constitutes a subsidy. According to Korea, as the interest rates for the pre-shipment loans under the KEXIM program are determined taking into account base rates reflecting market rates, the credit rating of the manufacturer of the capital goods covered by the pre-shipment loan and the collateral provided by the pre-shipment loan beneficiary, the pre-shipment loan program "as such" does not confer a benefit. Regarding the abovementioned extract from the KEXIM website, Korea considers that providing a longer term than is generally available does not mean that the rates are below market, since it depends on how those rates are adjusted to reflect the different terms. According to Korea, the size of a credit may or may not require different rates; it depends on factors extraneous to size alone. In the alternative, Korea submits that the PSL programme benefits from a safe haven pursuant to the first paragraph of item (k) of the Illustrative List.

(b) Evaluation by the Panel

7.125.
We recall that the EC replied to Question 138 from the Panel regarding the identification of the PSL programme, as set forth at para. 7,117 above. We also note that the EC stated that the PSL programme was introduced immediately after the establishment of KEXIM, and has been administered since that time pursuant to Article 18 of the KEXIM Act.76 The EC further argued that KEXIM is authorised to provide "Pre-delivery Export Loans, which are extended until the delivery date of the export goods and/or services concerned" pursuant to Article 11(4) of the KEXIM Operating Manual.77 In light of these considerations, we find that the EC has done enough to identify a PSL programme "as such". We consider that the legal basis for that programme is set forth in Article 18 of the KEXIM Act and Article 11(4) of the KEXIM Operating Manual. We shall conduct our analysis on the basis of those provisions.
7.126.
Regarding the existence of a "financial contribution" covered by the SCM Agreement, we have already found that KEXIM is a "public body" in the meaning of Article 1.1(a) (1) of the SCM Agreement. We further accept the EC's argument that the PSL programme, pursuant to Article 11 of the KEXIM Operating Manual, provides for a "direct transfer of funds" within the meaning of Article 1.1(a) (1) (i) of the SCM Agreement. We therefore find that the PSL programme constitutes a "financial contribution" covered by SCM Article 1.1(a) (1).
7.127.
As to whether the PSL programme confers a "benefit" and therefore constitutes a subsidy, the issue before us is whether or not the PSL programme mandates the conferral of a benefit by requiring the provision of PSLs on terms more favourable than Korean shipyards could obtain on the market. We do not consider that the EC has established a prima facie case to this effect.78 Neither Article 18 of the KEXIM Act nor Article 11(4) of the KEXIM Operating Manual even refer to the terms on which KEXIM shall offer PSLs, let alone require below-market loans. The EC has not identified any other provision regulating the terms of PSLs.79
7.128.
Regarding the KEXIM website material, we note that it was submitted by the EC in support of an argument regarding the "purpose"80 of the PSL programme. We recall, however, that the question we must answer is whether or not the PSL programme requires KEXIM to provide prohibited export subsidies. The intent behind the PSL programme is not relevant to this issue. In this respect, we agree with the following statement by the panel in Brazil – Aircraft (Article 21.5 – Canada II) :

In our view, a conclusion that PROEX III could be applied in a manner which confers a benefit, or even that it was intended to be and most likely would be applied in such a manner, would not be a sufficient basis to conclude that PROEX III as such is mandatory legislation susceptible of inconsistency with Article 3.1(a) of the SCM Agreement.81

7.129.
In light of the above, we reject the EC's claim that the PSL programme constitutes a subsidy. For this reason, there is no need for us to examine the EC's claim that the PSL programme "as such" is de jure export contingent, nor Korea's reliance on the first paragraph of item (k) of the Illustrative List.

4. Individual APRG transactions

7.130.
The EC has identified a number of KEXIM APRGs which it claims are prohibited export subsidies. The EC argues that these APRGs were provided on terms more favourable than the recipients could have obtained on the market. This argument is based on a comparison of the terms of the KEXIM APRGs with those of APRGs provided by certain other domestic banks and foreign banks.
7.131.
Korea denies that the APRGs identified by the EC constitute prohibited export subsidies. In the alternative, Korea asserts that KEXIM APRGs benefit from a safe haven provided for through an a contrario reading of item (j) of the Illustrative List. The EC submits that an item (j) defence is not available to Korea.
7.132.
The arguments of the parties raise a number of horizontal issues, mostly concerned with the market benchmarks proposed by the EC. We shall consider these horizontal issues before turning to the parties' transaction-specific arguments. In the event that we find any of the relevant APRG transactions inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement, we shall consider the availability of a defence under item (j) of the Illustrative List.

(a) Horizontal issues

7.133.
The parties have made arguments regarding the following horizontal issues: designation by the ship purchaser of a specific foreign APRG provider, use of foreign benchmarks, country risk spreads, use of domestic benchmarks, credit risk spread, past subsidies, and adverse inferences.

(i) Buyer's designation of foreign APRG-provider

7.134.
Korea asserts that certain shipyards are not always able to choose which financial institution would provide the APRG for its customers, since certain buyers require Korean shipyards to procure APRGs from designated foreign banks. Korea submitted evidence of a number of such designations. The EC has not responded to Korea's argument.
7.135.
In our view, the fact that a buyer designates the source from which a shipyard is to procure an APRG establishes a prima facie case that such APRG should not be treated as a market benchmark. In such cases, the designation of the APRG-provider by the buyer means that there is a risk that the APRG is not negotiated at arm’s length, since the shipyard is a captive buyer. The rate paid by the shipyard might therefore be higher than it would if the shipyard were able to shop around and compare offers from alternative suppliers.
7.136.
Korea has submitted evidence pertaining to one transaction (Exhibit Korea-59) demonstrating that the buyer, [BCI: Omitted from public version] designated [BCI: Omitted from public version] as the provider of two APRGs proposed by the EC as market benchmarks against which to assess the terms of KEXIM APRGs provided in respect of Daedong. For the reasons set forth in the preceding paragraph, and given the absence of any rebuttal by the EC, we reject the use of these two APRGs, dated 7 July 1999, as market benchmarks. That being said, we note that this nevertheless leaves one [BCI: Omitted from public version] APRG (provided on the same terms, and about which Korea has not adduced any evidence of buyer designation) for use as a market benchmark.
7.137.
Korea also submitted further evidence in Exhibits KOREA-58 and KOREA-83, concerning foreign APRG providers. This evidence is less probative, however. Exhibit KOREA-58 relates to an APRG transaction that is not proposed as a market benchmark by the EC. The evidence contained in Exhibit Korea-83 relates to APRGs provided to shipyards that fall outside the scope of these prohibited export subsidy claims. The evidence in these Exhibits is therefore of no direct relevance to our findings.

(ii) Foreign market benchmark

7.138.
Early in the proceedings, the Panel was under the impression that Korea was arguing that, as a matter of law, APRGs provided by foreign banks could not form part of the "market" against which to compare APRGs provided by KEXIM. During the second substantive meeting, however, Korea stated that:

"the Panel may wish to do as the EC requests and make a ruling that foreign lenders can be part of the market, but such a ruling would be completely beside the issue of choosing an appropriate benchmark. The EC cites the Appellate Body report in
US -- Lumber CVD Final as support for its position, but that only serves as an illustration of the "straw man" argument the EC is using."82

7.139.
In light of this statement, we do not consider that Korea disputes that foreign market benchmarks could be used as a matter of law.83 Rather, we understand Korea to argue that, as a matter of fact, the foreign market benchmarks relied on by the EC are not representative of APRG activities in the Korean market since the foreign APRG providers only participated in the APRG business on an exceptional basis, and were less familiar with that business.84
7.140.
In response, the EC submits that there is no reason to believe that foreign institutions are less capable of evaluating the creditworthiness of a shipyard or its technical capability to carry out the construction project until delivery. In response to Question 10 from the EC, Korea stated that its argument "is based on discussions with the shipyards and KEXIM. Korea has asked the shipyards and KEXIM for any further documentation and it will be submitted when provided to the Government of Korea".
7.141.
We do not consider that a vague reference to shipyard and KEXIM perceptions is sufficient to reject the foreign market benchmarks proposed by the EC. We also note that, despite Korea's response to Question 10 from the EC, no documentation supporting the alleged perception of the shipyards and KEXIM was submitted in the subsequent stages of the Panel proceedings. In the absence of more substantial arguments by Korea regarding alleged shortcomings in foreign banks' evaluation of the creditworthiness of shipyards, we are not persuaded by Korea's argument that foreign benchmarks are not appropriate as a result of their lack of familiarity with the APRG business. In our view, provided the terms of an APRG are negotiated at arm's length for fair market value, the fact that the provider engages in only a limited number of transactions should not be conclusive, and should not preclude the use of such transactions as market benchmarks.

(iii) Country risk spreads

7.142.
Korea has submitted evidence to the effect that at least one foreign bank proposed by the EC as a market benchmark included a 0.6 per cent country risk spread in its APRG rates for Korean shipyards. Korea submits that all other foreign banks would have done likewise. Korea submits that this means that foreign market APRG rates cannot be compared with KEXIM APRG rates without adjusting for country risk.
7.143.
The EC does not contest that a country risk spread would have been included by foreign APRG providers. However, the EC submits that the same spread should also have been included by Korean banks, since the APRGs were provided in a foreign currency (i.e., US dollars). The EC argues that the risk of providing an APRG to a Korean company in a foreign currency, as it is the case for most of the APRGs, is the same regardless of where the bank is based. The EC notes in this regard that the Comptroller's Handbook on Country Risk Management prepared by the US Comptroller of the Currency provides that "[c]ountry risk is not necessarily limited to a bank's exposures to foreign-domiciled counterparties".85
7.144.
The EC asserts that a Korean country premium has to be taken into account whenever a currency exposure is generated, regardless of whether a domestic or foreign bank issues the APRGs. The EC asserts that KEXIM quoted below-market rates because it failed to take such currency exposure into account. According to the EC, the country risk of Korea needs to be taken into account in the price as an add-on to cover the transfer risk resulting from the company needing to find foreign currency in the case where a government wants to keep the "strong" foreign currencies, and as an add-on resulting from the bank’s needs to obtain refinancing in the foreign currency (in the case of default by the shipyard).
7.145.
Korea submits that Korean domestic banks, by definition, cannot face the risk of their own country. Korea asserts that, for Korean banks, risks from events in their own country are no longer "country risk." Korea asserts that the Comptroller's Handbook does not imply that country risk can be applied to "every" domestic counterparty that is involved in "an export transaction". Rather, the handbook itself establishes that the country risk may be exceptionally applied to transactions with domestic counter parties under very limited circumstances, e.g., where the business of a domestic borrower is heavily relying on the businesses associated with that particular foreign country with respect to which the country risk is assessed. Korea asserts that this explanation is understandable because, in the situations where the business of a borrower (or guarantor) is heavily relying on transactions associated with a specific foreign country, events in such a foreign country will directly and significantly affect the general credit risks of the borrower (or guarantor) which in turn will significantly and directly affect the creditworthiness of the borrower (or guarantor). According to Korea, only in such specific circumstances would it make sense to take into account the country risk of such specific foreign country when assessing the creditworthiness of such borrower (or guarantor).
7.146.
Korea submits that no Korean shipyards deal exclusively with a specific foreign country such that the events in that foreign country would significantly and directly affect the creditworthiness of the Korean shipyards. Korea also asserts that among the buyers of Korean ships, the absolute majority of buyers come from high income OECD countries, such as the EC, Norway, USA and Japan. Korea submits that no foreign financial institution would apply country risk with respect to counterparties from such countries as they do not bear any country risks.
7.147.
Korea submits that, moreover, even if a Korean shipyard were exposed to the country risk of a particular foreign country by retaining significant "export receivables" from the buyers in that foreign country, the country risk it bears is the country risk of that particular "foreign" country, and not the country risk of "Korea". In other words, in such case, any Korean banks issuing APRGs to such Korean shipyard would apply the country risk of the said "foreign" country, not the "Korean" country risk.
7.148.
Korea also asserts that Korean country risk relates to Korea’s ability to honor its "external" financial commitments. Korea therefore asserts that Korean domestic banks cannot face similar "Korea risks" to those faced by foreign banks. Korea also submits that, in light of the definition of country risk, country risk factors such as the risk of expropriation of assets and the risk of currency manipulations must be understood to mean those that are of such nature that rather directly affect external financial obligations of Korea. Korea asserts that, by nature, these risks could not be the same as those risks faced by Korean domestic banks.
7.149.
Korea acknowledges that, during the period of the financial crisis, the Korean banks were facing credit risks that were generally increased throughout the country, but states that such risks were different from the "country risk factors" as faced by foreign banks. Korea submits that such increased risks during the crisis were taken into account by the Korean banks (as by foreign banks) as the "general credit risk" of the Korean shipyards.
7.150.
While Korea has provided evidence to the effect that foreign banks included country risk spreads when providing APRGs to Korean shipyards,86 we consider that the EC has established that
at least something equivalent to a country risk spread would also have been included by domestic APRG providers. A country risk spread is generally applied whenever a financial institution incurs international exposure. This will occur when a currency exposure is incurred. A currency exposure is incurred by both domestic and foreign providers of APRGs to Korean shipyards, because APRGs are provided to Korean shipyards in a foreign currency, i.e., US dollars.
7.151.
In addition, we note that an APRG guarantees the repayment of pre-payments in the event of default by the shipyard. The risk of default by the shipyard is related to the general economic conditions in Korea, and would therefore form part of the country risk assessment. This risk applies to both domestic and foreign banks.
7.152.
For these reasons, we proceed on the basis that APRG rates offered by Korean banks also reflect a country risk spread, or something equivalent thereto.

(iv) Domestic market benchmark

7.153.
Korea also rejects the use of certain Korean domestic banks as a market benchmark. Korea asserts that domestic rates should only be taken into account (in fixing an appropriate market benchmark) if they represent a statistically representative number of transactions. Thus, rates charged by domestic entities that only provided APRGs rarely or exceptionally should not be taken into account.
7.154.
The EC asserts that such domestic banks only provided APRGs exceptionally because they could not compete with the beneficial terms offered by KEXIM.
7.155.
In our view, the exceptional nature of any market APRG (be it domestic or foreign) should not preclude its use as an appropriate market benchmark for the purpose of determining the existence of "benefit". Provided it is negotiated on a commercial basis by a market operator, and is comparable in terms of duration etc., any APRG should be admissible as a market benchmark. Korea has submitted no evidence demonstrating that this was not the case for the domestic APRGs relied upon by the EC as market benchmarks.

(v) Credit risk spread

7.156.
The EC claims that APRGs issued by KEXIM before 12 March 1998 conferred a benefit because the terms did not include any credit risk spread.
7.157.
In Attachment 6 to its first written submission, Korea states that "[u]ntil March 1998, KEXIM did not take credit risks into account for its APRG transactions". There is therefore no disagreement between the parties regarding the factual element of the EC's claim. Regarding the legal issue of whether or not the absence of credit risk spreads confers a "benefit", we consider that market operators of necessity would take account of credit risk, and that the failure by KEXIM to include a credit risk spread would result in APRGs being offered on terms that are more favourable than those offered on the market. In this regard, we note that at page 62 of its first written submission, Korea stated (in respect of PSLs, but the same principle would apply in respect of APRGs) that "Chapter 2 of the Interest Rate Guidelines provides detailed standards for determining the interest rates including the base rates and spreads requiring in particular to take a credit risk spread into account in the same way as a private financial institution would." (emphasis supplied, footnote omitted) This would suggest that Korea accepts that failure to apply a credit risk spread is inconsistent with market behaviour (i.e., the behaviour of a private financial institution). In light of the above, we find that the following pre-12 March 1998 APRGs confer a "benefit" by virtue of KEXIM's failure to include a spread for credit risk:

[BCI: Omitted from public version.]

(vi) Past subsidies

7.158.
Korea submits that an alleged subsidy may only be challenged if it is conferring a benefit at the time that the dispute settlement proceeding is initiated. According to Korea, Article 1.1(a) (1) and (2), Article 3 and Article 4 do not apply to subsidies that were granted in the past, and that are not currently being maintained.
7.159.
The EC submits that there is no rule in the WTO that a violation is forgiven once it is in the past. According to the EC, Korea confuses the issue of whether a subsidy has been granted with countervailing duty principles, which only allow current benefit to be offset.
7.160.
Despite Korea's abovementioned argument, Korea stated in response to Question 108 from the Panel that it was "not making a general argument that the EC cannot challenge alleged past subsidies as a matter of principle". In light of Korea's reply, we see no need to rule on whether or not the EC is entitled to challenge "past" subsidies.87

(vii) Adverse inferences

7.161.
The EC requests adverse inferences on the basis of Korea's alleged failure to provide information regarding APRGs issued by independent entities after 28 May 2001. The EC requests the Panel to find in accordance with paragraph 7 of Annex V of the SCM Agreement that these banks either stopped issuing APRGs because they determined that they could not compete with KEXIM’s low premia, or that they continued issuing APRGs at comparatively higher premia than KEXIM.
7.162.
We note that the EC has requested an adverse inference on the basis of paragraph 7 of Annex V of the SCM Agreement. Korea argues that adverse inferences cannot be drawn on the basis of that provision in the context of claims brought under Part II of the SCM Agreement. We do not consider it necessary to resolve this legal issue, since in any event it is well established that WTO dispute settlement panels retain a residual authority88 to draw adverse inferences outside of the circumstances set forth in Annex V. Thus, even if that provision does not apply in respect of Part II claims, our residual authority to draw adverse inferences remains.
7.163.
As a factual matter, however, we consider that the EC has failed to establish that an adverse inference would be warranted. The EC request is based on Korea's alleged failure to provide information regarding APRGs issued by private banks to Daewoo-SME/Daewoo-HI after 28 May 2001, to Samho-HI/Halla-HI after 1 December 2000, and to STX/Daedong after 14 September 1999. The EC asks the Panel to infer (because of Korea's failure to provide the relevant information) that such APRGs were issued at rates higher than those charged by KEXIM. However, the EC has failed to provide any evidence that such APRGs were actually provided to the shipyards concerned after the dates specified by the EC, whereas Korea submits that it has provided all information regarding APRGs in its responses to the Annex V questions. In the absence of any evidence regarding the existence of the alleged APRGs, there is no basis for us to draw any inference regarding the terms on which such alleged APRGs were issued.89

(b) Transaction-specific issues

(i) Daewoo

7.164.
The EC has proposed a number of market benchmarks to argue that certain APRGs90 provided by KEXIM to DHI or DSME were below market rates. In particular, the EC has compared certain APRGs provided by KEXIM in 1998, 2000 and 2001 with a number of APRGs provided by [BCI: Omitted from public version] during that period.
7.165.
Korea rejects the market benchmarks proposed by the EC, and submits that the rates of KEXIM's APRGs should be compared with APRGs provided to those companies by KDB. Korea rejects the use of the EC's [BCI: Omitted from public version] benchmarks APRGs because, in Korea's view, those APRGs did not involve collateral of the same value as that provided in respect of the relevant KEXIM APRGs, i.e., Yangdo Dambo.91
7.166.
In response, the EC submits (on the basis of information provided by Korea in its reply to Question 14 from the EC) that the APRGs issued by [BCI: Omitted from public version] were guaranteed by cash deposits, which the EC claims is a stronger form of collateral than Yangdo Dambo.
7.167.
Information provided by Korea confirms the EC argument that the [BCI: Omitted from public version] APRGs were collateralized by cash deposits.92 The EC has provided a convincing explanation in support of its argument that cash deposits are a stronger form of collateral than Yangdo Dambo. The EC asserts that Yangdo Dambo "presents [] a wide spread depending on the financial situation of the shipyard, the quality of the shipyard work, its on-time delivery record, the evolution of the construction of the vessel, the sales price versus production cost of the ship, the technological requirements of the buyer and the ease with which these will be met".93 The EC also states that "[i]t is not unusual for instance that the final price paid by the purchaser of the ship be dependent upon a certain number of technical characteristics of the ship such as its speed for instance. In such a case, the final value of the ship will only be known at the end of the production process. The value of such a collateral is therefore highly dependent upon the quality of the shipyard."94 The value of Yangdo Dambo is therefore dependent on a number of variables, whereas the value of a cash deposit is self-evident. Korea has failed to rebut the EC's arguments. Instead, Korea asserts that the cash deposits provided in respect of the [BCI: Omitted from public version] APRGs covered only a small portion of the guarantee, compared with the 100 per cent coverage of the Yangdo Dambo provided in respect of the KEXIM APRGs. In order for this argument to prevail, Korea would need to demonstrate that the collateral value of the Yangdo Dambo exceeds that of the cash deposits. We consider that Korea has failed to do so, since Korea has made no attempt to do so.95 We consider such a demonstration to be particularly necessary since Korea has not challenged the EC's assertion that "cash deposits [] are one of the strongest forms of collateral compared to Yangdo Dambo".96
7.168.
We do not consider that the EC could have done more to prove its argument that the [BCI: Omitted from public version] collateral matched that of the KEXIM Yangdo Dambo. The EC was dependent on Korea for information regarding the value of the [BCI: Omitted from public version] cash deposits. The EC was also dependent on Korea for information regarding the value of the Yangdo Dambo required by KEXIM. However, in its reply to Question 67 from the Panel (concerning one of the [BCI: Omitted from public version] transactions), Korea indicated that it could not provide worksheets or other documentation regarding KEXIM's consideration of collateral. Korea was only able to provide very basic information regarding the Yangdo Dambo at issue.97 Given the number of variables that determine the collateral value of Yangdo Dambo, this limited information is not sufficient to conclude that the value of the Yangdo Dambo exceeds that of the relevant cash deposits.
7.169.
Furthermore, even though Korea seeks to rely on the limited value of the cash deposits provided in respect of the [BCI: Omitted from public version] APRGs, Korea has failed to establish the precise scope of those cash deposits. Thus, at note 155 to its first written submission, Korea reported the collateral for the [BCI: Omitted from public version] APRGs as "a pledge against bank deposits amount[ing] to 20 to 30% of the advance payments". Then, in reply to Question 14 from the EC, Korea stated that the cash deposits for the four [BCI: Omitted from public version] APRGs amounted to 10, 12, 20 and 20 per cent respectively, whereas the cash deposit for the [BCI: Omitted from public version] APRG was 30 per cent. As the party seeking to rely on the allegedly limited coverage of the relevant cash deposits, Korea should at least have stated clearly what that allegedly limited coverage was. Furthermore, Korea has failed to provide any evidence in support of its reporting of the coverage of the [BCI: Omitted from public version] cash deposits. Such evidentiary support is particularly necessary in this case, given the differences in Korea's reporting of the cash deposits at issue.
7.170.
In light of the above, there is no basis for us to accept Korea's argument that the [BCI: Omitted from public version] market benchmarks proposed by the EC should be rejected because the relevant collaterals were not comparable. In the absence of additional argumentation by Korea, we consider that the 1998, 2000 and 2001 KEXIM APRGs identified at Figure 11 of the EC's first written submission should be compared with those benchmarks in order to determine whether or not they were provided on terms more favourable than those available on the market. Since the KEXIM premia rates were less than the market benchmarks proposed by the EC, we find that the 1998, 2000 and 2001 KEXIM APRGs identified at Figure 11 of the EC's first written submission conferred a benefit and therefore constitute subsidies.
7.171.
The EC has also proposed APRGs offered by [BCI: Omitted from public version] and [BCI: Omitted from public version] as market benchmarks against which to compare KEXIM APRGs provided to Daewoo in 1997. It is not necessary for us to review the EC's argument, since we have already found that KEXIM's pre- March 1998 APRGs constituted subsidies because the terms did not include any credit risk spread.98

(ii) Samho/Halla

7.173.
The EC has proposed a number of market benchmarks to argue that certain APRGs provided by KEXIM to Samho / Halla in 2000 were below market rates. The relevant KEXIM APRGs are set forth at Figure 12 of the EC's first written submission. Korea asserts that the KEXIM APRGs are not comparable with the EC's proposed market benchmarks, since the latter were not collateralized whereas the former were provided on the basis of Yangdo Dambo. Korea submits that the KEXIM APRGs should instead be compared with a number of APRGs provided by KDB.102Korea also asserts that the KEXIM APRGs identified by the EC were not provided to Samho / Halla.
7.174.
Regarding Korea's reliance on KDB APRGs, we recall our finding that KDB is a public body. KDB rates therefore cannot be used as a market benchmark.
7.175.
Regarding Korea's argument on collateralization, the EC refers to its translation of a document submitted by Korea in the Annex V procedure, and asserts that the KEXIM APRGs were issued "on credit", and therefore without collateral.103 In its first written submission, however, Korea stated that KEXIM APRGs were always provided against "real property (land, buildings, factories as a whole), personal property, securities (i.e., 'Yangdo Dambo' [...] on significant items such as the hull of a vessel) and guarantees."104 In response to Question 67 from the Panel, Korea has demonstrated that KEXIM APRG transactions reported by Korea as being "on credit" actually involved the provision of collateral such as Yangdo Dambo.105 There is therefore no basis to doubt that the KEXIM APRGs provided to Samho / Halla and reported as being "on credit" were actually collateralized. Since the EC has failed to demonstrate that its proposed market benchmarks were also collateralized, we are unable to accept the use of those proposed market benchmarks for comparing with collateralized KEXIM APRGs. We therefore reject the EC's claim against the KEXIM APRGs identified in Figure 12 of its first written submission.
7.176.
In response to a Korean argument that some of the KEXIM APRGs included in Figure 12 of the EC's first written submission did not relate to Samho, the EC performed an additional benefit analysis in respect of other KEXIM APRGs provided to Samho / Halla.106 However, the EC's modified analysis is based on the same proposed market benchmarks, which (as described above) we are unable to compare with collateralized KEXIM APRGs. Since the EC has given us no reason to doubt Korea's assertion that all KEXIM APRGs were collateralized, we reject the EC's modified analysis in respect of the additional107 KEXIM APRGs identified by the EC. We therefore reject the EC's claim against those additional KEXIM APRGs provided to Samho / Halla.

(iii) STX/Daedong

7.177.
The EC submits that two KEXIM APRGs issued to STX/Daedong in 1999 (see Figure 13 of the EC's first written submission) constitute prohibited export subsidies. The EC claims that the terms of the KEXIM APRGs were more favourable than those of three APRGs provided to STX / Daedong by [BCI: Omitted from public version] in 1999. Korea opposes any comparison with APRGs provided by [BCI: Omitted from public version]. Korea asserts that those APRGs were not collateralized, whereas KEXIM's APRGs were. Korea submits that the KEXIM APRGs should instead be compared with APRGs provided by the Korea Exchange Bank ("KEB") in 2001 and KDB in 2002.
7.178.
We recall that we have already rejected the use of two of the three [BCI: Omitted from public version] APRGs.108 We shall therefore examine the EC's claim in light of the terms and conditions of the one remaining [BCI: Omitted from public version] APRG. Regarding comparability on the basis of collateralization, we recall Korea's argument that all KEXIM APRGs were collateralized. KEXIM APRGs should therefore be compared with collateralized market benchmarks. In this regard, we note that the remaining [BCI: Omitted from public version] APRG was collateralized "with cash deposits amounting to 5% of the advance payment amount and an export guarantee insurance".109 Since Korea has failed to argue that such collateral is not comparable with any collateral required in respect of the relevant KEXIM APRGs, we consider that it is appropriate to compare the two 1999 KEXIM APRGs with the single [BCI: Omitted from public version] APRG provided to STX / Daedong in 1999. On the basis of such comparison, we find that the KEXIM APRGs were below market, and therefore conferred a benefit.
7.179.
We recall that Korea has argued that the terms of the KEXIM APRGs should be compared with the terms of certain APRGs provided by KDB and KEB. In this regard, we recall our finding that KDB is a public body. Accordingly, KDB APRGs do not constitute a reliable market benchmark with which to asses the existence of benefit. As for KEB, we note that Korea has sought to rely on APRG rates offered by KEB in 2001, whereas the KEXIM APRGs at issue date from 1999. Given the absence of any temporal correlation between the KEXIM APRGs challenged by the EC and the KEB APRGs identified by Korea, and our preference for rates charged by entities without government ownership, we do not consider it appropriate to use KEB APRGs as a market benchmark for determining whether or not the KEXIM APRGs conferred a benefit. We also note the EC argument that KEB is not a reliable market benchmark as a result of government entrustment or direction110 and government ownership.111 We are not persuaded by the EC's argument concerning government entrustment or direction, since the only evidence provided by the EC of the government's alleged role is in relation to KEB's participation in the Daewoo workout, and has nothing to do with KEB's provision of APRGs to Samho / Halla. Regarding government ownership, the EC asserts that GOK has a minority shareholding in KEB. In choosing an appropriate market benchmark, we consider it preferable to choose when possible entities without any government ownership. The absence of government ownership generally removes the possibility that rates have been fixed on the basis of public policy, rather than commercial principles.
7.180.
In light of the above, we uphold the EC's claim against the KEXIM APRGs identified in Figure 13 of the EC's first written submission.

(iv) Hanjin

7.181.
The EC claims that two APRGs provided by KEXIM to Hanjin in 2002 constitute subsidies because they were provided on terms more favourable than those of two APRGs provided by [BCI: Omitted from public version] to Hanjin in the same year. The relevant APRGs are set forth at Figure 14 of the EC's first written submission.
7.182.
Korea opposes the comparison proposed by the EC because of differences in collateralization.Korea submits that the relevant KEXIM APRGs should instead be compared with certain APRGs provided by [BCI: Omitted from public version], KEB, KDB and [BCI: Omitted from public version] in 1997.112
7.183.
In its reply to Question 71, Korea submits that no security was deposited for the [BCI: Omitted from public version] APRGs, whereas Yangdo Dambo was provided for the relevant KEXIM APRGs. Korea submits that such differences in collateralization preclude any comparison between the [BCI: Omitted from public version] and KEXIM APRGs at issue. The EC submits that Korea's argument that the two APRGs issued by [BCI: Omitted from public version] were not collateralized is inconsistent with Korea's statement in the Annex V process113 that "[a]ll APRGs" reported by Korea, including therefore those from [BCI: Omitted from public version], "were issued with the collaterals of Yangdo Dambo".
7.184.
Despite Korea's contradictory statement in the Annex V process, Korea provided documentary evidence114 during these proceedings demonstrating that the two [BCI: Omitted from public version] APRGs at issue were not collateralized. The EC has not disputed that the KEXIM APRGs at issue were, by contrast, collateralized. Given these differences in collateralization, and the fact that the EC has not provided us with any basis for making adjustments to reflect such differences, we are unable to determine benefit on the basis of a comparison of the [BCI: Omitted from public version] APRGs with the KEXIM APRGs. We are therefore unable to accept the use of the [BCI: Omitted from public version] APRGs as a market benchmark, and reject the EC's claim against the Hanjin APRGs accordingly.
7.185.
In light of the above finding, it is not strictly necessary for us to consider the [BCI: Omitted from public version], KEB, KDB and [BCI: Omitted from public version] benchmarks proposed by Korea. We shall do so, however, for the sake of completeness. We have already indicated that KDB APRGs would not constitute an appropriate market benchmark, since KDB is a "public body". Although the EC does not argue that [BCI: Omitted from public version], KEB and [BCI: Omitted from public version] are public bodies, it asserts that they are entrusted or directed by the government.115 We must reject this argument, since the EC has not provided any evidence that these entities were entrusted or directed to provide APRGs to Hanjin. The only evidence of alleged government entrustment or direction of these entities relates to their participation in some of the restructurings at issue in these proceedings. Nevertheless, we note that the alternative APRGs proposed by Korea as market benchmarks all relate to 1997, whereas the KEXIM APRGs challenged by the EC relate to 2002. Given the lack of temporal correlation between these APRGs, the [BCI: Omitted from public version], KDB, KEB and [BCI: Omitted from public version] APRGs do not constitute appropriate market benchmarks for the purpose of assessing the KEXIM APRGs at issue.

(v) Samsung

7.186.
In respect of Samsung, the EC challenges four KEXIM APRGs, all of which were issued in 1997. The EC claims that these APRGs conferred a benefit because they were issued on terms more favourable than 1997 APRGs provided by [BCI: Omitted from public version] and [BCI: Omitted from public version]. The relevant APRGs are set forth at Figure 15 of the EC's first written submission.
7.187.
We recall that we have already found that KEXIM's pre- March 1998 APRGs to Samsung constitute subsidies because their terms did not include any credit risk spread. For this reason, there is no need for us to consider additional arguments by the parties regarding the comparability of these APRGs with the market benchmark APRGs proposed by the EC.
7.188.
For the sake of completeness, we note Korea's argument regarding the use of KDB APRGs as a market benchmark. We recall that KDB is a public body.116 As a public body, KDB does not constitute a market benchmark for the purpose of assessing whether or no the relevant KEXIM APRGs conferred a benefit.

(c) Export contingency

(i) Arguments of the Parties

7.189.
The EC submits that APRGs are provided for the specific purpose of guaranteeing the down payments for Korean goods intended for export. The EC asserts that any transaction financed through KEXIM's APRG programme must, by definition, be an export transaction, and that APRGs provided pursuant to the programme are expressly contingent on export within the meaning of Article 3.1(a) of the SCM Agreement.
7.190.
When asked by the Panel (Question 48) whether Korea contests the EC's claim that PSLs and APRGs under the KEXIM legal regime are contingent on export performance, Korea responded that it "has not taken any position as to whether [PSLs and APRGs] are contingent on export performance". In response to an oral question from the Panel at the second substantive meeting, Korea stated that it was not contesting the EC's claim of export contingency.

(ii) Evaluation by the Panel

7.191.
We find that the EC has established a prima facie case that KEXIM APRGs are contingent on export performance. In light of Korea's decision no to contest the EC's arguments regarding this matter, we find that KEXIM APRGs are "contingent... upon export performance" within the meaning of Article 3.1(a) of the SCM Agreement.

(d) Specificity

(e) Item (j) defence

7.193.
In light of our findings that certain APRGs constitute subsidies, and that such subsidies are contingent on export performance, we will be required to find that such APRGs are inconsistent with Articles 3.1(a) and 3.2 of the SCM Agreement unless we uphold Korea's claim that they benefit from a safe haven pursuant to item (j) of the Illustrative List.
7.194.
Korea submits that item (j) should be interpreted such that export credit guarantee programmes, or guarantee programmes against increases in the cost of exported products, that are provided at premium rates that cover the long-term operating costs and losses of the programmes should be found not to constitute export subsidies. In other words, Korea relies on an a contrario interpretation of item (j). Korea's reliance on item (j) raises a number of issues. First, is an a contrario interpretation of item (j) permissible? Second, if so, what are the relevant conditions to be fulfilled? Third, have those conditions been fulfilled in this case?

(i) Is an a contrario interpretation permissible?

7.195.
We note that the panel in Brazil – Aircraft – Article 21.5 was required to consider whether or not one of the items of the Illustrative List could be interpreted in an a contrario manner. In doing so, that panel observed that footnote 5117 to the SCM Agreement provides an explicit textual basis for determining whether and under what conditions the Illustrative List may be used to demonstrate that a measure is not a prohibited export subsidy.118 The panel noted that footnote 5 provides that "[m]easures referred to in Annex I as not constituting export subsidies shall not be prohibited under this or any other provision of this Agreement." The panel observed that, in its ordinary meaning, footnote 5 relates to situations where a measure is referred to as not constituting an export subsidy. The panel therefore considered whether or not the Illustrative List provision at issue contained any affirmative statement that a measure is not an export subsidy, or that a measure not satisfying the conditions of that provision is not prohibited, and thus falls within the scope of footnote 5.119 Finally, the panel noted that a broad reading of footnote 5 could place developing country Members at a permanent, structural disadvantage in the field of export credit terms, a result that it considered to be inconsistent with one of the objects and purposes of the WTO Agreement.120
7.196.
We find the reasoning121 expressed by the Brazil – Aircraft-Article 21.5 panel to be convincing, and consider it appropriate for us to be guided by it in these proceedings. We are of course aware that the Brazil – Aircraft-Article 21.5 panel was considering whether or not the first paragraph of item (k) of the Illustrative List could be interpreted a contrario. However, the panel's reasoning was based in the first instance on an interpretation of footnote 5 of the SCM Agreement, and was not confined to the text of item (k) in isolation. Furthermore, the panel's concerns regarding structural discrimination against developing country Members were based explicitly on considerations regarding item (j). Accordingly, there is no reason why the panel's reasoning should not also be applied in respect of the remaining provisions of the Illustrative List, including item (j).
7.197.
Korea argues that we should not follow the reasoning of the Brazil – Aircraft – Article 21.5 panel because it was invalidated by the Appellate Body. In particular, Korea relies on the statement by the Appellate Body in those proceedings that "[i]f Brazil had demonstrated that the payments made under the revised PROEX were not 'used to secure a material advantage in the field of export credit terms', and that such payments were 'payments' by Brazil of 'all or part of the costs incurred by exporters or financial institutions in obtaining credits', then we would have been prepared to find that the payments made under the revised PROEX are justified under item (k) of the Illustrative List."122 However, we do not accept that this amounts to a reversal of the panel's findings, nor a legal finding by the Appellate Body that an a contrario interpretation of the first paragraph of item (k) is permissible. This is because the Appellate Body explicitly stated that "[i]n making this observation, we wish to emphasize that we are not interpreting footnote 5 of the SCM Agreement, and we do not opine on the scope of footnote 5, or on the meaning of any other items in the Illustrative List." In light of this clarification by the Appellate Body, we consider that there is nothing in the Appellate Body statement that would cause us not to be guided by the abovementioned reasoning of the Brazil – Aircraft – Article 21.5 panel.
7.198.
Thus, in order to determine whether or not item (j) of the Illustrative List may be interpreted a contrario, we shall consider whether or not item (j) falls within the scope of footnote 5 of the SCM Agreement. That is to say, we shall consider whether item (j) contains any affirmative statement that a measure is not an export subsidy, or that a measure not satisfying the conditions of that paragraph is not prohibited. Item (j) contains no such affirmative statement. Item (j) merely describes certain circumstances in which particular programmes shall constitute export subsidies. Since item (j) therefore falls outside the scope of footnote 5, item (j) does not provide a basis on which to find that measures do not constitute prohibited export subsidies.
7.199.
Although the EC rejects the application of item (j) on the basis of the facts of this case, the EC submits that in law item (j) could be read to include a proviso, and thereby "refer" to export credit guarantees as not constituting export subsidies to the extent that the premium rates cover the long-term operating costs and losses of the programmes. As indicated above, however, we consider that item (j) does not refer to measures as not constituting export subsidies. The terms of item (j) merely refer to export credit guarantees etc. at premium rates which are inadequate to cover long-term operating costs and losses. It contains no explicit reference to export credit guarantees etc. that are adequate to cover long-term operating costs and losses. There is therefore no basis for claiming that item (j) somehow incorporates a proviso that brings it within the scope of footnote 5.
7.200.
Korea objects to the Brazil – Aircraft – Article 21.5 panel's interpretation and application of footnote 5, which we apply here. Korea considers that such an application is overly narrow. Korea argues that the negotiating history of footnote 5 shows that the drafters intended to expand, rather than restrict, the scope of footnote 5. In this regard, Korea makes the same argument as was advanced by the United States in Brazil – Aircraft – Article 21.5. The Brazil – Aircraft – Article 21.5 panel described and rejected this argument in the following terms:

The United States advances arguments based on the negotiating history of footnote 5 in support of its broad interpretation of that footnote to apply to the first paragraph of item (k). In this respect, it points out that in a Chairman's text of the SCM Agreement known as Cartland III, footnote 5 provided as follows:

"Measures expressly referred to as not constituting export subsidies shall not be prohibited under this or any other provision of this Agreement." (emphasis added).

As the United States correctly observes, a new Chairman's text (known as "Cartland IV") was released just a few days later. In that new text, the word "expressly" was dropped from the footnote, which took its present form. In the view of the United States, this change demonstrates that the drafters "intended to expand, rather than restrict" the scope of footnote 5, and that "they did not intend the sort of narrow construction of footnote 5 advanced by Canada and the EC."

We agree with the United States that the deletion of the term "expressly" appears to have broadened the scope of footnote 5 in Cartland IV beyond its scope in Cartland III. We do not agree, however, that it served to broaden footnote 5 to the extent suggested by the United States. As we discussed above, the Illustrative List contains – and already contained at the time of Cartland III and IV – a number of provisions that include affirmative statements that arguably represent authorizations to use certain measures. The language of Cartland III ("expressly referred to") could have precluded asserting that footnote 5 applied to any of these provisions, and it may be that the purpose of the modification was to rectify this situation. If on the other hand the intention of the drafters in changing footnote 5 had been to extend the scope of that footnote to cover situations where the Illustrative List merely referred to things that were export subsidies, they might have been expected to modify the structure of the second part of the footnote, and not merely delete the word "expressly". At the very least, we conclude that the implications of the negotiating history referred to by the United States are inconclusive and cannot lead us to disregard the ordinary meaning of the footnote.

Of course, it could be argued that, based on an a contrario argument, the Illustrative List permits admitted export subsidies even where those subsidies do not fall within the scope of footnote 5. As we have already indicated, however, the drafters have provided us with a specific textual provision that addresses the issue when the Illustrative List can be used to demonstrate that a measure is not a prohibited export subsidy. The fact that this footnote was adjusted on at least one occasion suggests that the drafters gave this issue consideration and provided the answer to this question. If we were to conclude that the Illustrative List by implication gave rise to "permitted" measures beyond those allowed by footnote, we would be calling into serious question the raison d'être of footnote 5.123

7.201.
We find the panel's reasoning to be convincing. For this reason, and in light of the fact that Korea has not attempted to rebut it, we see no reason not to be guided by that reasoning in the present case. Although Korea notes that the Appellate Body has not opined on the Brazil – Aircraft – Article 21.5 panel's reasoning,124 this fact does not make the panel's reasoning any less convincing to us. On the basis of that reasoning, we reject Korea's argument concerning the negotiating history of footnote 5.
7.202.
Korea also argues that the only purpose of item (j) is to indicate when export subsidies do not exist.125 Korea submits that failure to permit an a contrario reading of item (j) and the first paragraph of item (k) would render those provisions meaningless, since a complaining party would never rely on them to demonstrate the existence of an export subsidy. According to Korea, this is because reliance on these provisions would require a complaining party to demonstrate more facts than the basic legal provision in Article 3. In particular, Korea argues that since one could establish a violation of Article 3.1(a) simply by showing export contingency, a complainant would never have recourse to item (j) to demonstrate export subsidization, because that would require the complainant to also demonstrate that premium rates are inadequate to cover the long-term operating costs of the relevant programme (i.e., that the conditions of item (j) are met).
7.203.
We cannot accept Korea's argument that the only purpose of item (j) is to indicate when export subsidies do not exist. To accept this argument would require reading item (j) in a sense exactly opposite that of its plain language: on its face, item (j) defines certain circumstances in which export credit guarantee programmes are export subsidies. Item (j) simply does not address export guarantee programmes that do cover their long-term operating costs and losses. Indeed, what is the point of footnotes 1 and 5 being included in the SCM Agreement to clarify the relationship between Articles 1 and 3 of the SCM Agreement on the one hand, and the Illustrative List on the other, if footnotes 1 and 5 can ultimately be ignored and the Illustrative List read in a sense exactly the opposite of its plain language?
7.204.
As for Korea's argument that item (j) would never be used by a complaining party to establish the existence of a prohibited export subsidy, we disagree. As item (j) provides, if a complaining party establishes that another Member's export guarantee programme fails overall to cover its long-term operating costs and losses, that is sufficient for a finding that the programme as a whole constitutes a prohibited export subsidy. Given the per se nature of the items set forth in the Illustrative List, no further separate analysis of the programme under Articles 1 and 3 would be necessary.126 Furthermore, in arguing that all a party would need to do to pursue a claim under Article 3.1(a) is establish export contingency, Korea overlooks the need under Article 3.1(a) to demonstrate the existence of subsidization. Thus, a complainant cannot establish a violation of Article 3.1(a) simply by showing export contingency; the complainant must also demonstrate subsidization, and therefore benefit, on the basis of Article 1.1 of the SCM Agreement.
7.205.
Furthermore, we recall that the Brazil – Aircraft –Article 21.5 panel's reasoning was based explicitly on the structural disadvantages for developing country Members that would result from an a contrario interpretation of item (j). The panel found:

6.59. The same situation exists in respect of item (j) of the Illustrative List. Brazil argues that its interpretation of the first paragraph of item (k) is necessary to allow it to meet export credit terms provided by developed country Members through export credit guarantees. If footnote 5 is interpreted broadly to encompass the first paragraph of item (k), however, it presumably would also apply to item (j) and thus "permit" export credit guarantees at premium rates adequate to cover long-term operating costs and losses, even where the guarantees constituted a subsidy contingent upon export performance within the meaning of Article 3.1(a). As Canada points out, however, in the case of a government guarantee, a lending bank establishes financing terms in light of the risk of the guarantor government, not the borrower. Developed countries generally present a lower risk of default than developing countries, and a developing country may often be perceived as posing a higher risk than even the borrower to whom a guarantee might be extended. As a result, while developing countries in theory could utilise any "safe harbour" under item (j) to provide loan guarantees at the same premium rates as developed countries, the effect of guarantees by developing country Members on the interest rate of the guaranteed export credits would be minimal or non-existent in most cases. In other words, a broad reading of footnote 5 would, in respect of item (j), allow developed countries to support export credits at interest rates that would be consistently lower than those of export credits supported by developing countries.

6.60. If, on the other hand, we interpret footnote 5 in accordance with its ordinary meaning, and conclude that it does not apply to items such as the first paragraph of item (k) and item (j), then all WTO Members are faced with a common set of rules in respect of export credit practices. First, they can ensure that those practices do not confer a benefit within the meaning of Article 1 and are therefore not subsidies. Because the existence of benefit is determined based on the existence of a benefit to a recipient, and without regard to whether there is a cost to the government, all Members compete on a level playing field in respect of this assessment, i.e., a measure which constitutes an export subsidy when provided by Brazil ipso facto will also constitute a subsidy when provided by Canada, and vice versa.127

7.206.
We share that panel's concerns regarding the structural disadvantages that would result for developing country Members from an a contrario interpretation of item (j). As noted by that panel, an interpretation leading to such structural disadvantages "would be at odds with one of the objects and purposes of the WTO Agreement generally and the SCM Agreement specifically".128
7.207.
In light of the above, we find that an a contrario interpretation of item (j) is not permissible. Strictly speaking, therefore, it is not necessary for us to continue with our analysis of the two remaining issues outlined above. For the sake of completeness, however, we shall do so.

(ii) If an a contrario interpretation of item (j) were permissible, what conditions would need to be fulfilled?

7.208.
Assuming arguendo that item (j) could operate as an affirmative defence (with which, as noted, we disagree), it would need to be demonstrated that the APRG programme constitutes an export credit guarantee or insurance programme, or an insurance or guarantee programme against increases in the cost of exported products or of exchange risk, that operates at premium rates which are adequate to cover the long-term operating costs and losses of that programme.

(iii) Are the relevant conditions for the application of item (j) as an affirmative defence fulfilled?

7.209.
As the party seeking to rely on item (j) as an affirmative defence, the burden of proof is on Korea to establish that the relevant conditions have been fulfilled. Korea submits that KEXIM APRGs are export credit guarantees or, at least, guarantees against increases in the costs of exported products. Korea further submits that the APRG programme covers its long-term operating costs, as evidenced by the profitability of that programme from 1997 to 2002.
7.210.
The EC contests Korea's assertion that APRGs are export credit guarantees, or guarantees against increases in the cost of exported products. The EC asserts that export credit guarantees are provided to banks or exporters in respect of credits they have provided to foreign customers, whereas APRGs involve guarantees provided to the foreign customers in respect of payments they have made to shipbuilders. The EC also asserts that APRGs guard against the overall expenses of the exporter or credit risks taken by the purchaser, not against increases in the cost of the exported product. The EC argues that Korea’s broad interpretation would allow any subsidy to an exporter or to exported products that is formulated as a "guarantee programme" to be covered by item (j) since any such subsidy would tend to reduce the cost of manufacturing the exported goods for the exporter or of buying the exported goods for the purchaser.
7.211.
The EC further asserts that, in making APRGs and preshipment loans, KEXIM assumes a risk that relates to the creditworthiness of the domestic exporters. The EC submits that the export credit financing referred to in item (j) concerns foreign risk. According to the EC, the underlying rationale of these provisions is that domestic banks typically do not have the means of assessing overseas risks of a potential buyer of an export product (or of recovering money abroad).
7.212.
The EC does not challenge Korea's assertion regarding the long-term profitability of the APRG programme.

Do APRGs constitute export credit guarantees?

7.213.
From a purely textual perspective, based entirely on the plain meaning of the words, we consider that an instrument may only be designated as an "export credit guarantee" if it guarantees an export credit. An instrument will guarantee an export credit if it covers default by a borrower in respect of an export credit provided to that borrower.
7.214.
Korea has not explicitly argued that APRGs guarantee export credits. Rather, Korea has sought to establish a link between the APRGs and export credits by arguing that "APRGs are issued to protect the shipowner against a contractual default by the shipbuilder. It is not disputed that Korean exporters who export capital goods which qualify for loans under KEXIM policies on export loans are also eligible for APRGs. There is, therefore, a close connection between the export loan / credit financing and the APRGs even though there is no complete concurrence."129 Thus, although Korea does not argue that APRGs guarantee export credits (Korea asserts, instead, that APRGs guarantee against contractual default by the shipbuilder), it argues that there is a sufficiently close connection between credits to exporters130 and APRGs for the latter to be treated as export credit guarantees. We note, however, that there is no direct connection between APRGs and loans to exporters, since there may well be cases in which APRGs are issued but no KEXIM loans (or, in theory at least, any loans at all) are provided to the exporters.
7.215.
In reply to Question 59 from the Panel, Korea stated that "[t]he fact that APRGs may be granted when export credits in the narrow sense are not does not prevent APRGs from being qualified as export credit guarantees because it still is a guarantee accessory to an export transaction similar to a loan guarantee which covers a default by the borrower." If we understand Korea correctly, it is seeking to demonstrate a closer link between APRGs and loans to exporters by treating advance payments by the buyers as loans to exporters, and the APRGs as guarantees of those loans. We are unable to accept this argument. First, we see no basis for treating the advance payment as a loan to the shipbuilder. Receiving and making pre-payment is not the same as accepting and offering a loan. A loan is to be repaid, whereas a pre-payment is a payment for a good or service to be provided. Indeed, the fact that Korea merely argues that the APRG is "similar to" a loan guarantee suggests that Korea itself is not of the view that the advance payment constitutes a loan.131 Second, even if the advance payment were treated as a loan, the mere fact that a guarantee is issued in respect of a loan in the context of (or "accessory to") an export transaction does not necessarily make that guarantee an export credit guarantee. An instrument will only constitute an export credit guarantee if it guarantees an "export credit". Korea has neither argued nor demonstrated that advance payments guaranteed by APRGs constitute "export credits".132
7.216.
Korea submits that other Members' export credit agencies also provide APRGs. We understand Korea to argue that, because APRGs are offered by other Members' export credit agencies, they should be treated as export credit guarantees. Korea also asserts at note 34 to its replies to questions from the Panel after the second meeting that "if the EC’s narrow and simplistic construct is correct, then a number of EC Member States are prima facie in violation of Part II of the SCM Agreement as their programs would be outside of the parameters of the OECD Arrangement and not protected by any safe harbors." The conformity of the practices of other Members with Part II of the SCM Agreement is not an issue in these proceedings. In resolving the present dispute, therefore, we take no account of the practices of other Members, nor of the potential implications of our findings for the practices of such Members.
7.217.
For the above reasons, we find that KEXIM APRGs do not constitute export credit guarantees within the meaning of item (j) of the Illustrative List.

Do APRGs constitute guarantees against increases in the cost of the exported product?

7.218.
Regarding the issue of guarantee against increases in cost, Korea asserts (response to Question 60 from the Panel) that an APRG guarantees against an increase in the cost associated with the working capital necessary to produce the ship, because the fact that the shipowner makes an advance payment (guaranteed by the APRG) reduces the amount of money that the shipyard needs to borrow to finance the cost of producing the vessel (i.e., the pre-payment serves as interest-free working capital).
7.219.
The EC submits that item (j) does not apply to guarantees against increases in exporters' costs generally, but only to guarantees against increases in "the cost of exported products".
7.220.
We note that item (j) applies to guarantee programmes against increases in the cost of "exported products". The fact that a shipyard has access to interest-free working capital (in the amount of the advance payments guaranteed by APRGs) has no bearing on whether or not the (total) cost of the vessel will increase. It provides neither a guarantee for the shipbuilder that the cost of production of the vessel will not increase, nor a guarantee to the buyer that the price of the vessel will not increase.133 Irrespective of the pre-payment, the cost of the vessel could still increase. APRGs provide no guarantee that other costs, and therefore the (total) cost of the exported product, will not increase. Accordingly, there is no basis for us to find that APRGs constitute "guarantee programmes against increases in the cost of exported products" within the meaning of item (j) of the Illustrative List.
7.221.
Although the above findings indicate that item (j) could not operate as an affirmative defence in the present case, for the sake of completeness we shall address the issue of the adequacy of APRG premium rates. In this regard, we note that the EC has not disputed Korea's assertion that the APRG premia are adequate to cover the long-term operating costs and losses of the programme. We therefore consider that Korea has established prima facie that this condition is fulfilled.
7.222.
In light of our findings that item (j) may not be interpreted a contrario, and that APRGs are neither export credit guarantees nor guarantees against increases in the cost of exported products, we find that Korea has failed to demonstrate that item (j) is applicable as an affirmative defence for those individual APRGs found to be in violation of Article 3.1(a) of the SCM Agreement.134

(f) Conclusion

7.223.
To conclude on the individual APRG transactions challenged by the EC, we find that the following KEXIM APRGS constitute prohibited export subsidies contrary to Articles 3.1(a) and 3.2 of the SCM Agreement:

[BCI: Omitted from public version.]

5. Individual PSL transactions

7.224.
The EC claims that a number of individual KEXIM PSLs constitute prohibited export subsidies, contrary to Articles 3.1(a) and 3.2 of the SCM Agreement. The EC's claim is based on a comparison of the terms of individual PSLs with a market benchmark constructed primarily on the basis of an index of corporate bond prices.
7.225.
Korea has made a number of arguments concerning the EC's proposed market benchmark. Korea has also proposed alternative market benchmarks of its own. In particular, Korea submits that the terms of KEXIM PSLs should be compared with the terms of other financing instruments used by the relevant shipyards, including corporate bonds issued by those shipyards. In the alternative, Korea submits that the individual PSL transactions benefit from a safe haven provided for in the first paragraph of item (k) of the Illustrative List. We shall examine the parties' arguments regarding the alternative market benchmarks proposed by Korea before addressing issues concerning the market benchmark proposed by the EC. Only then will we examine the individual transactions identified by the EC in its claim. In the event that we uphold any of the EC's claims against individual transactions, we shall then consider Korea's reliance on the first paragraph of item (k) of the Illustrative List.

(a) Alternative benchmarks proposed by Korea

(i) Corporate bonds issued by shipyards

7.226.
Korea notes that the EC has proposed a market benchmark constructed on the basis of an index of corporate bond rates. Korea submits that, if a market benchmark is to be based on corporate bonds, it should at least be based on corporate bonds issued by the individual shipyards at issue. Accordingly, Korea has submitted data concerning corporate bonds issued by DHI/DSME, Samho, STX/Daedong, Hyundai Mipo and HHI.
7.227.
The EC asserts that the bonds actually issued by the shipyards are not appropriate market benchmarks. The EC submits that many of the bonds issued were guaranteed, and that Korea failed to submit details regarding the terms of the guarantees. The EC also complains that Korea often failed to report the identity of the guarantor.135
7.228.
In determining whether or not it is appropriate to use the corporate bond data submitted by Korea as a market benchmark, we shall examine the evidence provided by Korea for each shipyard separately.

DHI/DSME

7.229.
In response to Question 17 from the EC, Korea provided a table containing information in respect of six corporate bonds issued by DHI/DSME in 1997 and 1998.136 The table reports "N/A" in the "collateral / guarantee" column. It is unclear whether Korea intended to report "not applicable" or "not available". The EC argues that at least three of those bonds were guaranteed.137 Although Korea has not disputed this, it has not provided information regarding the terms on which such guarantees were offered. Without this information, we are unable to use the bond data provided by Korea as a market benchmark. In particular, there is no means for us to ensure that the Korean corporate bond data and the PSL rates are directly comparable. Nor is there any means for us to make adjustments for any differences in guarantee / collateralization that may exist. In addition, information submitted by the EC also indicates that the guarantor for at least one of those bonds was KDB, which we have already found to be a public body.138 This confirms our decision not to establish a market benchmark on the basis of this data, since there is a risk that corporate bonds guaranteed by public bodies are not a reliable indicator of market rates.
7.230.
Korea also submitted corporate bond data for 1999 and 2000.139 The data comprise quarterly balances, with average interest rates. The parties disagree as to whether or not such information constitutes a proper basis for a market benchmark.140 We consider that we do not need to address this issue, however, as Korea has in any event failed to report details of any collateral or guarantees provided in respect of these bonds, even though Exhibit EC – 129 demonstrates that many DHI/DSME bonds were issued against guarantees. Again, the absence of such information precludes us from using such data as a market benchmark.
7.231.
In light of the above, we are unable to use the data submitted by Korea in respect of corporate bonds issued by DHI/DSME as a reliable market benchmark.

Samho

7.232.
Korea provided information regarding corporate bonds issued by Samho at para. 233 of its first written submission, and in response to question 17 from the EC.
7.233.
The EC claims that the bonds cannot be used as a reliable market benchmark because they were guaranteed by [BCI: Omitted from public version], thus reflecting the credit rating of [BCI: Omitted from public version] rather than Samho.
7.234.
Korea has not rebutted the EC's argument regarding the [BCI: Omitted from public version] guarantee, nor has it provided any information regarding the terms of that guarantee. We are therefore unable to determine what the rate for the bonds would have been without the guarantee. Furthermore, Korea has reported two-year bond data, which cannot be readily compared with six-month PSL rates. In these circumstances, we are unable to use the Samho corporate bond data submitted by Korea as a basis for establishing a market benchmark against which to judge the KEXIM PSLs.

STX/Daedong

7.235.
Korea provided information regarding corporate bonds issued by STX/Daedong at para. 236 of its first written submission, and in response to question 17 from the EC. Korea asserts that the STX/Daedong bonds constitute particularly appropriate market benchmarks, since they were collateralized, and could therefore be readily compared with the collateralized PSLs.
7.236.
The EC notes that, according to the information provided by Korea, the STX/Daedong bonds were issued in Japanese yen. The EC submits that it is improper to compare the corporate bond interest rates quoted in Japanese yen with the PSL rates quoted in Korean won, as interest rates may differ greatly based on the underlying currency. The EC also asserts that the collateral backing the STX/Daedong bonds was the factory as a whole, whereas the factory was not used as collateral for PSLs.
7.237.
We note that Korea has not rebutted the EC's arguments against the use of the STX/Daedong bond data as a market benchmark. We agree with the EC that yen rates for corporate bonds cannot be compared directly with won rates for PSLs. Furthermore, in the absence of additional information from Korea regarding the value of the collateral underlying the STX/Daedong bonds (i.e., factory) and PSLs (i.e., Yangdo Dambo) respectively, we are unable to ensure that any comparison of the corporate bond and PSL rates is not affected by differences in the value of the relevant collateral. In addition, we note that Korea has reported three-year bond data, which cannot be readily compared with six-month PSL rates. For these reasons, we are unable to use the STX/Daedong corporate bond data submitted by Korea as a basis for establishing a market benchmark against which to judge the STX/Daedong PSLs.

Hyundai / Mipo

7.238.
Korea provided information regarding corporate bonds issued by Hyundai Mipo at para. 238 of its first written submission.
7.239.
The EC submits that Korea's data should not be used as a market benchmark because Korea failed to provide any supporting evidence in the form of exhibits, or otherwise. The EC also argues that, even if the Panel considers Korea’s proposed benchmark as relevant, the corporate bond rates proposed by Korea are generally higher than the rates of the KEXIM pre-shipment loans to Hyundai-MIPO.
7.240.
First, we note that Korea failed to provide any detailed information regarding the Mipo bond rates. For example, it failed to provide any of the more detailed information (concerning quarterly average rates) that it had provided for other shipyards in the form of Exhibits KOREA 18 – 21. In addition, we note that Korea failed to inform the Panel whether any of the Mipo bonds were guaranteed or collateralized. For these reasons, we do not consider it appropriate to establish a market benchmark on the basis of the corporate bond information submitted by Korea.

HHI

7.241.
Korea provided information regarding corporate bonds issued by HHI at para. 240 of its first written submission.
7.242.
The EC asserts that the HHI PSL rates are lower than the HHI corporate bond rates submitted by Korea.
7.243.
We note that Korea has reported three-year bond data, which it would be inappropriate to compare with six-month PSL rates (without further adjustment). For this reason, we are unable to establish a market benchmark on the basis of the HHI corporate bond data submitted by Korea.

Conclusion

7.244.
For the above reasons, we are unable to establish a market benchmark on the basis of the data submitted by Korea regarding corporate bonds issued by the shipyards at issue.

(ii) Other benchmarks proposed by Korea

7.245.
Korea submits that individual PSL transactions can be reviewed in light of usance letters of credit, overdraft loans, corporate bonds, commercial papers, facility loans and short-term borrowings taken / issued by the individual shipyards at issue.
7.246.
The EC asserts that the other sources of financing mentioned by Korea should not be used as a market benchmark, as the EC is unaware of the conditions and collateral with which such financing was provided.
7.247.
In our view, Korea has failed to provide enough information for us to use the abovementioned alternative instruments as a reliable market benchmark. With the exception of facility loans, the only information submitted by Korea regarding collateral requirements is that it "depend[s] on the financial strength of the borrower".141 For facility loans, Korea reports a need for collateral in the form of "real properties, factory as a whole",142 but provides no indication of the impact of such collateral on the facility loan spreads. Thus, there is not enough information for us to be sure that facility loans and KEXIM PSLs are comparable on the basis of collateralization (nor is there sufficient information for us to make any adjustments that might be required). The information provided in respect of the terms / maturity of these various instruments is similarly insufficient. The only concrete term information relates to usance letters of credit.143 For the other instruments, Korea merely reports what terms "normally" or approximately ("about") are applicable.144 There is therefore no means for us to ensure that these other instruments and KEXIM PSLs are directly comparable in terms of maturity (nor is there sufficient information for us to make any adjustments that may be required). In addition, because Korea has not reported the identity of the providers of the abovementioned instruments, we are unable to establish whether or not they represent market (as opposed to public body) rates. Furthermore, we note that Korea has reported data in respect of usance letters of credit issued in US dollars for Hanjin.145 We do not consider it appropriate to compare the terms of instruments issued in different currencies.

(iii) Conclusion

7.248.
For the above reasons, we are unable to establish a market benchmark on the basis of the abovementioned alternative instruments.

(b) The EC's proposed market benchmark

7.249.
The EC's claim is based on a comparison of the terms of individual PSLs with a market benchmark constructed primarily on the basis of an index of corporate bond prices.146 For shipyards with a credit rating agency rating of BBB or higher (i.e., investment grade), the EC generally compares the actual KEXIM PSL rates with a market benchmark constructed on the basis of the Korea Stock Dealers Association ("KSDA") general index of six-month corporate bond prices, with certain adjustments for differences in maturity and collateral. For each investment grade shipyard, the EC identifies a credit rating agency rating, and applies KSDA index prices for corporate bonds with the same rating. For non-investment grade shipyards, having agency ratings lower than BBB, the EC compares the actual PSL rates with rates constructed on the basis of KEXIM's own Interest Rate Guidelines (generally using the KEXIM spreads for the relevant KEXIM rating), again with certain adjustments for differences in maturity and collateral.
7.250.
Korea objects to the market benchmark proposed by the EC. Korea's objections concern alleged differences in the rating practices of KEXIM / private banks and corporate bond credit rating agencies, the use of indexed bond pricing, the choice of maturity of the KSDA index data, risk management under the PSL disbursement process, and the admissibility of certain EC arguments / data.

(i) Alleged differences in rating practices

7.251.
Korea submits that the EC's methodology for investment grade transactions is flawed, as the credit rating assigned by KEXIM to a particular shipbuilder cannot be systematically matched to the credit ratings assigned to the same shipbuilder by corporate bond rating agencies. Korea asserts that corporate bond ratings and KEXIM's credit ratings are based on different levels of underlying credit risk.147 In particular, Korea argues that the basic credit rating for an unsecured bond is the same as the credit rating for the issuing company, whereas the credit rating by KEXIM is based on a "facility rating", which includes not only the credit risk of the borrowing company, but also risk-reducing factors reflecting the structure of the facility concerned, including collateral. As a result of facility rating, Korea argues that the default rate for a bank credit rating (such as KEXIM's PSLs) is lower than that for the corresponding corporate bond rating. According to Korea, for the same corporate entity, the corporate bond rating must be higher than the bank credit rating in order to have equivalent risk. Korea submits that banks analyse three elements when assigning credit ratings: the probability of default, loss given (i.e., in the event of) default, and expected loss. Korea states, on the basis of a KEXIM document,148 that loss given default "is the critical element that distinguishes corporate bonds from bank loans, since corporate bond rating in Korea, which is generally targeting for unsecured senior bonds with long-term maturity, does not take into account [loss given default]". Korea asserts that there is therefore a significant difference in expected loss between bank loan rating and corporate bond rating, with the result that yields on corporate bonds should be higher than those of bank loans for the same entity.
7.252.
Korea also argues that there is a fundamental difference between the rating practices of banks and those of credit rating agencies. Korea submits that banks generally employ a "point in time" approach, under which the time period for validity of a risk assessment is generally one year from the date of assessment. Korea states that private banks therefore analyse the risk in the "current situation". Korea asserts that credit rating agencies, on the other hand, rate corporate bonds on a "through the cycle" approach, covering the whole period of maturity, and taking into account the projected condition of the bond issuer in the projected worse part of the economic or industry cycle within that period. Korea therefore submits that the rating by credit ratings would be worse than that of banks.
7.253.
The EC submits that the credit ratings by corporate bond credit rating agencies are comparable to credit ratings by KEXIM. With regard to DSME, the EC notes that most of the DSME bonds issued between 1997 and 1999 were guaranteed / collateralized. The EC argues that the rating of the bonds reflected the guarantee, just as KEXIM's ratings reflected the collateral for the PSL.
7.254.
The EC also asserts that credit exposure from investment grade and BB rated private placement loans are comparable to credit exposure from public debt with the same rating. The EC argues that, in case of discrepancy, the more pessimistic rating is usually the one with the most predictive power. The EC argues that there should at least be correlation between corporate bond ratings and KEXIM ratings for ratings better than or equal to BB. The EC asserts that a credit rating assesses exposure risk in terms of the repayment capacity of the obligor, taking into account all possible collateral. The EC therefore submits that a private loan (such as a PSL) and a bond having the same ratings will present the same obligor repayment capacity and the same credit exposure risk, and should therefore be remunerated with the same interest rate.
7.255.
Regarding rating practices, the EC asserts that Standard & Poor's definition of a rating does not refer to worst-case scenario projections. Instead, it refers to the creditworthiness of the obligor and guarantees. In addition, the EC states that KEXIM's rating must analyse more than the "current situation", otherwise it would not have continued issuing PSLs for DSME when it was bankrupt.
7.256.
Regarding the relevance of guarantees in determining credit risk exposure, Korea states that most of the bonds issued in Korea are non-collateralized, and that the KSDA bond rates quoted by the EC are also non-collateralized. Korea also notes that the EC stated that the DSME corporate bonds were collateralized, just like the KEXIM PSLs, and queries why the DSME corporate bonds could not have been used as a benchmark.
7.257.
With regard to the EC's argument that the credit exposure from investment grade and
BB-rated private loans and public debt is comparable, Korea notes that the source relied on by the EC actually refers to the "worse incidence or default rates" for public placements, and "better loss severities on private placements". Korea also notes that the source further provides that private placement "offer superior experience with respect to all of incidence, severity and economic loss" than publicly issued bonds. Korea asserts that the EC's source therefore confirms Korea's position regarding the lower risk exposure of private placements vis-à-vis public placements with the same credit rating.
7.258.
In reviewing Korea's argument, we are struck by KEXIM's statement that loss given default "is the critical element that distinguishes corporate bonds from bank loans, since corporate bond rating in Korea, which is generally targeting for unsecured senior bonds with long-term maturity, does not take into account [loss given default]".149 KEXIM states that loss given default depends in part on the conditions of the facility and protection measures, including "collateral, collateral margin requirements,... and other support measures including guarantees and insurances".150
7.259.
We also note Korea's reliance on a Moody's document which states:

Moody's typically rates bank loans anywhere from on a par with a given borrower's senior implied rating to three refined categories above senior implied. The rating differential between a firm's bank loans and other debt obligations is not a reflection of a greater or lesser probability of default. Rather it is a consideration of higher expected recovery values for bank loans over the same borrower's bonds as a result of loan structure and security."151

7.260.
To the extent that there are discrepancies between the rating practices of private banks and credit rating agencies, the above extracts from the KEXIM and Moody's documentation submitted by Korea indicates that such discrepancies are largely caused by differences in collateral / guarantee (in the sense that bank loans such as KEXIM PSLs are normally guaranteed / collateralized, whereas corporate bonds are not) and maturity (in the sense that corporate bonds generally have a longer maturity than private bank loans such as KEXIM PSLs).152153 We shall therefore examine to what extent differences in collateralization and maturity are addressed in the EC's proposed market benchmark.
7.261.
Although the KSDA bond index is based on unsecured bonds, whereas KEXIM PSLs are collateralized, the EC's proposed market benchmark includes the same adjustment for collateral / guarantees as made by KEXIM in respect of PSLs (i.e., [BCI: Omitted from public version] per cent reduction in credit risk spread).154 For the most part, therefore, we consider that any rating discrepancy caused by dif