|
|
|
espa�ol - fran�ais - portugu�s |
Search
|
|
|
|
espa�ol - fran�ais - portugu�s |
|
Search
|
UNITED
STATES � ANTI-DUMPING MEASURES ON
Report of the Panel (Continued)
118. As an initial matter, it should be noted that the notion of "disguised" dumping margins is inherently circular. Dumping margins are "disguised" by a single average methodology only if one assumes that the "correct" dumping margins are those calculated using multiple sub-period averages. If, by contrast, one assumes that the Anti-Dumping Agreement requires a comparison of single averages, then the results of the single average calculation are correct (not "disguised"), and the multiple-averaging methodology "inflates" the dumping margins. In short, whether a particular methodology "disguises" or "inflates" the dumping margins depends on one's assumption regarding the "correct" methodology.
119. In this regard, it should be noted that the provisions of Article 2.4.2 represent a carefully negotiated bargain among the WTO Members concerning the methodology that is to be used to calculate dumping margins. For example, prior to the negotiation of the current Anti-Dumping Agreement, a number of major users of anti-dumping laws (including the United States) regularly calculated dumping margins by comparing an average normal value to the export prices for individual transactions and then "zeroing" any negative margins found. This methodology was challenged under the Tokyo Round Anti-Dumping Code, and a number of WTO Members sought to prohibit this methodology in the Uruguay Round negotiations.59
120. The result of these efforts was a compromise. The average-to-transaction methodology was not entirely prohibited by the Article 2.4.2; however, it was restricted to exceptional situations in which export prices differ significantly among purchasers, regions or time periods.60 At the same time, the general rule was established that, for non-exceptional cases, dumping margins should be calculated based on the comparison of "an average" normal value to "an average" export price (unless the transaction-to-transaction methodology was used).
121. By arguing that the single-average methodology "disguises" dumping margins, the United States is essentially asking the Panel to undo the bargain reflected in Article 2.4.2. In other words, the United States is asking the Panel to ignore the language of Article 2.4.2 (which requires a comparison of single averages) and instead affirmatively determine that "zeroing" is permitted even in non-exceptional cases, when there is no claim that export prices differed significantly among purchasers, regions or time periods. The Panel plainly should not accept that invitation.
122. The notion of "disguised" dumping margins also ignores the nature of exchange rates. As noted in Korea's Oral Statement, exchange rates are not conditions of sale; they are tools for converting amounts from one currency to another.61 Moreover, they are necessarily volatile and imperfect tools.
123. In hindsight, one might say that the Korean won was over-valued (at roughly 900 won per dollar) throughout much of 1997, until it "crashed" in November. The won then overshot the proper exchange rate and remained under-valued (at a rate as high as 1960 won per dollar in December 1997), until it finally rebounded to a mid point of around 1400 won per dollar at the end of March 1998.
124. Throughout all of these exchange rate shifts, the prices for POSCO's Korean and US sales changed little.62 However, the results of a comparison of those prices using the shifting exchange rates changed dramatically. Prior to November 1997, when the won was apparently over-valued, the Korean prices seemed higher than the US prices. In November and December 1997, when the won was apparently under-valued, the Korean prices seemed lower than the US prices.
125. The US arguments regarding "disguised" dumping margins implicitly assume that the dumping margins calculated when the won was over-valued are more probative than the dumping margins calculated when the won was under-valued. The multiple-averaging methodology also reflects the same bias: It calculates one dumping margin for the period when the won was over-valued and a separate margin for the period when the won was under-valued - and then effectively ignores the latter through "zeroing."
126. A fair comparison would not allow such a bias. It would not base a finding of dumping solely on an analysis of a period in which the won was over-valued, just as it would not base a finding of no dumping based solely on an analysis of a period in which the won was under-valued. Instead, a fair analysis would have given both periods equal weight, to avoid the distortions inherent in an analysis that examined only a period in which the exchange rate was biased in one direction.
127. In short, the distortions caused by these exchange rate movements could have been minimized by the use of an averaging period that included both the period of over-valuation and the period of under-valuation. By contrast, the multiple averaging period effectively included only the period of over-valuation. It thus allowed the imperfections in the exchange rate movements to cause the distorted results that it claimed that it was trying to avoid.
128. The facts concerning POSCO's "local sales" were as follows: The orders for these sales were placed in dollars (not won). The invoices for these sales set forth both the original order price in dollars and an amount in won calculated by applying the exchange rate on the date of invoice to the dollar price. The payments for these sales were made in won, with the payment amount calculated by applying the exchange rate on the date of payment to the agreed-upon dollar price. When the exchange rates on the dates of invoice and payment were not the same, the won amount of the payment did not match the won amount on the invoice (and the resulting exchange gain or loss was recorded in POSCO's accounting records). All of these facts were verified by the DOC. Based on those facts, the DOC's verification report concluded that "local sales are dollar denominated."63
129. The issue in this case concerns the methodology used by the United States to include these "local" sales in the calculation of normal value. As described in Korea's First Submission, the United States decided to base its calculations on the won amounts shown on the invoices for these sales (which were calculated by applying the exchange rate on the date of the home-market invoice to the agreed-upon dollar price). These won amounts were then included in the calculation of an average normal value in won. The average normal value in won was then converted into dollars using the exchange rate on the date of the US sales.
130. In essence, then, the dollar prices for these sales were converted into won at one rate (the rate on the date of the home-market invoice) and then converted back into dollars at another rate (the rate on the date of the US sale). Not surprisingly, this methodology resulted in substantial distortions. Indeed, Korea's First Submission identified one instance where this methodology inflated normal value by more than 70 percent.64
131. The issue before the Panel is whether, given the undisputed facts, the US methodology and the justifications given for adopting it were consistent with the requirements of the Anti-Dumping Agreement.
132. As explained in Korea's First Submission, the methodology for converting currencies in anti-dumping investigations is governed by Article 2.4.1 of the Anti-Dumping Agreement. That provision is, however, by its terms applicable only "[w]hen the price comparison under this paragraph [i.e., Article 2.4] requires a conversion of currencies." The provisions of Article 2.4.1 do not apply, therefore, when a conversion of currencies is not "required."65
133. In the circumstances of the SSPC and SSSS cases, it was clear that the conversion of the dollar-denominated prices for the local sales was not required. The United States could have included those sales in its normal value calculations based on the dollar prices, without using the converted won amounts. And, if the United States had used the dollar-denominated prices, then there would have been no need for the second conversion from won back to dollars. Thus, neither leg of the double-conversion was "required." The US methodology cannot, therefore, be reconciled with the requirements of Article 2.4.1.
134. The double-conversion methodology employed by the United States suffered from a further problem: Because the United States used the exchange rate at the time of the home-market invoice to convert the dollar prices into won, while using the exchange on the date of the comparable US sales to convert the won amounts back into dollars, it penalized POSCO for any changes in exchange rates between those two dates. This effect was not merely theoretical. In fact, as demonstrated in Korea's First Submission, the use of the inconsistent exchange rates actually increased normal value by more than 70 percent for one product comparison.66
135. The changes in the exchange rate between the date of the home-market invoice and the date of the comparable US sales was not within POSCO's control. Consequently, the use of a methodology that penalized POSCO for these changes was patently unfair. The double-conversion methodology was, therefore, inconsistent with the "fair comparison" requirement of Article 2.4.
136. As mentioned, the DOC's determinations indicated that the double-conversion methodology was adopted for three reasons: (1) the payments for the sales were made in won, (2) the sales were recorded in POSCO's accounting ledgers in won, and (3) the exchange rates used to convert the dollar amounts into the won amounts reflected on the invoices did not correspond to the exchange rates normally used by the DOC.67 Of these, the United States regards the exchange rate differences as the "primary basis" for its decision.68
137. Significantly, none of these rationales explains why the double-conversion was "required." Thus, they do not provide a basis for finding that the double-conversion was consistent with Article 2.4.1 of the Anti-Dumping Agreement, which permits currency conversions only when "required."
138. Moreover, the rationales offered by the DOC are irrelevant to the critical issue in this case. In its preliminary oral responses to the Panel's questions, the United States has conceded that, if the sales in question had been denominated in dollars, its double-conversion methodology would have been inappropriate. Thus, the DOC's decision to apply the double-conversion methodology could only be justified by evidence establishing that the sales in question were not dollar-denominated. But the three justifications offered by the United States do not address that issue.
139. First, the fact that the payments were actually made in won does not contradict the fact that the basic terms of the sales were fixed in dollars, not won. As discussed above, the verified facts demonstrated that the purchase orders showed only a dollar price and that this agreed-upon dollar price was used to determine first the won amount set forth on the invoice (in addition to the dollar price) and later the different won amount of the customer's payment. It is also undisputed that the won amounts set forth on the invoice were not used to determine the won amount of the customer's payment.
140. In this regard, the DOC's reliance on the fact that the payments were actually made in won is somewhat ironic. The DOC did not use the amount in won that the customer actually paid as the price for these sales. Instead, the DOC used the amount in won shown on the invoice as the price for these sales - despite the fact that the won amounts shown on the invoice did not correspond to the customer's payments. In short, the DOC justified its use of one won amount as the price for these sales based on the fact that the customer paid a different won amount. That result is plainly illogical. The only rational conclusion to be drawn from the verified facts is that "local sales are dollar denominated," which is precisely what the DOC's own Verification Report stated.69
141. Second, the fact that local sales are recorded in won in POSCO's accounting ledgers is apropos of nothing. In accordance with generally accepted accounting principles, POSCO keeps its accounting ledgers in a single currency, the won. Thus, all of POSCO's sales - regardless of the currency in which they are denominated - are recorded in won. Indeed, POSCO's export sales to the United States - which the United States accepted as being denominated in dollars - are recorded in POSCO's accounting records in won.70
142. Third, the differences between official US and Korean exchange rates had no bearing on the question whether the sales prices were fixed in dollars. The fundamental point is that the won amounts shown on the invoices and the won amounts paid were calculated based on the agreed-upon dollar prices, and were not based on some agreed-upon won amount.
143. Finally, it should be noted that the reliance by the United States on the differences between US and Korean exchange rates was fundamentally unfair. In effect, the United States penalized POSCO by applying the distortive double-conversion methodology simply because the US Federal Reserve and the Korean Exchange Bank did not set the same exchange rates. The differences between the Federal Reserve and Korean Exchange Bank plainly were not within POSCO's control, and they could not be predicted by POSCO (since, among other things, the Federal Reserve rates each day were set long after the close of business in Korea). By penalizing POSCO for events beyond its control, the United States failed to make the "fair comparison" required by Article 2.4.
144. In the proceedings before this Panel, the United States has offered a number of alternate justifications for the double-conversion methodology. As discussed below, those additional justifications are without merit. Before turning to those justifications, however, it is worth noting a somewhat separate point: A close examination of the US arguments before this Panel reveals that the United States has, in fact, conceded that its double-conversion methodology was flawed.
145. As mentioned, the DOC claimed that the use of the Korean won amounts recorded in POSCO's accounting system was justified because the exchange rates used to calculate those amounts (which were based on official Korean Exchange Bank rates) differed significantly from the Federal Reserve exchange rates normally used by the DOC for its currency calculations. The exchange rate comparison presented by the DOC as evidence of this difference was, however, deeply flawed. The comparisons of POSCO's exchange rates to the Federal Reserve rates that were actually relied by the DOC showed differences of less than 1 percent in all cases. The only comparisons where the DOC identified a difference of more than 1 percent involved situations where the DOC mistakenly compared POSCO's exchange rates to modified exchange rates calculated by the DOC, and not to the actual Federal Reserve exchange rates.71
146. In its First Submission, the United States concedes that the DOC made this error. Footnote 161 of its Submission admits that "Korea is correct that the Department mistakenly used adjusted exchange rates in the SSPC case." It is clear, then, that the DOC's determination was based on a factual error and, as a result, it cannot be sustained.
147. The United States argues that the factual error in the SSPC case was harmless, because its post hoc comparison of exchange rates also identifies four dates in November for which there were differences of more than 1 percent between the Federal Reserve rate and the rates used by POSCO. But this post hoc argument does not cure the defect in the DOC's determination. The DOC based its argument on the fact that the exchange rates actually used by POSCO to convert specific local sales differed from the Federal Reserve rates. By contrast, the United States has not pointed to any evidence that the exchange rates on the four dates it has identified were actually used by POSCO to convert local sales prices. For example, if there were no local sales on those dates, the comparisons presented by the United States in its brief would have no bearing on the manner in which the sales under investigation were actually recorded in POSCO's accounting records.
148. Finally, it should be noted that the comparisons offered by the United States in its First Submission are flawed, because they ignore the significant time differences between the United States and Korea. The Federal Reserve rates are based on a survey of New York banks at 12:00 noon on each date. But 12:00 noon in New York on any given day is 2:00 in the morning the next day in Korea. Or, to put it the other way, 12:00 noon in Korea on any date is 10:00 at night the previous day in New York. Thus, a comparison that matches exchange rates on the "same" date actually compares exchange rates determined 14 hours apart.72
149. When exchange rates are shifting rapidly, as they were in November 1997, this 14-hour difference can be critical. Indeed, if one compares the exchange rates at the same time (rather than the same date), the differences identified in the US brief simply disappear. The following table illustrates this point:
150. In its First Submission, the United States also conceded that the first step in the double conversion - that is, the initial conversion from dollars to won - was made using a methodology that did not comply with the requirements of the Anti-Dumping Agreement. Thus, it contended that:
Of course, the United States blames POSCO for failing to follow the requirements of Article 2.4.1 in the initial conversion from dollars to won. But POSCO was making that conversion solely for its internal accounting purposes, and not for purposes of calculating dumping margins under the Anti-Dumping Agreement. It was the United States that adopted POSCO's internal conversion from dollars to won as the starting point for the dumping margin calculations.
151. In any event, whatever flaws there may have been in POSCO's internal conversion cannot possibly justify the US decision to reject the actual dollar prices for the local sales - because the actual dollar prices were not affected by POSCO's internal conversion. In this regard, it must be remembered that the only conversions made by POSCO converted the dollar-prices shown on the purchase orders for the local sales into won. POSCO never converted from won into dollars. Indeed, the fact that POSCO's conversions were made from dollars into won was explicitly conceded by the DOC's verification report.74
152. Thus, when the United States complains about the flaws in POSCO's currency conversions, it is, in fact, complaining about the conversions that were used to determine the won amounts shown on the invoices. The won amounts shown on the invoices were, of course, the amounts the United States used as the starting point for its calculation of normal value. Consequently, the United States has conceded that the won amounts it used for the starting point of its calculations were determined in a manner that is inconsistent with the requirements of Article 2.4.1 of the Anti-Dumping Agreement. The violation of the Anti-Dumping Agreement is, therefore, manifest.
153. The United States has offered four basic responses to the arguments presented by Korea on the double-conversion methodology: First, it asserts that these arguments raise factual issues on which the Panel should defer to the DOC's determinations. Second, it claims that it complied with the requirements of Article 2.4.1 by making the currency conversions for the local sales using the exchange rates on the "date of sale."75 Third, it contends that, while Article 2.4.1 "presupposes" that a conversion will be required, it does not prevent investigating authorities from making conversions when no conversion is required.76 And, fourth, it suggests that there was no double-conversion by the DOC, because the first conversion from dollars into won was made by POSCO for internal accounting purposes and merely adopted by the DOC. As discussed below, none of these arguments provides a basis for upholding the double-conversion methodology.
154. The United States has attempted to shield its decision to double-convert the dollar amounts for the local sales by claiming that its decision is entitled to special deference because it is a "factual determination."77 That argument is, however, misplaced.
155. The facts concerning the "local sales" are not in dispute. The undisputed facts demonstrate that:
One might argue about the effect these undisputed facts should have on the dumping calculations. But the facts themselves are not, and never have been, in dispute.83
156. The US decision to apply the double-conversion methodology was, of course, based on an assessment of the facts of the case. It was not, however, based on a factual findings that differed from those advanced by POSCO. For example, the DOC did not dispute that the orders for the local sales were denominated solely in dollars (and not in won). It did not dispute that the won amounts on the invoices for the local sales did not correspond to the won amounts actually paid. And it did not dispute that the won amounts in all cases were calculated by applying exchange rates fixed by the official Korean Exchange Bank to the dollar-denominated prices for the local sales. Indeed, it would have been plain error for the DOC to dispute these facts, because there was absolutely no evidence on the record refuting them.
157. Given these undisputed facts, the question before the DOC was whether, in its calculation of normal value, it was appropriate to use the actual dollar prices for these sales, or to use won amounts that (the undisputed facts showed) were calculated by multiplying the agreed-upon dollar prices by the official exchange rates for the date of invoice. The DOC made the decision that the appropriate methodology was to use the converted won amounts. According to the DOC's determinations, this decision was based on three considerations: (1) the payments for the sales were made in won, (2) the sales are recorded in POSCO's accounting ledgers in won, and (3) the exchange rates used to convert the dollar amounts into the won amounts reflected on the invoices did not correspond to the exchange rates normally used by the DOC.84
158. The question before the Panel is whether the three grounds identified by the DOC are sufficient to justify the methodology the DOC adopted. This is not a factual issue, and it does not require the Panel to resolve any factual questions.
159. As mentioned, the United States has repeatedly claimed that it complied with the requirements of Article 2.4.1 by making the currency conversions for the local sales using the exchange rates on the "date of sale."85 This argument is, however, utterly irrelevant.
160. The problem in this case is not that the United States selected the wrong exchange rate. Rather, the problem is that the United States used a double conversion, applying different exchange rates to convert the same sale in an inconsistent manner. The inconsistency arose because the United States made the final conversion from won to dollar using the exchange rate on the date of the US sale, while the initial conversion from dollars to won had been made using the exchange rate on the date of the home-market invoice. Because the exchange rate on the date of the US sale could (and did) differ dramatically from the exchange rate on the date of the home-market invoice, the double-conversion created significant distortions.
161. As mentioned, the United States also claims that, while Article 2.4.1 "presupposes" that a conversion will be required, it does not prevent investigating authorities from making conversions when no conversion is required.86 This interpretation is without merit.
162. Article 2.4.1 is the only provision in the entire Anti-Dumping Agreement that addresses currency conversion issues. If that provision does not apply to a particular currency conversion, then there are no rules in the Agreement governing the conversion.
163. Article 2.4.1 by its terms does not apply when the conversion is not "required." Consequently, if the Anti-Dumping Agreement does permit conversions when a conversion is not required (as the US contends), then those non-required conversions would not be subject to the disciplines of Article 2.4.1. Moreover, since there are no other provisions of the Agreement addressing currency conversions, these non-required conversions would not be subject to any disciplines at all. Thus, the US argument leads inevitably to the conclusion that (while required conversions are closely regulated by the Agreement) the Agreement grants the investigating authorities unfettered discretion to make non-required conversions and to use any conversion methodology at all when they do so. Such a result is plainly absurd.
164. In short, it is apparent that the Anti-Dumping Agreement permits currency conversions only when such conversions are "required."
165. The United States cannot meet that test in this case. As explained in Korea's First Submission, POSCO had submitted the dollar-denominated prices for the local sales to the DOC. The DOC was able to verify that the reported dollar prices matched the dollar amounts on the order sheets and invoices, and it was also able to verify that the customers' actual payments in won were calculated by multiplying the dollar prices by the official exchange rate for the date of payment. Accordingly, there was no reason the DOC could not have used the dollar-denominated prices. Consequently, the use of the converted won amounts was not "required" - and thus was not permitted under Article 2.4.1 of the Anti-Dumping Agreement.
166. The United States has also suggested that Article 2.4.1 does not apply to its double-conversion methodology, because the initial conversion from dollars to won was actually made in the first instance by POSCO for internal accounting purposes.87 It appears that the United States believes that, because it did not itself actively convert the dollar prices into won, it is not "guilty" of a double-conversion. That position is, however, untenable.
167. In the SSPC and SSSS investigations, the United States chose to use the converted won amounts (which were calculated figures that did not represent the fixed prices for the sales) rather than the actual dollar amounts (which were the fixed prices for the sales). The United States then chose to convert those won amounts back into dollars using the exchange rate on the date of the US sale - which was not the same as the exchange rate that had been used previously to convert the dollar amounts into won. The overall methodology therefore involved a double-conversion of the dollar amounts using inconsistent exchange rates. This double-conversion methodology is improper - regardless of whether the initial conversion was "made" by the United States or simply adopted by the United States from the figures in POSCO's accounting records.
168. As demonstrated in Korea's First Submission, the "double-conversion" methodology distorted the results of the DOC's comparisons - inflating the dumping margins for one comparison by more than 70 percent. This distortion occurred because the exchange rates on the dates of the home-market invoices (which were used to convert the dollar prices into won) could differ significantly from the exchange rates on the dates of the US sales (which was used to convert the won amounts back into dollars).
169. The inflation of the dumping margins in this manner was plainly unfair. It penalized POSCO for changes in the exchange rates that were beyond its control.88 Consequently, as Korea explained in its First Submission, the double-conversion methodology violated the "fair comparison" requirements of Article 2.4.
170. The United States has not even attempted to respond to these points, because there is simply no justification for the unfair results caused by its double-conversion methodology. Its silence on this point is striking confirmation of Korea's claims.
59 Terence P. Stewart, ed., II The GATT Uruguay Round: A
Negotiating History, 1986-1992, at 1537-43 (1993) (ROK Ex. 77).
60 In such exceptional circumstances, Article 2.4.2 permits
investigating authorities to calculate dumping margins by comparing an average
normal value to the prices for individual export transactions.
61 See Cotton Yarn, para. 494 (an exchange rate "is a
mere instrument for translating into a common currency prices that have
previously been rendered comparable in accordance with the second sentence of
[Tokyo Round Code] Article 2:6," the predecessor of Article 2.4 of the
Anti-Dumping Agreement).
62 See POSCO Rebuttal Brief (SSPC), at 20-21 (ROK Ex. 9);
POSCO Case Brief (SSSS), at 21-22 (ROK Ex. 20).
63 SSSS Sales Verification Report, at 14 (ROK Ex. 19).
64 See Korea�s First Submission, para. 3.58.
65 See Korea�s First Submission, paras. 4.66 - 4.69.
66 See Korea�s First Submission, para. 3.58.
67 SSPC Final Determination, at 15456 (ROK Ex. 11); SSSS Final
Determination, at 30678 ("the local sales were paid in won and recorded in
POSCO�s accounting records in won, and the exchange rates used by POSCO were
dissimilar from those used by the Department") (ROK Ex. 24).
68 US First Submission, para. 173.
69 SSSS Sales Verification Report, at 14 (ROK Ex. 19).
70 See Korea�s First Submission, paras. 3.49 - 3.54.
71 See Korea�s First Submission, para. 3.60.
72 As noted in Korea�s First Submission, the 14-hour time
difference between Korea and New York means that the Federal Reserve rates
determined at 12:00 noon in New York are not announced until nine hours after
the close of business (5:00 p.m.) in Korea.
73 US First Submission, para. 178.
74 See SSSS Sales Verification Report, at 14 (ROK Ex.
19).
75 See, e.g., US First Submission, para. 175.
76 See US First Submission, para. 177.
77 See US First Submission, para. 180.
78 Provided as ROK Exhibit 78 is a sample order sheet from a
verified local sale. In the order sheet, one can see clearly that the
transaction is denominated in dollars because there is a "D" for dollars in the
currency box (Number 10). Also provided is a guide to reading the order sheet
from the DOC Verification Report.
79 Provided as ROK Exhibit 79 is a sample invoice from a
verified local sale.
80 Provided as ROK Exhibit 80 is a sample shipping list from a
verified local sale.
81 SSSS Sales Verification Report, at 14 (ROK Ex. 19).
82 Id.
83 The United States has belatedly made in its Oral Statement
the argument that POSCO did not provide sufficient, timely evidence to the DOC
demonstrating that the economic value of the "local sales" was fixed in US
dollars and not in won. US Oral Statement, paras. 35-40. For the reasons
provided in Korea�s response to the Panel�s Question No. 3 regarding
double-conversion, Korea believes that this US argument is untimely and
meritless.
84 SSPC Final Determination, at 15456 (ROK Ex. 11); SSSS Final
Determination, at 30678 ("the local sales were paid in won and recorded in
POSCO�s accounting records in won, and the exchange rates used by POSCO were
dissimilar from those used by the Department") (ROK Ex. 24).
85 See, e.g., US First Submission, para. 175.
86 See US First Submission, para. 177.
87 See US First Submission, para. 183.
88 Moreover, because the adoption of the double-conversion
methodology was predicated on the differences between the official Federal
Reserve and Korean Exchange Bank rates, POSCO was unfairly penalized for the
existence of those differences and for the fact that it used the official Korean
Exchange Bank rates in the normal course of business.
|
|
