Vinh Hoan Corp. v. United States

234 F. Supp. 3d 1332, 2017 CIT 81, 2017 Ct. Intl. Trade LEXIS 82

• Court: United States Court of International Trade
• Decided: July 10, 2017
• Docket: Consol. 13-00156
• Status: Published

Opinion

OPINION AND ORDER

Kelly, Judge

Before the court for review is the U.S. Department of Commerce’s (“Department” or “Commerce”) final results of its second redetermination filed pursuant to the court’s' decision and remand order in Vinh Hoan Corporation v. United States, 40 CIT -, 179 F.Supp.3d 1208 (2016) (“Vinh Hoan II”). See Final Results of Redeterraination Pursuant to Vinh Hoan Corporation et al. v. United States, Consol. Court No. 13-00156, Slip Op. 16-53 (May 26, 2016), Jan. 27, 2017, ECF No. 203-1 (“Second Remand Results”); see also Vinh Hoan II, 40 CIT -, 179 F.Supp.3d 1208.

In Vinh Hoan II, the court remanded four aspects of Commerce’s first reconsideration after remand of its final determination in this eighth antidumping duty (“ADD”) administrative review of certain frozen fish fillets from the Socialist Republic of Vietnam (“Vietnam”). Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1237; see also Final Results of Redetermination Pursuant to Vinh Hoan Corporation et al. v. United States, 44-46, 79-82, Aug. 3, 2015, ECF No. 132 (“First Remand Results”). Specifically, the court ordered Commerce to explain or reconsider its: 1) selection of surrogate value data to value respondents’ sawdust factor of production; 2) selection of surrogate value data to value respondents’ rice husk factor of production; 3) decision to construct a value for respondent Vinh Hoan Corporation’s (“Vinh Hoan”) fish oil byproduct rather than selecting the best surrogate value data for fish oil placed on the record; and 4) methodology for calculating byproduct offsets. See Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1237.

For the reasons that follow, Commerce’s determinations in the Second Remand Results to value sawdust and rice husk using Indonesian import data comply with the court’s remand order, are supported by substantial evidence, and are sustained. Commerce’s byproduct offset calculation, as applied in the Second Remand Results, is also sustained. However, Commerce’s determination to value fish oil using a constructed value continues to be unsupported by substantial evidence. The court accordingly remands Commerce’s decision to construct a value for Vinh Hoan’s fish oil byproduct rather than select the best available existing surrogate value data source for fish oil for further explanation consistent with this opinion.

BACKGROUND

The court presumes familiarity with the facts of this case as discussed in Vinh Hoan II and in Vinh Hoan Corporation v. United States, 39 CIT -, 49 F.Supp.3d 1285 (2015) (“Vinh Hoan I”), and here summarizes the facts relevant to the present discussion of the Second Remand Results.

In the final determination of this eighth administrative review of the ADD order, Commerce selected Indonesia as the primary surrogate country. See Certain Frozen Fish Fillets From [Vietnam], 78 Fed. Reg. 17,350 (Dep’t Commerce Mar. 21, 2013) (final results of ADD administrative review and new shipper reviews; 2010-2011), as amended, 78 Fed. Reg. 29,323 (Dep’t Commerce May 20, 2013) (“Final Results”), and accompanying Issues and Decision Memorandum for the Final Results of the Eighth Administrative Review and Aligned New Shipper Reviews for Certain Frozen Fish Fillets from [Vietnam] 27, June 19, 2013, ECF No. 27-3 (“Final Decision Memo”). Commerce selected Indonesian import data for HTS 4401.30 (“Sawdust and Wood Waste and Scrap”) to value respondents’ sawdust factors of production (“FOP”), determining that data to be the best available information because it satisfied all of the surrogate value (“SV”) data selection criteria. See Final Decision Memo at 32. Commerce likewise determined that Indonesian import data for HTS 1213.00 (“Cereal Straw and Husks, Unprepared, Whether or Not Chopped, Ground, Pressed or in the Form of Pellets”) is the best available information for valuing rice husk because that data satisfied all of the SV data selection criteria as well. See id. at 33. Commerce valued the fish oil byproduct offset using Indonesian import data for HTS 1504.20.9000, but capped the fish oil SV at a value constructed for unrefined fish oil, id. at 38, “starting with fish waste and adding the [FOPs] used by Vinh Hoan to produce fish oil, and included surrogate financial ratios to ensure the value is on an as-sold basis.” [Eighth] Administrative Review, and Aligned [Ninth] New Shipper Reviews, of Certain Frozen Fish Fillets from [Vietnam]: Surrogate Values for the Final Results, A-552-801, at 6, PD 436, bar code 3124119-01 (Mar. 13, 2013) (“Final Surrogate Value Memo”). 1

In Vinh Hoan I, the court remanded Commerce’s primary surrogate country selection of Indonesia, determining that the selection was contrary to law and not supported by substantial evidence. Vinh Hoan 1, 39 CIT at -, 49 F.Supp.3d at 1296-1321. The court accordingly reserved judgment on all challenges to Commerce’s SV data source selection to value respondents’ FOPs, including the selection of Indonesian import data for sawdust and rice husk, awaiting further explanation of Commerce’s selection of Indonesia as the primary surrogate country. See id., 39 CIT at -, 49 F.Supp.3d at 1321. The court also sought further explanation of Commerce’s choice of FOP SVs. Id., 39 CIT at -, 49 F.Supp.3d at 1309-11. The court granted Defendant’s request for a remand for Commerce to reconsider its calculation of respondent’s fish oil byproduct offset. 2 Id., 39 CIT at -, 49 F.Supp.3d at 1321-22.

On first remand, Commerce again selected Indonesia as the primary surrogate country and determined to use Indonesian import data under HTS 4401.30 to value respondents’ sawdust FOP, First Remand Results 62-65, and Indonesian import data under HTS 1213.00 to value respondents’ rice husk FOP. Id. at 57-61. Regarding the offset for the fish oil byproduct, Commerce continued to “cap” Indonesian import data for HTS 1504.20.9000 at a value representative of Vinh Hoan’s fish oil, derived from a build-up of FOPs used to produce unrefined fish oil, because the import data value was greater than the value for whole fish, the main input. Id. at 79. Commerce determined that “[i]t would be illogical to value an unrefined by-product like fish oil at a value greater than that of the main input, a value that also approaches that of the finished product, frozen fish fillets.” IOd. at 80. Commerce also made certain adjustments to its fish oil calculation. See id. at 44-46.

After remand, Vinh Hoan continued to challenge Commerce’s SV selections for valuing FOPs in light of the further explanation provided by Commerce. First Remand Results 40-45. Relevant here, Vinh Hoan questioned the specificity of the Indonesian import data under HTS 4401.30 used to value sawdust by presenting evidence that the import data includes “more complex and value-added products, such as ‘cat litter product’ and ‘wood fire starters.’ ” Pl.’s Comments Final Results of Re-determination Pursuant to Remand 4-5, Oct. 23, 2015, ECF No. 149 (“Vinh Hoan Comments on First Remand”). Vinh Hoan also challenged the selection of Indonesian import data to value rice husk on the grounds that the data was non-specific and aberrational because the import data covered by HTS 1213.00 is a basket category that covers too many items to be specific and representative of the value of rice husk. Id. at 2-4. Vinh Hoan doubted whether large volumes of imports under this category could reasonably be specific to rice husk in light of evidence that: (1) Indonesia is a significant producer of rice with seemingly little need to import this byproduct; (2) large volumes of Indonesian imports came from countries that are not significant rice producers; and (3) the rice husk SV exceeded the value of rice. See id. Vinh Hoan further argued that the SV assigned by Commerce was too high in relation to the SV assigned to subject merchandise to accurately reflect the role played by rice husk as a fuel source in producing Vinh Hoan’s fish oil byproduct. See id. at 30-31. Vinh Hoan also argued that Commerce unreasonably “capped” the SV selected for fish oil, rendering the valuation of fish oil unsupported by substantial evidence and contrary to law. 3 Id. at 6-15.

In Vinh Hoan II, the court remanded Commerce’s SV data selection for sawdust and rice husk for further explanation. Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1227-29. The court determined that Commerce’s explanation that HTS heading 4401.30 specifically names sawdust “fails to address record evidence that the import data covered other materials and possibly higher value-added materials,” id. at 1228, and remanded for Commerce to address the detracting evidence. See id. at 1229. The court also determined that Commerce had not addressed evidence that detracted from its finding that Indonesian import data for rice husk is specific and non-aberrational and, accordingly, that Commerce’s selection of Indonesian import data to value rice husk was not supported by substantial evidence. Id. at 1228-29. The court also remanded Commerce’s decision to construct a value for fish oil because Commerce failed to explain why it is reasonable to depart from its normal methodology of choosing the best existing SV data source to value respondent’s fish oil byproduct by using a constructed value in lieu of SV data derived from an existing data source. See id. at 1224. The court stated that, until Commerce acknowledged that its methodology was in fact a constructed value rather than a SV derived from an existing data source, the court could “not review whether Commerce’s choice of Indonesian import data is reasonable when it is unclear how, to what extent, or even if Commerce used Indonesian import data for fish oil in calculating a SV for Vinh Hoan’s fish oil.” See id. As a separate matter, the court also held that Commerce failed to provide a reasonable explanation for its byproduct offset calculation methodology. See id. at 1225.

In the Second Remand Results, Commerce made the following determinations. First, Commerce reconsidered its SV selection for sawdust. After reopening the record to obtain additional Indonesian historical import data for HTS 4401.30, covering “Sawdust and Wood Waste and Scrap,” reported on a monthly basis, Commerce continued to find Indonesian import data for HTS 4401.30 to be the best available information to value sawdust because the data is specific and non-aberrational. See Second Remand Results 11-14. Commerce also reconsidered its SV selection for rice husk. See id. at 14-23. Reopening the record to obtain Indonesian historical import data for HTS 1213.00, covering “Cereal Straw and Husks, Unprepared, Whether or Not Chopped, Ground, Pressed or in the Form of Pellets,” Commerce found that the Indonesian import data for HTS 1213.00 used in the final results to value rice husk is aberrational. Id. at 19-21; see Final Decision Memo at 33-34. Although Commerce determined that both Indonesian data (from the Central Bureau of Statistics (“ICBS”)) and Indian import data on the record satisfy its SV data selection criteria, Commerce selected the ICBS data over Indian import data to value respondent’s rice husk FOP because the ICBS data is from the primary surrogate country. Id. at 21-23.

Additionally, on second remand, Commerce continued to construct a value for fish oil using import data for Indonesian Global Trade Atlas category HTS 1504.20.9000 (“Fish Fats & Oils & Their Fractions Exc Liver, Refined or Not, Not Chemically Mod”), which covers both refined and unrefined fish oil, capping that data at a value derived from a build-up of FOPs used to produce unrefined fish oil. 4 Second Remand Results 23. Commerce “capped” the import data at the value for unrefined fish oil because Commerce considered it unreasonable to use as a SV Indonesian import data for fish oil with a higher value than data for whole fish, the main input. 5 Id. at 23-24. Finally, Commerce determined that it need not reconsider its methodology for calculating a byproduct offset, which deducted the absolute value of a byproduct with a negative SV (that is, a byproduct sold at a loss) from normal value, because there were no longer any byproduct values greater than the revenues for the byproducts. See id. at 25-26. 6

JURISDICTION AND STANDARD OF REVIEW

The court continues to have jurisdiction pursuant to section 516A(a)(2)(B)(iii) of the Tariff Act of 1930, as amended, 19 U.S.C. § 1516a(a)(2)(B)(iii) (2012), 7 and 28 U.S.C. § 1581(c) (2012), which grant the court authority to review actions contesting the final determination in an administrative review of an anti-dumping duty order. “The court shall hold unlawful any determination, finding, or conclusion found ... to be unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i). “The results of a redetermination pursuant to court remand are also reviewed ‘for compliance with the court’s remand order.’ ” Xinjiamei Furniture (Zhangzhou) Co. v. United States, 38 CIT -, -, 968 F.Supp.2d 1255, 1259 (2014) (quoting Nakornthai Strip Mill Public Co. v. United States, 32 CIT 1272, 1274, 587 F.Supp.2d 1303, 1306 (2008)).

DISCUSSION

I. Sawdust Surrogate Value

In Vinh Hoan II, the court held that Commerce cannot rely exclusively upon the fact that the word “sawdust” appears in the heading to conclude that the data source is specific and non-aberrational. Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1227-28. The court further held that Commerce’s selection of Indonesian import data under HTS 4401.30, covering “Sawdust and Wood Waste and Scrap,” is unsupported by substantial evidence because Commerce failed to respond to arguments and record evidence demonstrating that the HTS category includes higher value-added products. Id. The court specifically drew attention to Commerce’s failure to explain why it could discount the possibility that the data contained significant volumes of non-specific, higher value-added merchandise without any analysis. Id., 40 CIT at -, 179 F.Supp.3d at 1228 n.18. The court remanded for Commerce to explain its determination that Indonesian import data is specific and non-aberrational, given record evidence that higher value-added products are included within the category and the significant range of values contained in the import data. Id., 40 CIT at -, 179 F.Supp.3d at 1227-28.

In the Second Remand Results, Commerce determined that the record lacked information to examine whether the sawdust SV generated in the First Remand Results is aberrational. Second Remand Results 5-6. Accordingly Commerce reopened the record to permit Vinh Hoan to place additional Indonesian historical data for HTS 4401.30 for sawdust on the record to allow Commerce to examine whether the sawdust SV is aberrational. Id. at 5. Commerce examined the data and continued to find that. import data for HTS 4401.30 is specific to Vinh Hoan’s sawdust, M. at 10-11, and non-aberrational. Id. at 11-14. Commerce explained its conclusions that the import data is non-aberrational despite the wide range of average unit values (“AUV”) and that the category is specific to sawdust used by Vinh Hoan despite the fact that it includes higher value-added product. Id. at 10-14. Therefore, Commerce’s SV data selection complies with the court’s remand order.

In NME cases, Commerce obtains a normal value by adding the value of the FOPs used to produce the subject merchandise and “an amount for general expenses and profit plus the cost of containers, coverings, and other expenses.” 19 U.S.C. § 1677b(c)(l). Commerce values the FOPs “based on the best available information regarding the values of such factors in a market economy country or countries.” Id. Commerce’s methodology for selecting the best available information evaluates data sources based upon their: (1) specificity to the input; (2) tax and import duty exclusivity; (3) contemporaneity with the period of review; (4) representativeness of a broad market average; and (5) public availability. Import Admin., U.S. Dep’t Commerce, Non-Market Economy Surrogate Country Selection Process, Policy Bulletin 04.1 (2004), available at http://enforcement.trade.gov/pohcy/bull04-l.html (last visited July 5, 2017) (“Policy Bulletin 04.1”). Commerce’s practice for selecting the best'available information to value individual FOPs favors selecting a data source that satisfies the breadth of its selection criteria where possible. See Final Decision Memo at 11.

On second remand, Commerce supported its determination that the import data is specific despite the fact that it may contain imports of value-added products by refei’encing Vinh Hoan’s questionnaire response that it consumes “pressed sawdust,” which Commerce concluded has been processed into a value-added product that is closer in form to HTS 4401.30 than to Bangladeshi price quotes for “unprocessed sawdust.” See Second Remand Results 10. To address Vinh Hoan’s claim that Indonesian import values are affected by especially small but expensive shipments from Singapore, the United States, and Germany, Commerce removed import data from those countries for quantities of less than 100 kilograms (“kg”) and found no difference in historic AUVs for HTS 4401.30. 8 Id. at 6-7. Commerce also explained its determination that the import data is specific, despite Vinh Hoan’s claim that the data must have included nonspecific merchandise because a SV derived from this data is too high relative to other fuel sources Vinh Hoan reported using, by noting the absence of record information relating various fuel sources to the energy they provide. 9 Id at 9. Lastly, Commerce discounted the notion that the import data for HTS 4401.30 could not be specific to sawdust because export data indicates that exports under this HTS are cheaper than imports. See id. at 9-10. Commerce determined that this disparity does not affect its determination that the import data is specific because the category is a basket category that may include higher priced value-added products. See id. Commerce reasoned that its selection of import data under this HTS is supported by the fact that imports are higher priced because Vinh Hoan used pressed sawdust, a higher value-added product. Id.

Commerce justified its determination that import data for HTS 4401.30 for the period of review (“POR”) is not aberrational by comparing the import data to historical import data. See Second Remand Results 12-13. Although Commerce acknowledged that the SV generated for the POR is the highest AUV in recent years, Commerce found that the AUV for the POR is only eight percent higher than the previous year and only three times higher than the lowest AUV in 2009. Id. at 13. Therefore, Commerce , found that the import data for HTS 4401.30 is not so much higher than the historical values on the record that it could be aberrationally high. Id. Commerce’s reconsideration complies with the court’s order to explain why the data is specific and non-aberrational in light of the detracting evidence highlighted by Vinh Hoan.

No party continues to challenge Commerce’s SV data selection for sawdust. Commerce has further explained why the import data, which includes higher value-added products, is specific to the sawdust Vinh Hoan reported using in production. Further, Commerce’s explained its conclusion that the import data is not aberrational by comparing the AUV for this POR to historical AUVS and concluding that the AUVs from the POR do not different significantly, Therefore, Commerce has complied.with the court’s order.

Il Rice Husk Surrogate Value

. In .Vinh. Hoan II the court remanded Commerce’s selection of Indonesian import data, .for HTS 1213.00 to value rice husk. See Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1228-29. The. court determined that Commerce’s conclusion that the data is. specific and non-aberrational was not supported by substantial evidence in light of evidence that the SV derived from this data was too high relative to the role rice husk played as a fuel source in Vinh Hoan’s production of fish oil and given the substantial increase in the Indonesian value from the preliminary to the final- determination. See id. at 1229. The court ordered Commerce on remand to either explain its SV data selection in light of this detracting evidence or reconsider its determination. Id.

On remand, Commerce placed Indonesian historical data for HTS 1213.00 on the record. See Second Remand Results 15. Commerce continued to find that the import data is specific to the input because cereal husks are among the items covered by the plain terms of the heading. Id. at 18. However, after comparing Indonesian import data to import data on the record from other countries on the surrogate country list, Commerce determined that the Indonesian import data for HTS 1213.00 is aberrational. Id. at 20. Therefore, Commerce declined to use Indonesian import data for HTS 1213.00 to value Vinh Hoan’s rice husk FOP. See id. at 21. Instead, Commerce selected Indonesian ICBS data as the best available information because it the data is representative of a broad market average, publicly available, tax and data exclusive, contemporaneous, reliable, and from the primary surrogate country. Id. at 21-23.

As already discussed, Commerce uses the “best available information” to value FOPs with a SV from a market economy country, 19 U.S.C. § 1677b(c)(1), seeking a SV that is specific to the input, tax and import duty exclusive, contemporaneous with the period of review, representative of a broad market average, and publically available. Policy Bulletin 04.Í. Here, Commerce supported its conclusion that the Indonesian import data for HTS 1213.00 is aberrational by noting that the Indonesian data, when compared to rice husk benchmark data for other countries on the surrogate country list, is so much higher that Commerce considered the data aberration-ally high. Second Remand Results 20. Specifically, Commerce found that the Indonesian value is “over five times higher than the Philippine value, and over 150 times higher than the Indian value for HTS 1213.00[], and the ICBS data.” 10 Id. Therefore, Commerce declined to use either the Philippine data or the Indonesian import data for HTS 1213.00 in its Second Remand Results, as it had in its Final Results. See id. at 21. Instead, on remand, Commerce determined that Indonesian ICBS data is the best available information to value rice husk because it satisfies the breadth of Commerce’s SV data selection criteria. See id. Although Commerce found that Indian import data for HTS 1213.00 equally meets its SV data selection criteria, Commerce preferred the ICBS data because it is from the primary surrogate country. See id. at 22. Commerce explained that it prefers to rely on factor costs from a single surrogate country because doing so “better reflects the tradeoff between labor costs and other factors’ costs, including capital, based on them relative prices.” Id Commerce also determined that the two Bangladeshi price quotes on the record are not the best available information for valuing rice husk because Commerce could not determine whether the price quotes are reliable. 11 Id. at 14-15.

No party continues to challenge Commerce’s SV data selection to value rice husk. Commerce has complied with the court’s remand order regarding rice husk.

III. Byproduct Offset Calculation

The court held that Commerce failed to provide a reasonable explanation for why its byproduct offset calculation methodology is appropriate even where it has the effect of reducing normal value, which appears to be at odds with the logic behind a byproduct offset. See Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1224-25. The court remanded Commerce’s determination for further explanation of why granting an offset that reduces normal value is reasonable or reconsideration of its byproduct offset calculation, which subtracts the absolute value of the byproduct value from normal value. See id.

On second remand, Commerce determined that the propriety of its byproduct offset calculation, which deducted the absolute Value of a byproduct with a negative SV (i.e., sold at a loss) from normal value in the Final Results, is moot because the changes in SV on remand result in no negative byproduct values. See Second Remand Results 26. No party questions Commerce’s conclusion, and the issue of the propriety of Commerce’s prior calculation that had the effect of increasing normal value is no longer a live case or controversy. Commerce’s has complied with the court’s order regarding the byproduct offset calculation.

IV. Fish Oil Constructed Value

Vinh Hoan II remanded for further explanation or reconsideration Commerce’s decision to use a “cap” to limit the SV chosen for Vinh Hoan’s fish oil byproduct. Vinh Hoan II, 40 CIT at -, 179 F.Supp.3d at 1222-24. Commerce , purported to use a SV based on Indonesian import data for HTS 1504.20.9000 while “capping” the value at a constructed value using fish oil FOPs derived from Vinh Hoan’s production data. See id. Therefore, Commerce, in effect, used a constructed value in place of a SV for fish oil based upon Indonesian pricing data. Id., 40 CIT at -, 179 F.Supp.3d at 1224. As a result, the court held that Commerce deviated, without explanation, from its standard methodology for valuing FOPs using a SV for fish oil as the best available information. 12 Id.

On second remand, Commerce continues to use a constructed value. See Second Remand Results 23-25. Commerce still has not explained its rationale for constructing a value in this case rather than choosing the best existing SV data source for fish oil from the existing alternative sources. Therefore, the court remands this issue again for Commerce to explain why constructing a value constitutes the best available information, rather than using an existing alternative SV data source for fish oil on the record, or reconsider its determination.

As already discussed, Commerce values FOPs in NME cases “based on the best available information regarding the values of such factors in a market economy country or countries.” 19 U.S.C. § 1671(b)(c)(l). Commerce uses the same methodology to “offset production costs incurred by a respondent with the sale of by-products generated during the production process.” See Final Decision Memo at 34. Commerce’s methodology for selecting the best available information evaluates data sources based upon their: (1) specificity to the input; (2) tax and import duty exclusivity; (3) contemporaneity with the period of review; (4) representativeness of a broad market average; and (5) public availability. Policy Bulletin 04.1. Although Commerce has discretion to decide what constitutes the best available information, see QVD Food Co. v. United States, 658 F.3d 1318, 1323 (Fed. Cir. 2011), Commerce must ground its selection of the best available information in the overall purpose of the ADD statute, calculating accurate dumping margins. See CS Wind Vietnam Co., Ltd. v. United States, 38 CIT -, -, 971 F.Supp.2d 1271, 1277 (2014) (citing Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed. Cir. 1990)).

Commerce has not explained, or squarely acknowledged, its use of a constructed value methodology in place of a surrogate value for fish oil. On second remand, Commerce stated it had determined that Indonesian import data for HTS 1504.20.9000 is the best available information by examining “the SVs on the record versus the value of the by-product, and whether the value of the by-product would lead to an unreasonable result.” See Second Remand Results 23. Yet Commerce does not actually use the import data for HTS 1504.20.9000 as a SV. Instead, Commerce builds a constructed value for the fish oil using fish oil FOPs and calls this value a “cap.” 13 See id. at 35-36. Commerce iden tified the FOPs used to produce fish oil from Vinh Hoan’s SV questionnaire responses. See id. at 24-25. Commerce described its calculation as applying the “verified FOPs consumed by Vinh Hoan to produce unrefined fish oil during the [period of review]” to period-of-review-specific SVs. Id. at 25. Therefore, it is apparent that Commerce selected SVs for the FOPs used to produce fish oil to construct a SV for the byproduct, rather than selecting actual SV data for fish oil on the record. However, because Commerce is not applying the calculated “capped” value to existing import data to remove any data values above the “cap,” it is apparent that Commerce is simply substituting a constructed value for a surrogate value. 14 See id. Commerce claims that the Indonesian import data is the best available data, but then claims that it cannot use that data without adjustment. Commerce cannot justify its decision to construct a value by relying on the extent to which Indonesian import data for HTS 1504.20.9000 best satisfies its SV data selection criteria, and then discard that import data for HTS 1504.20.9000. If Commerce is going to deviate from its practice of selecting the best SV data source -for a particular FOP, it must acknowledge it is doing so and explain why it is reasonable to conclude that the constructed value for that FOP yields more accurate margins than the other' SV data on the record for that FOP. 15 It may be reasonable for Commerce to construct a value; that is not yet for the court to say. But the court cannot assess the reasonableness of using a constructed value for fish oil when Commerce justifies .that methodology by claiming it is something other than what it actually is. 16

Commerce has not explained why a constructed value is a better choice than any of the other SV choices on the record; it has only explained why the Indonesian import data for HTS 1504.20.9000 is better than any of the other choices on the record. Although Commerce compares the extent to which the Indonesian import data for HTS 1504,20.9000 better satisfies its SV data selection criteria than would the Indonesian price quote on the record, see Second Remand Results 23, this analysis is of no help in discerning why a constructed value of fish oil FOPs using import data is superior to the alternative SV data sources on the record for fish oil. The constructed value does not use the import data for Indonesia HTS 1504.20,9000 in any way. Therefore, Commerce cannot justify its determination to construct a value based on the superiority of HTS 1504.20.9000 Indonesian import data to price quotes on the record. On remand, Commerce must explain why none of the SV data sources for fish oil on the record lead to a reasonable value and otherwise explain why a constructed value is superior to the alternative SV data sources for fish oil on the record or reconsider its determination. Commerce does state that using “Vinh Hoan’s own information in its production of fish oil is necessarily the most representative, and specific, value.” 17 Second Remand Results 25. However, Commerce has multiple factors that it considers in assessing the best available information. See Policy Bulletin 04.1. The court continues to defer consideration of Vinh Hoan’s arguments regarding the accuracy of Vinh Hoan’s fish oil constructed value calculations until Commerce has adequately explained the reasonableness of its practice.

CONCLUSION

The court sustains Commerce’s surrogate value data selections for rice husk and sawdust. The court also sustains Commerce’s byproduct offset calculation. However, the court again remands Commerce’s determination to construct a value for Vinh Hoan’s fish oil byproduct. Therefore, in accordance with the foregoing, it is

ORDERED that Commerce’s remand redeterminations regarding fish oil is remanded for further consideration consistent with this opinion; and it is further

ORDERED that Commerce shall file its third remand redetermination with the court within 60 days of this date; and it is further

ORDERED that the parties shall have 30 days thereafter to file comments on the third remand redetermination; and it is further

ORDERED that the parties shall have 15 days to file their replies to comments on the third remand redetermination.

1 . On June 19, 2013, Defendant submitted in-dices to the public and confidential administrative records, which identify the documents that comprise the public and confidential administrative records to Commerce's final determination. The indices to these administrative records can be located at ECF No. 27. All further references to documents from the administrative records are identified by the numbers assigned by Commerce in these administrative records.

2 . The court made the following determinations in Vinh Hoan I. The court remanded for further consideration Commerce’s: primary surrogate country selection; use of facts available for Vinh Hoan’s consignment constructed export price sales; determination not to adjust Vinh Hoan’s normal value margin calculation to exclude glazing weight; and fish oil byproduct calculation (pursuant to Defendant's request). Vinh Hoan, 39 CIT at -, 49 F.Supp.3d at 1327. The court sustained Commerce’s inclusion of sample sales in Vinh Hoan’s margin calculation. IcL The court deferred judgment on Vinh Hoan's challenges to Commerce’s SV selection, in light of remand on the issue of primary surrogate country selection. Id., 39 CIT at -, 49 F.Supp.3d at 1321.

3 . Vinh Hoan argued that capping the SV had the result of understating the by-product offset, "thus overstating Vinh Hoan’s overall margin results.’’ Vinh Hoan Comments on First Remand at 6.

4 .Commerce noted that Vinh Hoan reported the FOPs it consumed during the production of fish oil. Second Remand Results 24-25. Commerce stated that ”[t]he verified FOPs consumed by Vinh Hoan to produce unrefined fish oil during the POR were applied to POR-specific SVs from the primary surrogate country, Indonesia,” in Commerce's calculation of Vinh Hoan’s normal value. Id. at 25. Commerce then added surrogate ratios for overhead, selling, general, and administrative expenses, and profit to the value of materials used to produce Vinh Hoan’s fish oil byproduct. Id. Commerce also adjusted for yield loss (i.e,, the amount by weight of fish waste, which Commerce found to be the main input used to produce fish oil, that would be lost in the production of fish oil). See id. at 38-39.

Commerce does not indicate that it changed its general- calculation methodology from its Final Results on remand or on second remand. In the Final Results, Commerce calculated material costs for fish oil on a per-kilogram basis by multiplying a per-kilogram value for each FOP by a usage rate calculated based on Vinh Hoan’s usage data and adding the per-kilogram costs of manufacture together to derive a cost of materials for fish oil. See Eighth Administrative Review of Certain Frozen Fish Fillets from [Vietnam]: Final Results Analysis Memorandum for Vinh Hoan Corporation, A-552-801, at 6, Attach. II, CD 261, bar code 3124243-01 (Mar. 13, 2013) ("Final Analysis Memo”); see Final Decision Memo at 38. Commerce then added a per-unit overhead cost to obtain a total per-kilogram.manufacturing costs. See Final Analysis; Memo at Attach. II. Commerce then multiplied the total per-kilogram manufacturing costs by the selling, general, and administrative expense ratio and added that product to the total manufacturing costs to obtain a constructed value for fish oil. See id. Lastly, Commerce added a profit ratio to obtain a fully-loaded constructed value. Id.

5 . Commerce attributed this anomaly to the fact that the import data on the record for GTA HTS 1504.20.9000.9000 includes prices for refined fish oil while Commerce found that Vinh Hoan produced only unrefined- fish oil. Id. at 35.

6 . Commerce’s changes on remand resulted in revised weighted-average" dumping margins for mandatory respondent Vinh Hoan of $0.13 per kilogram ("kg"). Id. at 41. The weighted-average" dumping margins for man-datoiy respondent Anvifish Joint Stock Company changed to $1.26 per kg on remand.

7 .' Further citations to the Tariff Act'of 1930, as amended, are to the relevant provisions of Title 19 of the U.S. Code, 2012 edition.

8 . Commerce notes that it removed the imports from these three countries for quantities of less than 100 kg as an exception to its practice of not selectively removing import data to comply with the court's request that it analyze whether the data is aberrational. Second Remand Results 7. Moreover, Commerce notes that the inclusion or exclusion of these data has no impact on the historic AUVs for HTS 4401.30 “because these values represent broad market averages for different periods of time.” |cl at 7-8.

9 . Commerce also notes that it could not adequately compare certain record information provided about the extent to which coal produces more energy than sawdust because the information provided specific values for anthracite and bituminous coal while Vinh Hoan did not report using a specific (ype of coal as an FOP in its questionnaire responses. Second Remand Results 8-9.

In any event, Commerce questions Vinh Hoan’s assumption that the company would always consume the most economical energy source. See Second Remand Results 9. Commerce notes that Vinh Hoan reported consuming sawdust, coal, rice, and electricity as energy sources. Id. Commerce noted that the fact that Vinh Hoan consumes various energy sources undermines the notion that Vinh Hoan would select its energy inputs purely on the basis of which was most economical. See id.

10 .- Oommerce discounted the significance of -the Philippine data because it found the quantity of imports was too low and from only one country that record evidence indicates does not have a rice industry, See Second Remand Results 20.

11 . Specifically, Commerce noted that it found that the price quotes were not contemporaneous, not representative of broad market averages, that the record does not demonstrate the quotes are tax and duty exclusive. Second Remand Results 14 (citing First Remand Results 25-27). Commerce also referenced its earlier findings that the SR Apparels quote was unreliable because it lacked adequate facts about the conditions under which the price quotes were solicited and whether they were self-selected from a broader range of quotes. Id. For the Seraph International price quote, Commerce determined that, although the quote is accompanied by an affidavit indicating how it was obtained, it could not determine whether these price quotes were self-selected from the affidavit. See id. at 14-15.

12 . The court required Commerce to explain how the fish oil cap is not just a rejection of the import data in favor of a constructed value. Vinh Hoan II, 40 CIT at —, 179 F.Supp.3d at 1222. Further, the court stated that

Commerce may have a good reason to go beyond its [methodology of selecting the best existing SV data source] and construct a value, but Commerce needs to state what it is doing and explain why it is reasonable so that the court may review Commerce's methodology and determination. The court cannot review whether Commerce’s choice of Indonesian import data is reasonable when it is unclear how, to what extent, or even if Commerce used Indonesian import data for fish oil in calculating a SV for Vinh Hoan’s fish oil.

Id., 40 CIT at -, 179 F,Supp,3d at 1224.

13 . Although Commerce refers to its methodology as “capping” the Indonesian surrogate value, the term "cap” is a misnomer. There is no indication in the parties’ briefs or in the record documents that Commerce uses the “cap” as a threshold by which to retain import values below and discard import values above, as the name “cap” suggests. Rather, Commerce has constructed a value based on FOPs for Vinh Hoan’s fish oil production, and uses that constructed value as the surrogate value. See, e.g., Final Decision Memo at 38; Final Surrogate Value Memo at 6. In this way, the value used is not reflective of the Indonesian HTS data at all. Because the value used is not dependent upon the actual Indonesian import data, the fish oil byproduct value used by Commerce is in fact a constructed value, rather than a surrogate value adjusted with a cap.

14 . Commerce describe its methodology as using the data that “Vinh Hoan reported ... coupled with POR-specific SV from the primary surrogate country and adjusted by surrogate ratios, to calculate a fully loaded unrefined fish oil SV.” Second Remand Results 36.

15 . Commerce has the power to use facts available when it lacks necessary information on the record. 19 U.S.C. § 1677e(a). However, it must explain why the information it does have is insufficient.

16 . Commerce argues that it has developed a practice of constructing a value for an FOP in past cases using the same methodology it used in this case where it concluded that the constructed value represented the best available information. Second Remand Results 35-36 (citing Clearon Corp. and Occidental Chemical Corp., et al. v. United States, Court of International Trade Consolidated Court No, 13-00073, Final Results of Redetermination Pursuant to Remand at 7-11, available at http://enforcement.trade.gOv/remands/l 5-91. pdf (last visited July 5, 2017) ("Chloro Isos Remand”); Drill Pipe from the People’s Republic of China; Issues and Decision Memorandum for the Final Determination, A-570-965, at 26-29 (Jan. 3, 2011), available at http://ia.ita.doc.gov/frn/summary/prc/2011-390-l.pdf (last visited July 5, 2017) ("Drill Pipe from PRC Final Decision Memo”). However, in neither case does Commerce purport to use import data for the input in question. In both cases, Commerce simply constructs a value using other SV data on the record. See Chloro Isos Remand at 8 (wherein Commerce calculated the byproduct offset by deducting any costs associated with converting ammonia gas and sulfuric acid into ammonium sulfate from the surrogate value of the downstream product (i.e., ammonium sulfate, not the byproducts in question, which were ammonia gas and sulfuric acid); Drill Pipe from PRC I & D at 28 (wherein Commerce acknowledged it constructed a value based upon SV data for the inputs applied to the components of tool joints from the selected surro gate country and not based upon SV data for tool joints). Commerce does not justify its use of a constructed value by claiming that it is using a SV for the input in question and merely capping that SV. See Chloro Isos Remand at 8, Drill Pipe from PRC I & D at 28. Moreover, in both cases Commerce found that the SV data for the input in question was not the best available information and determined that the constructed value was the best available information to value the input in question. See Chloro Isos Remand at 9-10; Drill Pipe from PRC I & D at 26-28. Here, Commerce concludes that import data for fish oil is the best available information to value fish oil and then proceeds to construct a value for fish oil based upon SV data for the inputs used to produce fish oil. See Second Remand Results 23-25,

17 . Commerce may indeed have good reason to conclude that using Indonesian HTS data would lead to an unreasonable value for Vinh Hoan’s fish oil. But it does not follow that the best alternative is to construct a value. Nor could the idea that using the import data as is would lead to an unreasonable SV, see Def.’s Resp, Pis.' Comments Second Remand Rede-termination 11-14, Apr, 26, 2017, ECF No. 212, justify constructing a value, as Commerce has not explained why constructing a value yields superior data to using the alternative SV data sources for fish oil on the record.