JUDGMENT OF THE GENERAL COURT (Fifth Chamber)

20 September 2019 (*) (1)

(Dumping — Implementing Regulation (EU) 2017/1146 — Imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in China, manufactured by Jinan Meide Castings Co., Ltd — Definitive anti-dumping duty — Resumption of the procedure following the partial annulment of Implementing Regulation (EU) No 430/2013 — Article 2(7)(a), (10) and (11) of Regulation (EC) No 1225/2009 (now Article 2(7)(a), (10) and (11) of Regulation (EU) 2016/1036) — Normal value — Fair comparison — Non-matching product types — Article 3(1) to (3) and Article 9(4) and (5) of Regulation No 1225/2009 (now Article 3(1) to (3) and Article 9(4) and (5) of Regulation 2016/1036) — Determination of injury)

In Case T‑650/17,

Jinan Meide Casting Co. Ltd, established in Jinan (China), represented by R. Antonini, E. Monard and B. Maniatis, lawyers,

applicant,

v

European Commission, represented by J.‑F. Brakeland, M. França and N. Kuplewatzky, acting as Agents,

defendant,

ACTION under Article 263 TFEU for the annulment of Commission Implementing Regulation (EU) 2017/1146 of 28 June 2017 re-imposing a definitive anti-dumping duty on imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China, manufactured by Jinan Meide Castings Co., Ltd (OJ 2017 L 166, p. 23).

THE GENERAL COURT (Fifth Chamber),

composed of D. Gratsias (Rapporteur), President, I. Labucka and I. Ulloa Rubio, Judges,

Registrar: S. Bukšek Tomac, Administrator,

having regard to the written part of the procedure and further to the hearing on 7 March 2019,

gives the following

Judgment

I.      Background to the dispute

1        The applicant, Jinan Meide Casting Co. Ltd, is a company established in China that produces threaded tube or pipe cast fittings, of malleable cast iron, for the domestic market and export.

A.      Background to the dispute in Case T‑424/13

2        The background to the dispute on which the General Court ruled in the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), as set out in paragraphs 1 to 51 of that judgment, may be summarised as follows.

3        On 16 February 2012, the European Commission published a notice of initiation of an anti-dumping proceeding concerning imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China, Thailand and Indonesia (OJ 2012 C 44, p. 33).

4        The investigation of dumping and injury covered the period from 1 January to 31 December 2011 (‘the investigation period’). The examination of trends relevant for the assessment of injury covered the period from 2008 to the end of the investigation period.

5        With regard to exports from China, the Commission selected a sample of three exporting producers, representing 88% of the volume of exports made by the cooperating companies. The applicant was one of that sample.

6        The Commission refused to grant those three exporting producers market economy treatment (‘MET’), as provided for in Article 2(7)(b) of Council Regulation (EC) No 1225/2009 of 30 November 2009 on protection against dumped imports from countries not members of the European Community (OJ 2009 L 343, p. 51) (now Article 2(7)(b) of Regulation (EU) 2016/1036 of the European Parliament and of the Council of 8 June 2016 on protection against dumped imports from countries not members of the European Union (OJ 2016 L 176, p. 21)). However, it granted them individual treatment of their dumping margin, pursuant to the second subparagraph of Article 9(5) of Regulation No 1225/2009 (now the second subparagraph of Article 9(5) of Regulation 2016/1036).

7        The Commission considered that, for the purpose of determining normal value, it was appropriate to choose India as a market economy third country within the meaning of Article 2(7)(a) of Regulation No 1225/2009 (now Article 2(7)(a) of Regulation 2016/1036). Only one Indian producer (‘the analogue country producer’) agreed to provide the data necessary for the determination of normal value.

8        On 14 November 2012, the Commission adopted Regulation (EU) No 1071/2012 imposing a provisional anti-dumping duty on imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China and Thailand (OJ 2012 L 318, p. 10; ‘the provisional regulation’).

9        On 13 May 2013, the Council of the European Union adopted Implementing Regulation (EU) No 430/2013 imposing a definitive anti-dumping duty and collecting definitively the provisional duty imposed on imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China and Thailand and terminating the proceeding with regard to Indonesia (OJ 2013 L 129, p. 1).

10      Article 1(1) of Implementing Regulation No 430/2013 provided:

‘A definitive anti-dumping duty is hereby imposed on imports of threaded tube or pipe cast fittings, of malleable cast iron, excluding bodies of compression fittings using ISO DIN 13 metric thread and malleable iron threaded circular junction boxes without having a lid, currently falling within CN code ex 7307 19 10 (TARIC code 7307191010) and originating in the People’s Republic of China (“PRC”) and Thailand.’

11      Article 1(2) of Implementing Regulation No 430/2013 provided, in respect of the applicant’s exports, that the rate of the definitive anti-dumping duty applicable to the net free-at-Union-frontier price, before duty, of the product described was 40.8%.

B.      The action in Case T‑424/13

12      By application lodged at the Registry of the General Court on 7 August 2013, the applicant brought an action (‘the initial action’) seeking the annulment of Implementing Regulation No 430/2013, in so far as it applied to it (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 52).

13      The initial action was based on five pleas in law (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 57).

14      The first plea alleged infringement by the EU institutions of the applicant’s rights of defence and of various provisions of Regulation No 1225/2009, in that those institutions had refused to disclose to the applicant the information relevant for the determination of normal value. In the context of the first plea, the applicant raised three complaints. In particular, by the first of them, it complained that the EU institutions had refused it access to the normal value calculations after it had received authorisation from the analogue country producer to examine the data underlying those calculations (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 57).

15      The second plea was based mainly on manifest errors of assessment and errors of law, on the ground that the institutions had rejected claims for adjustments to the normal value, in respect of raw materials and productivity, submitted by the applicant and, in the alternative, on a failure to state reasons (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 57). The third plea was based on manifest errors of assessment and errors of law and on infringement of the principle of non-discrimination, on the ground that the institutions had followed an unreasonable methodology for determining the normal value for non-matching products (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 57). The fourth plea was based on infringement of essential procedural requirements, on the ground that the Commission had notified the conclusions relating to the MET late (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 57). The fifth plea alleged errors of fact and manifest errors of assessment and an infringement of Regulation No 1225/2009, on the ground that the determination of the injury suffered by the European Union industry was based on incorrect data with regard to the volume of dumped imports from China (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 57).

16      In its judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), the Court rejected the fourth plea (paragraphs 59 to 89 of that judgment), as well as the second and third complaints raised in the first plea (paragraphs 108 to 127 of that judgment).

17      By contrast, the Court upheld the first complaint raised in the first plea (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraphs 128 to 221). The Court concluded that Implementing Regulation No 430/2013 had to be annulled, without it being necessary to examine the second, third and fifth pleas in the initial action (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 221).

18      In paragraph 1 of the operative part of the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), the Court decided:

‘[To annul] Council Implementing Regulation (EU) No 430/2013 of 13 May 2013 imposing a definitive anti-dumping duty and collecting definitively the provisional duty imposed on imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China and Thailand and terminating the proceeding with regard to Indonesia, to the extent that it applies to Jinan Meide Casting Co. Ltd’.

C.      Background to the dispute subsequent to the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13)

19      On 28 October 2016, the Commission published a notice relating to the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), concerning Implementing Regulation No 430/2013 (OJ 2016 C 398, p. 57; ‘the 28 October 2016 notice’).

20      In the third recital of the 28 October 2016 notice, the Commission stated that, in accordance with Article 266 TFEU, the request of the exporting producer concerned for disclosure of the normal value calculations using confidential data of the analogue country producer should be re-examined in the light of the particular circumstances relating to that exporting producer.

21      In the fourth recital of the 28 October 2016 notice, the Commission pointed out that the annulment of Implementing Regulation No 430/2013 concerned one step of the administrative proceeding, namely the disclosure of information to the exporting producer. It therefore considered that, in complying with the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), it had the possibility to remedy the aspects of the proceeding which had led to the annulment, while leaving unchanged those parts which were not affected by the judgment, and that the findings reached in Implementing Regulation No 430/2013 which had not been contested within the time limits for a challenge, or which had been contested but rejected by the Court’s judgment or not examined by the Court, and therefore did not lead to the annulment of Implementing Regulation No 430/2013, remained valid.

22      In the fifth and sixth recitals of the 28 October 2016 notice, the Commission stated that, in view of the above, it would reopen the anti-dumping investigation concerning imports of malleable fittings originating in China which had led to the adoption of Implementing Regulation No 430/2013, in so far as it concerned the exporting producer concerned, and resumed that investigation at the point at which the irregularity occurred, by publishing that notice in the Official Journal of the European Union and that that reopening was limited in scope to the implementation of the Court’s judgment with regard to the applicant.

23      The Commission provided the applicant with several successive versions of the dumping margin calculations and various documents relating to those calculations and the data provided by the analogue country producer (letters dated 23 December 2016, 31 January, 14 February and 12 April 2017 and emails dated 21 April and 29 May 2017).

24      The applicant submitted comments by letters dated 19 January and 2 May 2017.

25      On 8 March 2017, the Commission held a hearing with the applicant and two importers of the product concerned and, on 15 March and 25 April 2017, it held two other hearings with the applicant.

26      On 28 June 2017, the Commission adopted Implementing Regulation (EU) 2017/1146 re-imposing a definitive anti-dumping duty on imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China, manufactured by Jinan Meide Castings Co., Ltd (OJ 2017 L 166, p. 23; ‘the contested regulation’).

27      In recitals 4 to 6 of the contested regulation, the Commission justified the detailed rules for the implementation of the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378) in similar terms to those of recitals 4 to 6 of the 28 October 2016 notice (see paragraphs 20 to 22 above).

28      Article 1(1) of the contested regulation provides:

‘A definitive anti-dumping duty is hereby imposed on imports of threaded tube or pipe cast fittings, of malleable cast iron, excluding bodies of compression fittings using ISO DIN 13 metric thread and malleable iron threaded circular junction boxes without having a lid, currently falling within CN code ex 7307 19 10 (TARIC code 7307191010) and originating in the People’s Republic of China and manufactured by Jinan Meide (TARIC additional code B336).’

29      Article 1(2) of the contested regulation provides:

‘The rate of the definitive anti-dumping duty applicable to the net, free-at-Union-frontier price, before duty, shall be 39[.]2%.’

II.    Procedure and forms of order sought

30      By application lodged at the Registry on 25 September 2017, the applicant brought the present action.

31      On 22 December 2017, the Commission lodged the defence.

32      The reply and the rejoinder were lodged, respectively, on 20 February and on 3 April 2018.

33      By letter of 4 April 2018, the parties were informed that the written part of the procedure had been closed and that they were able to request that a hearing be held under the conditions laid down in Article 106 of the Court’s Rules of Procedure. By letter of 6 April 2018, the applicant requested that a hearing be held.

34      On 1 February 2019, by a measure of organisation of procedure, the Court asked the parties a number of questions with a request for a written answer and invited them to produce certain documents. The parties replied to the Court by procedural documents dated 22 February 2019.

35      On 1 March 2019, by a new measure of organisation of procedure, the Court invited the Commission to produce a supplementary document. The Commission replied to that invitation by pleading dated 5 March 2019.

36      The hearing was held on 7 March 2019. At the hearing, the applicant confirmed that, as it had indicated in a written answer to a question put by the Court, it was withdrawing the fifth plea in law of the application.

37      The applicant claims that the Court should:

–        annul the contested regulation;

–        order the Commission to pay the costs.

38      The Commission contends that the Court should:

–        dismiss the action;

–        order the applicant to pay the costs.

III. Law

39      As stated in paragraph 36 above, the applicant withdrew the fifth plea in law of the application. The present action is therefore based on only four pleas in law. The first plea alleges infringement of Article 2(7)(a) of Regulation No 1225/2009, due to errors by the Commission in the determination of normal value. The second plea alleges an infringement of Article 2(10) of that regulation (now Article 2(10) of Regulation 2016/1036) and of Article 2.4 of the Agreement on Implementation of Article VI of the General Agreement on Tariffs and Trade 1994 (GATT) (OJ 1994 L 336, p. 103; ‘the anti-dumping agreement’), which is part of Annex 1A of the Agreement establishing the World Trade Organisation (WTO) (OJ 1994 L 336, p. 3), owing to the Commission’s erroneous rejection of some of the applicant’s requests for adjustment. The third plea alleges infringement of Article 2(7)(a), Article 2(10) ab initio and (a) and Article 2(11) of that regulation (now Article 2(7)(a), Article 2(10) ab initio and (a) and Article 2(11) of Regulation 2016/1036), owing to errors by the Commission in determining the normal value of non-matching product types. The fourth plea alleges infringement of Article 3(1) to (3) of the regulation at issue (now Article 3(1) to (3) of Regulation 2016/1036), owing to the Commission’s use of incorrect import data and, moreover, infringement of Article 3 and Article 9(4) and (5) of the same regulation (now Article 9(4) of Regulation 2016/1036), as well as a failure to state reasons, on the ground that, in the contested regulation, the Commission did not explicitly adopt injury and causal link findings.

40      At the outset it should be noted that, in recital 19 of the contested regulation, the Commission stated that the law applicable to the reopening of the anti-dumping investigation was Regulation No 1225/2009, which was the substantive law at the time of the adoption of the regulation annulled by the Court. It added that, in any event, Regulation 2016/1036, which repealed and replaced Regulation No 1225/2009 with effect from 19 July 2016, was a codification of the latter regulation and its subsequent amendments. In the application, the applicant agreed with those considerations, provided that the expression ‘basic regulation’ used in the contested regulation is interpreted as referring to Regulation No 1225/2009.

41      In this respect, according to settled case-law, if the legal basis of an act and the applicable procedural rules must be in force when the act is adopted, compliance with the principles governing the temporal application of the law and the requirements relating to the principles of legal certainty and the protection of legitimate expectations require the application of the substantive rules in force at the date of the facts in issue, even if those rules are no longer in force when an EU institution adopts an act (see, to that effect, judgment of 14 June 2016, Commission v McBride and Others, C‑361/14 P, EU:C:2016:434, paragraph 40 and the case-law cited).

42      It follows that, in the present case, if the contested regulation were to be adopted on the basis of Regulation 2016/1036 and in accordance with the procedural rules laid down in that regulation, the lawfulness of that regulation is to be assessed in the light of the rules of substantive law which were applicable to the facts covered by the anti-dumping investigation, namely the rules laid down in Regulation No 1225/2009. In so far as the pleas in law of the application exclusively concern the application of those rules of substantive law, in the context of the examination of those pleas in paragraphs 44 to 412 below, reference should therefore be made only to Regulation No 1225/2009 (‘the basic regulation’).

43      In the first place, the Court considers it appropriate to examine the third plea in law.

A.      The third plea, alleging infringement of Article 2(7)(a), Article 2(10) ab initio and (a) and Article 2(11) of the basic regulation, owing to errors made by the Commission in determining the normal value of non-matching product types

44      The third plea consists of two parts. The first alleges that the methodology adopted by the Commission to determine the normal value of the non-matching product types (‘the contested methodology’) is unreasonable and the second alleges that that methodology does not reflect the full degree of dumping being practised, contrary to Article 2(11) of the basic regulation.

1.      Preliminary observations

45      At the outset it should be recalled that, pursuant to Article 2(7)(a) of the basic regulation, two methods are possible for determining normal value when the exporting country is a non-market economy country.

46      As the Court of Justice has noted, it follows from the wording and the scheme of the provisions of Article 2(7)(a) of the basic regulation that the main method for determining normal value in such a case is that of ‘the price or constructed value in a market economy third country’ or ‘the price from such a third country to other countries, including [the European Union]’. Failing that, the stated alternative method of determining the normal value is that that value is to be determined ‘on any other reasonable basis, including the price actually paid or payable in the [European Union] for the like product, duly adjusted if necessary to include a reasonable profit margin’ (judgment of 22 March 2012, GLS, C‑338/10, EU:C:2012:158, paragraph 24).

47      According to the Court of Justice, the objective of the priority given to the main method prescribed by those provisions is to obtain a reasonable determination of the normal value in the country of export through the choice of a third country in which the price for a like product is formed in circumstances which are as similar as possible to those in the country of export, provided that it is a market economy country. It follows that the application of that main method can be excluded only if it cannot be applied (see, to that effect, judgment of 22 March 2012, GLS, C‑338/10, EU:C:2012:158, paragraphs 25 and 26).

48      Moreover, it is apparent from both the wording and the scheme of Article 2(10) of the basic regulation that an adjustment to the export price or the normal value may be made only to take account of differences in factors which affect the prices and therefore their comparability. That means, in other words, that the purpose of the adjustment is to re-establish the symmetry between normal value and export price, with the result that, if the adjustment has not been validly made, that implies a contrario that it has created an asymmetry between those two values (see, to that effect, judgment of 16 December 2011, Dashiqiao Sanqiang Refractory Materials v Council, T‑423/09, EU:T:2011:764, paragraph 42 and the case-law cited).

49      In particular, Article 2(10)(a) of the basic regulation provides that an adjustment is to be made for differences in the physical characteristics of the product concerned and that the amount of the adjustment is to correspond to a reasonable estimate of the market value of the difference. However, that provision does not specify how such a reasonable estimate should be arrived at. Furthermore, it should be noted that that provision does not require, in order to re-establish the symmetry between the normal value of the like product and the export price of the product concerned, the amount of the adjustment thus assessed to accurately reflect such a market value, but only to constitute a reasonable estimate thereof.

50      As follows from the case-law, within the limits defined, on the one hand, by Article 2(7)(a) of the basic regulation and, on the other hand, Article 2(10) of that regulation, the Commission has a wide discretion both in assessing the normal value of a product and in assessing facts justifying the fairness of the comparison of normal value and export price made, with the vague concepts of reasonableness and fairness to be applied by the Commission in the context of those provisions having to be made concrete by it on a case-by-case basis, depending on the relevant economic context (see, to that effect, judgments of 7 May 1987, NTN Toyo Bearing and Others v Council, 240/84, EU:C:1987:202, paragraph 19, and of 16 December 2011, Dashiqiao Sanqiang Refractory Materials v Council, T‑423/09, EU:T:2011:764, paragraphs 40 and 41 and the case-law cited).

51      However, in making those assessments, the Commission must ensure that it takes as a basis values and parameters that can be verified as the normal result of market forces, including effective competition (see, to that effect and by analogy, judgment of 10 September 2015, Fliesen-Zentrum Deutschland, C‑687/13, EU:C:2015:573, paragraphs 66 to 68).

52      In addition, the methods used must be consistent with the final objective of the calculation of the dumping margin which, as follows from Article 2(11) of the basic regulation, is to reflect the full degree of dumping being practised (see, to that effect, judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council, C‑376/15 P and C‑377/15 P, EU:C:2017:269, paragraph 54).

53      In general, it is for the Courts of the European Union to verify that, in choosing the methods for determining normal value and ensuring a fair comparison between normal value and export prices, the Commission did not fail to take into account essential elements in order to establish the appropriate nature of those choices and that all the elements of the file were examined with all the care required (see, to that effect and by analogy, judgment of 22 March 2012, GLS, C‑338/10, EU:C:2012:158, paragraph 22).

2.      The first part, based on the fact that the Commission adopted an unreasonable methodology to determine the normal value for non-matching product types

54      The applicant claims that the contested methodology is based on the erroneous assumption that the market value of the physical differences is reflected in the export prices whereas, according to the Commission’s own findings, the same export prices reflect the dumping at least partially. Furthermore, it claims that that methodology is based on the erroneous assumption that the export prices of the non-matching product types reflect a level of dumping equivalent to that found for those product types for which there was a directly comparable product type (‘the directly comparable product types’). According to the applicant, that assumption is unreasonable and unverifiable. The applicant claims that the Commission therefore infringed Article 2(7)(a) and Article 2(10) ab initio and (a) of the basic regulation. In addition, at the hearing, the applicant pointed out that there were a large number of alternative methods that the Commission could have used under the applicable provisions.

55      The Commission replies that the average normal value was corrected by the market value of the physical differences, in accordance with Article 2(10)(a) of the basic regulation. It asserts that, in the light of the definition of market value by the International Valuation Standards Council (IVSC), the market value must be considered to be reflected in the export prices. According to the Commission, it must be presumed that the first independent customer pays the market value and that the export price is the price paid on the EU market. The Commission further asserts that its method made it possible to reduce the impact of product types with a high dumping margin and that, consequently, the overall dumping margin was lowered to the benefit of the applicant. In its written replies to the Court’s questions and at the hearing, the Commission emphasised that the contested methodology was applied, following the applicant’s requests and taking into account the impossibility of making an individual determination, product type by product type, of the differences in physical characteristics. It submits, in essence, that the lawfulness of that methodology must be examined in the light of Article 2(10) of the basic regulation and not in the light of Article 2(7)(a) or Article 2(11) thereof, which concern different steps in the determination of the dumping margin.

56      Before examining the applicant’s arguments, it is necessary to go back to the content of the contested methodology and the proportion of the applicant’s export volume affected by the application of that methodology.

57      As follows from recital 68 of the provisional regulation, the Commission had initially calculated the applicant’s dumping margin on the basis of a comparison between the weighted average normal value of each directly comparable like product type established for the analogue country and the weighted average export price of the corresponding type of product concerned. Therefore, it had not included in that calculation transactions falling under the non-matching product types. Subsequently, it accepted the applicant’s request to take those transactions into account in the calculation of the dumping margin, which, according to the uncontested information contained in the application, represented 44% of its total export volume to the European Union. Thus, recital 18 of Implementing Regulation No 430/2013 states that, for not directly comparable product types, the normal value was based on the arithmetical average normal value for the directly comparable product types, adjusted by the market value of the differences in the physical characteristics pursuant to Article 2(10)(a) of the basic regulation.

58      According to point 2.2.3 of the Annex to the Commission’s information document of 23 December 2016, entitled ‘Calculation of the dumping margin of Jinan Meide Casting CO., Ltd “JMCC”’ (‘the 23 December 2016 document’), the Commission implemented the contested methodology in the following manner.

59      The Commission determined that the linear average export price of all directly comparable product types was 16 Chinese yuan (CNY)/kg (around EUR 2.12/kg). The linear average of the normal value for the corresponding similar product types was determined as CNY 20.91/kg (approximately EUR 2.77/kg). Next, in order to determine an adjustment for physical differences between directly comparable product types and non-matching product types, the Commission calculated the ratio between the average export sales price of each non-matching product type and the linear average export price of the directly comparable product types. It then applied those ratios, as a percentage, to the linear average of the normal value and thus obtained the normal value of each non-matching product type, adjusted for physical differences.

60      For example, the average unit export price of the non-matching product type with product control number 0002FF00BN was CNY 11.83/kg (approximately EUR 1.57/kg), which is equivalent to 73.92% of the linear average export price of the directly comparable product types. The Commission applied that ratio of 73.92% to the linear average of the normal value of those product types and thus obtained the normal value for the non-matching product type at issue, adjusted for physical differences, of CNY 15.46/kg (approximately EUR 2.05/kg).

61      The applicant contested the validity of that methodology in the administrative procedure preceding the adoption of Implementing Regulation No 430/2013 and before the Court, in connection with the third plea in law of its initial action (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraphs 57 and 123). However, as was pointed out in paragraph 17 above, the Court upheld the first complaint raised in the first plea, relating to the Commission’s refusal to disclose the normal value calculations, without examining the second, third and fifth pleas in the action. Although the applicant repeated its observations in connection with the reopening of the procedure, the Commission did not re-examine that issue, with the result that it recalculated the applicant’s dumping margin by applying the contested methodology once again.

62      However, it should be noted that, as can be deduced from point 2.2.3 of the 23 December 2016 document and, as confirmed by the parties in a written reply to a question put by the Court, for the purpose of calculating normal value and the dumping margin, in addition to directly comparable product types and non-matching product types, the Commission identified a third category of product types, namely ‘quasi-matching’ product types.

63      In this respect, as regards the category of ‘quasi-matching’ product types, it follows from the indications in point 2.2.3 of the 23 December 2016 document that it includes product types sold by the applicant which differ from directly comparable product types only by the fact that their surface is not galvanised. The product types belonging to that category had their normal value adjusted on the basis of the normal value of the corresponding galvanised like product type. For a product type with a black surface (B), normal value was set at 80% of the normal value of the corresponding galvanised like product type. For the other surfaces (A, E and M), the same normal value of the galvanised like product type in its entirety was used.

64      According to the indications of the parties in their written replies to the questions put by the Court, the distribution of the 1 528 product types sold for export among the three categories set out in paragraph 62 above is as follows.

65      In the first place, 202 product types, representing 55% of the applicant’s total export volume, were considered to be directly comparable product types. For those product types, the dumping margin was therefore determined by calculating the normal value, as indicated in recitals 17 and 19 of Implementing Regulation No 430/2013, on the basis of the domestic sales of the analogue country producer made in the ordinary course of trade or on the basis of constructed value. In the second place, 343 product types, representing 17% of that total volume, were considered to be ‘quasi-matching’ product types, for which the dumping margin was determined by adjusting the normal value on the basis of the methodology described in paragraph 63 above. In the third place, the remaining product types, namely 983 product types representing 28% of the same total volume, were considered to fall within the category of non-matching product types, for which normal value was calculated and adjusted in accordance with the contested methodology.

66      It therefore follows from the above that the contested methodology was applied for the purpose of determining the dumping margin of part of the applicant’s exports representing between a quarter and a third of their total volume, that is, a significant proportion of that volume. The use of that methodology is therefore likely to have had a significant impact on the calculation of the dumping margin determined for all those exports.

67      It is necessary to examine, first of all, the applicant’s argument that the contested methodology is based on the claim that the applicant’s export prices reflect the market value of the differences in physical characteristics, which would be in contradiction with the Commission’s finding that those export prices are at least partially dumped.

68      It should be noted that, in view of the information in paragraphs 57 to 59 above, the contested methodology can be described as a combination of two steps, namely, on the one hand, the determination of normal value in accordance with the main method provided for in Article 2(7)(a) of the basic regulation (see paragraphs 46 and 47 above) and, on the other hand, the application of an adjustment for physical differences, under the conditions laid down in Article 2(10)(a) of the same regulation. In sum, as follows from that information, the Commission considered that the differences between the export prices charged by the applicant for the non-matching product types and the same prices for the directly comparable product types constituted a reasonable estimate of the value of the physical differences between those product types for the purpose of making a normal value adjustment.

69      The applicant does not call into question, as such, the lawfulness of the first step of that methodology, namely the reference to the average unit price of the like product on the Indian market for the determination of normal value, but calls into question only the lawfulness of the second step, namely the use of the ratio between the price of each non-matching product type and the average unit export price of directly comparable product types to determine the amount of the normal value adjustment for the purpose of a fair comparison, within the meaning of Article 2(10) of the basic regulation.

70      It must be stated that the applicant’s argument is based on a correct premiss. Indeed, a price likely to be affected by dumping cannot form the basis for a reasonable estimate of the market value of differences in physical characteristics within the meaning of Article 2(10)(a) of the basic regulation, since such a price may not be the result of normal market forces.

71      Therefore, it seems paradoxical for the Commission to adjust the normal value of a given like product type by means of a value potentially affected by dumping, while it seeks to establish that normal value on the basis of its value in a market economy third country with the objective of identifying a price formed in circumstances which are as comparable as possible to those in the exporting country.

72      Indeed, by definition, it is not possible for the Commission to assume that such a value, which is potentially affected by dumping, was formed under market economy conditions. Thus, at that stage of the procedure, the Commission cannot exclude that that value may be the result of an artificial undervaluation leading to an amount lower than that at which that value would have been fixed if it had resulted only from forces freely exerted on the market.

73      Furthermore, it must be noted that the use, for the purpose of a fair comparison, of an adjustment to the normal value corresponding to an amount determined on the basis of export prices for which the Commission specifically seeks to assess the undervaluation due to dumping does not reflect a consistent approach.

74      In this respect, it should be noted that, in order to be able to determine the dumping margin in a reasonable and objective manner, the calculation of the normal value of a given product type must be based, in principle, on data independent of the export prices for which the Commission specifically seeks to assess, by establishing that normal value, the undervaluation to which they are subject.

75      The normal value is the reference value against which the export price potentially affected by dumping is compared. That comparison is distorted if, in the calculation of such a reference value, a constitutive element of the export price to be compared is introduced.

76      It is true that the Commission was entitled to make an adjustment to the normal value in this case since the normal value had been determined on the basis of the average unit value of the directly comparable product types on the Indian market. As the Commission stated in a written answer to a question put by the Court, the uniform application of that average unit value for the determination of the normal value of all non-matching product types was not appropriate in this case, as it had found that the variation in the average export prices of the different non-matching product types was significant, ranging from less than CNY 10/kg (around EUR 1.32/kg) to more than CNY 100/kg (around EUR 13.2/kg). Thus, in the absence of an adjustment, such a uniform application of the average unit value would not have ensured comparability between the normal value and the export price of each non-matching product type.

77      However, the Commission has not demonstrated that the use of a constituent element of the export prices of the non-matching product types, in order to correct the normal value to which those prices are compared, was such as to re-establish the symmetry between those prices and that normal value in accordance with the objective of Article 2(10) of the basic regulation. In particular, there was no indication that the ratio between the export price of each non-matching product type and the average unit export price of the directly comparable product types correctly reflected the value of the physical differences between the latter category of product type and the non-matching product type in question.

78      It is true that it cannot be excluded that, in some cases, the difference between the export prices of certain specific non-matching product types and the export prices of directly comparable product types may correspond to the market value of the physical characteristics of the non-matching product types in question. Furthermore, Article 2(10)(a) of the basic regulation does not require that the adjustment fully reflect that value, but only that it constitute a reasonable estimate.

79      That said, in the present case, the contested methodology is based on the presumption that that price difference corresponds to the market value of the physical differences for all non-matching product types.

80      As the applicant has pointed out, without being challenged on this point by the Commission, that presumption necessarily implies the assumption that exports of all those product types are affected by dumping at a level equivalent to that found for directly comparable product types.

81      Indeed, the presumption that the price difference between the two categories of product types at issue corresponds to the market value of the differences in physical characteristics is tantamount to assuming that, if there were no such differences in physical characteristics, the non-matching product types would be sold for export at the same price as the directly comparable product types.

82      Such a presumption therefore implies, by definition, that the dumping margin likely to affect the prices of those two product type categories is at the same level. Otherwise, the price differences between the two categories of product types at issue may result, at least in part, from the differences in the dumping margin and therefore cannot be considered with sufficient reliability to reflect only the differences in physical characteristics.

83      That being said, as the applicant rightly claims in the context of the second argument made in connection with this part of the plea, the assumption of equivalent dumping margins for the two categories of product types at issue cannot be considered to be reasonable or verifiable.

84      On the one hand, it is difficult to reconcile that assumption with the use of the calculation of the dumping margin by product type, which assumes, on the contrary, that that dumping margin may be different according to the type of product concerned and that it is necessary to make that calculation in order to properly reflect the full degree of dumping being practised, pursuant to Article 2(11) of the basic regulation (see, to that effect, judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council, C‑376/15 P and C‑377/15 P, EU:C:2017:269, paragraph 54).

85      On the other hand, the table of detailed calculations of the applicant’s dumping margin, product type by product type, which is annexed to the document which the Commission disclosed to the applicant on 29 May 2017 and forwarded to the Court as part of a measure of organisation of procedure (‘the 29 May 2017 table’), does not allow that assumption to be substantiated.

86      That table shows that, for the directly comparable or ‘quasi-matching’ product types, there is a particularly large range of dumping margins between negative values below -50% and positive values close to 500%. The Commission has not provided any evidence to support the hypothesis that this would be different for the non-matching product types.

87      Moreover, like the applicant, the Court finds that, as is illustrated by the description of the contested methodology in point 2.2.3 of the 23 December 2016 document (see paragraph 59 above) and as confirmed by the simplified example of that methodology in the application, the application of that methodology should in principle result in an identical rate for the average dumping margin of the directly comparable product types and that of each of the non-matching product types.

88      Indeed, as the applicant has explained, on the basis of the theoretical assumption that (i) the average normal value and the average export price of the directly comparable product types are 13 and 10 respectively and (ii) that the export price of a given non-matching product type is 7, the application of the Commission’s methodology results in the ratio of that price of 7 to the average export price of 10 — that is, a rate of 70% — being applied to the average normal value of 13 in order to obtain the normal value, after adjustment, of the relevant non-matching product type, that is, 9.1. As the applicant points out, that calculation results, in the simplified form set out in the application and which is not challenged by the Commission, in a dumping margin for the non-matching product type which is identical to the average dumping margin of the directly comparable product types, namely 30% in the fictitious example used by the applicant.

89      The example provided by the Commission in point 2.2.3 of the 23 December 2016 document, which is based on the figures actually used to establish the adjusted normal value of non-matching products, produces a similar result.

90      As was mentioned in paragraph 59 above, the Commission determined an average unit price of the product concerned for export of CNY 16/kg and an average unit price of the like product on the Indian market of CNY 20.91/kg. If the average dumping margin of the directly comparable product types was calculated solely on the basis of the comparison between those two prices, it would be 30.7%.

91      As the Commission points out, the application of the contested methodology to determine the adjusted normal value of the non-matching product type with product control number (PCN) 0002FF00BN, whose average unit export price was CNY 11.83, results in an amount of CNY 15.46. It must be noted that that calculation would result in a dumping margin for the non-matching product type at issue at the rate of 30.7%, that is, the same rate as the average dumping margin referred to in paragraph 90 above.

92      It is true that, as follows from a written reply from the Commission to a question put by the Court, the calculation of the dumping margin for each product type and for the product concerned as a whole is more complex than in the above examples and involves additional operations.

93      Indeed, as explained by the Commission, for each product type, the difference between the export sales price and the normal value is multiplied by the quantity exported to obtain the total amount of dumping. That amount of dumping is then related to the total amount of exports to obtain the dumping margin for the product type at issue, which is expressed as a percentage of the cost, insurance, freight (CIF) price at the EU border and before duty. As shown in the 29 May 2017 table, the same operations are carried out for the determination of the dumping margin for the whole product concerned after adding the dumping amounts of each product type.

94      Furthermore, as the applicant stated in the application, in practice, for a given product type, the export price which is used in the fair comparison between the normal value and that price is the ex-works price of the product concerned, whereas the export price to which the amount of dumping is related in order to determine the dumping margin for that product type is a CIF price, that is, a price including all costs of transport to the EU border.

95      That being said, the 29 May 2017 table shows that those differences between the simplified examples of the applicant and the Commission and the dumping margin calculations carried out in practice by the Commission did not significantly alter the result of the contested methodology. It is apparent from that table that the dumping margins calculated for the 983 non-matching product types are within a value range of between 24% and 28%. However, the small magnitude of those dumping margins is not comparable to that found for the dumping margins of directly comparable and ‘quasi-matching’ product types, which, as was noted in paragraph 86 above, range from negative values below -50% to positive values close to 500%.

96      It follows from all of the foregoing that the Commission has not demonstrated that, by the contested methodology, it made a reasonable estimate of the market value of the differences in physical characteristics between the non-matching product types and the directly comparable product types. Therefore, it did not demonstrate that the application of that methodology resulted in a fair comparison between the normal value and export prices. Furthermore, it did not demonstrate that the adjustment to the normal value of the non-matching product types made in this way preserved the reasonable determination of that normal value, that is, a determination based on values and parameters which can be considered to be the normal result of market forces. The application of the erroneous contested methodology is therefore not in accordance with Article 2(7)(a) or Article 2(10) ab initio and (a) of the basic regulation.

97      It is true that an error in the reasoning of the author of the contested act or the method it used cannot be sufficient to warrant the annulment of that act if, in the particular circumstances of the case, that error could not have had a decisive effect on the outcome (see, to that effect and by analogy, judgments of 9 July 2008, Alitalia v Commission, T‑301/01, EU:T:2008:262, paragraph 307 and the case-law cited, and of 15 December 2010, CEAHR v Commission, T‑427/08, EU:T:2010:517, paragraph 161 and the case-law cited).

98      However, it must be noted that this is not the case here.

99      First, as follows from paragraph 66 above, given the significant proportion of the applicant’s export volume concerned by the application of the contested methodology, it cannot be excluded that it had a significant impact on the rate of the applicant’s dumping margin adopted in paragraph 1 of the operative part of the contested regulation.

100    Secondly, the reasonableness of the estimate of the value of the physical differences between the non-matching product types and the directly comparable product types and the fairness of the comparison resulting from the adjustment based on that estimate cannot be assessed in the light of the existence or non-existence of more appropriate alternative methodologies.

101    Indeed, as the Court of Justice has found, while it follows from a combined reading of Article 2(10) and (11) of the basic regulation that the calculation of the dumping margin must be made on the basis of a fair comparison, the notion of ‘fairer comparison’ appears nowhere in those provisions. In any event, since a method cannot be considered to be a means of ensuring a fair comparison, it cannot be argued that the use of another method for calculating the normal value would not have ensured a fairer comparison (see, to that effect, judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council, C‑376/15 P and C‑377/15 P, EU:C:2017:269, paragraph 71). Those considerations apply mutatis mutandis to the notion of reasonable estimate in Article 2(10)(a) of the basic regulation.

102    In the present case, invited by the General Court in connection with a question with a request for a written answer to indicate whether it had considered methodologies other than the contested methodology, in relation to the non-matching product types, the Commission indicated that it had chosen the latter after having excluded three possibilities. First, in view of the very large number of product types exported by the applicant (around 1 500), in particular non-matching product types (almost 1 000), the Commission excluded an individual determination of the normal value of each non-matching product type. Secondly, for the same reasons, the Commission also considered it impossible to identify products with close similarities. Thirdly, as was stated in paragraph 76 above, the Commission also rejected the possibility of applying a uniform normal value for all non-matching product types due to the significant variation in export prices of those different product types.

103    Furthermore, in the same written reply and at the hearing, the Commission considered that it was not appropriate to consider relying on prices of product types distributed by the analogue country producer on the Indian market but not produced by the producer itself. It claimed that it had no information on the prices of those product types, that it was possible that those product types, imported from China, had been dumped and that their use in this case would essentially amount to a comparison of Chinese exports to the European Union and Chinese exports to India. It further argued at the hearing that Article 2(7)(a) of the basic regulation did not allow it to apply, for part of the product types concerned, the main methodology for determining normal value provided for in those provisions and the alternative methodology for the other part. In the Commission’s view, it was therefore not possible to use the EU producers’ prices for non-matching product types. The Commission further argued that, should it have been able to use those prices, it would in any event have had to make adjustments in view of the large number of product types sold for export by the applicant.

104    However, it should be observed that, even if the Commission were justified in taking the view that all those alternative methods would either have been inappropriate or impossible to implement or would not necessarily have avoided the use of adjustments such as those applied under the contested methodology, it follows from paragraph 96 above that the Commission did not demonstrate that the application of the contested methodology made it possible to make a fair comparison between the normal value and export prices and that it preserved the reasonable nature of the determination of that normal value.

105    In any event, the Commission has not demonstrated the absence of any possible alternative methodology.

106    In this respect, as the Court of Justice has stated, in a case such as the present case, where the analogue country producer neither produces nor sells a certain product type, the EU institutions may either decide to exclude that product type from the definition of ‘the product under consideration’ or construe the normal value for that product type so as to take into consideration export transactions of that same product type for the purposes of calculating the dumping margin (see, by analogy, judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council, C‑376/15 P and C‑377/15 P, EU:C:2017:269, paragraph 70).

107    Moreover, assuming that, in the present case, the large number of non-matching product types made it difficult to apply either of those solutions, it should be noted that, invited by the Court to express its views in this connection at the hearing, the Commission did not demonstrate that it would not have been able to make methodological choices in accordance with the applicable rules if it had examined at an earlier stage of the anti-dumping investigation the issues of calculating the normal value of non-matching product types and any necessary adjustments.

108    It should be recalled that, as was stated in paragraph 57 above, in the provisional regulation, the Commission had initially excluded non-matching product types from the calculation of normal value and it was only after having taken note of the applicant’s comments in this respect that, within the framework of Implementing Regulation No 430/2013, the institutions decided to take those product types into account in the calculation.

109    The Commission’s explanations at the hearing do not show that, prior to the adoption of the provisional regulation, it could not have considered other methods to take into account the non-matching product types in the calculation of the dumping margin, since it was in a position to ascertain at that stage, on the basis of a comparison of the analogue country producer’s data and the applicant’s data, that the analogue country producer produced only a limited number of the product types sold for export by the applicant.

110    In particular, the Commission did not provide any concrete evidence to exclude the possibility that, in order to make a reasonable estimate of the market value of the physical differences in order to make the necessary adjustments to the normal value of the non-matching product types, it used the data available at the time with regard to producer prices in the European Union (see recital 109 of the provisional regulation).

111    Admittedly, as the Commission has explained, in the light of the case-law referred to in paragraphs 46 and 47 above, the wording of Article 2(7)(a) of the basic regulation precluded it from using, in order to determine the normal value of part of the product types concerned, the main methodology provided for in those provisions and, at the same time, the alternative methodology for the other part, which includes in particular the possibility of using prices in the European Union. Similarly, the Commission is entitled to argue that it could exclude the use of the main methodology only if it could not be applied.

112    However, the wording of Article 2(7)(a) of the basic regulation does not preclude, once the normal value has been obtained by applying the main methodology provided for in those provisions, an adjustment to that normal value being made within the framework of Article 2(10) of that regulation using prices other than the domestic prices of the analogue country or export prices from the analogue country, provided that they result from normal market forces, in particular competitive pressure.

113    It follows from those same provisions that the legislator has not ruled out that ‘the price actually paid or payable in [the European Union] for the like product, duly adjusted if necessary to include a reasonable profit margin’ may serve, under certain conditions, as a reasonable basis for determining normal value. It therefore does not appear to be excluded, a fortiori, that, for the purpose of a fair comparison, a reasonable estimate of the market value of the physical differences can be based, in the absence of other data available, on the difference between the price of the non-matching product type at issue and the average price of the directly comparable product type at one or more Union producers.

114    It follows from all the foregoing that the first part of the third plea is well-founded and is such as to lead to the annulment of the contested regulation.

115    That finding is not called into question by the Commission’s arguments.

116    In the first place, contrary to its argument, the Commission did not comply with the provisions of Article 2(10)(a) of the basic regulation by determining the market value of the physical differences between the product types under consideration on the basis of the export price of the non-matching product types, which is, according to its argument, the price paid in the European Union for that good by the first independent customer.

117    In this respect, in the Commission’s view, the presumption that the market value of a good is reflected in the price paid by the first independent customer is confirmed by the definition of the concept of market value in accordance with the international valuation standards defined by the IVSC. According to that definition, market value is ‘the estimated amount for which an asset or liability should exchange on the valuation date between a willing buyer and a willing seller in an arm’s length transaction, after proper marketing and where the parties had each acted knowledgeably, prudently and without compulsion’.

118    For the reasons set out in paragraphs 70 to 75 above, the fact that the export price of the non-matching product types constitutes the price paid by the first independent customer in the European Union is clearly not sufficient to be considered as a reasonable estimate of the market value. In view of the objective of Article 2(7)(a) and (10) of the basic regulation, that notion does not only imply that the price in question is paid by an independent customer in the context of an arm’s length transaction as defined by the IVSC, but that it must also be possible to ensure that that price is the normal result of market forces. However, this cannot be the case here since that price is likely to be affected by dumping.

119    In the second place, the Commission’s argument that the application of the contested methodology had the effect of reducing the applicant’s dumping margin is not relevant. Assuming that this is the case, it should be noted that that methodology was applied to 28% of the applicant’s export volume, that is, a significant proportion of it. Therefore, it cannot be excluded that, should the estimate of the market value of the physical differences between the like product type and the non-matching product types have been based on a reasonable methodology and in accordance with the applicable provisions, that dumping margin would have been reduced even more significantly.

120    In the third place, the Commission’s argument that, in essence, the lawfulness of the contested methodology must be assessed only in the light of the requirements of Article 2(10)(a) of the basic regulation, and not in the light of the requirements of Article 2(7)(a) or (11) of that regulation, on the ground that that methodology is only in relation to the fair comparison stage cannot be accepted.

121    On the one hand, by definition, the adjustment to the normal value made in accordance with the contested methodology has an impact on the level at which that normal value was determined and, consequently, on the determination of the dumping margin. Consequently, if the application of that methodology has the effect of resulting in a determination of those two parameters which is not in accordance either with the objectives of Article 2(7)(a) of the basic regulation or with those of Article 2(11) of the same regulation, an infringement of those provisions may, or must, be found by the EU judicature.

122    On the other hand, in any event, the applicant invoked an infringement of Article 2(10), ab initio and (a) of the basic regulation. As has been explained in paragraph 77 above, the Commission has not demonstrated that, by applying the contested methodology, it had re-established the symmetry between normal value and the export price of the non-matching product types.

123    In the fourth place, the Commission’s argument that, in the context of the main methodology provided for in Article 2(7)(a) of the basic regulation, it was not obliged to rely on a ‘reasonable basis’ within the meaning of the alternative methodology provided for in that article is manifestly irrelevant. Indeed, as the Court of Justice has found, the reference in the context of the provisions of that article relating to the alternative methodology to ‘any other reasonable basis’ implies a fortiori that, in the context of the main methodology, the basis on which the Commission determines the normal value must be reasonable (see, to that effect, judgment of 22 March 2012, GLS, C‑338/10, EU:C:2012:158, paragraph 25). In addition, as has been repeatedly noted, Article 2(10)(a) requires that the estimate of the market value of the physical differences used for the purpose of the fair comparison be reasonable.

124    In the fifth place, the Commission cannot rely on the fact that it applied the contested methodology following the applicant’s request to take into account, in the context of the dumping margin, export transactions relating to non-matching product types.

125    As was recalled in paragraphs 57 and 108 above, in the provisional regulation, the Commission had simply excluded those transactions from the calculation of the dumping margin. As the applicant points out in the second part of this plea in law, according to the Court of Justice, such an exclusion is contrary to the objective of the different methods of calculating the dumping margin which, in accordance with the objective of Article 2(11) of the basic regulation, is to reflect the full degree of dumping being practised. Thus, according to the Court of Justice, the corollary from such an exclusion is that it is impossible for the Commission to measure the impact that those transactions might have on that calculation, and, therefore, it cannot ensure that the dumping margin calculated reflects the full degree of dumping being practised (judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council, C‑376/15 P and C‑377/15 P, EU:C:2017:269, paragraph 55).

126    Moreover, as the applicant correctly stated at the hearing, it only asked the Commission to take into account transactions relating to non-matching product types when determining the normal value. By contrast, it did not ask the Commission to use the contested methodology to calculate the normal value of those product types. On the contrary, as follows from the Court’s examination of the third complaint raised in the first plea in law of the initial action, before the adoption of Implementing Regulation No 430/2013, the applicant had objected to the adoption of that methodology and had even proposed to use an alternative method (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 123).

127    It follows from all the foregoing that the first part of the third plea must be upheld.

3.      The second part, based on the fact that the Commission infringed Article 2(11) of the basic regulation by adopting a methodology which leads to the de facto exclusion of transactions relating to non-matching product types

128    The applicant claims that, by adopting a methodology that results in an assumption of dumping for the non-matching product types at the same level as that for the directly comparable product types, the dumping margin eventually obtained does not reflect the full degree of dumping being practised, contrary to Article 2(11) of the basic regulation. It asserts that that methodology results in the de facto exclusion of transactions relating to those product types and is based on the erroneous assumption that those transactions had no real impact on the overall dumping margin. In its reply, it states that the Commission’s position on this point contradicts the judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council (C‑376/15 P and C‑377/15 P, EU:C:2017:269).

129    The Commission replies that no product type has been excluded. Moreover, the Commission asserts that, in any event, even if it were found that the methodology applied did exclude de facto non-matching product types, that methodology is not an infringement of Article 2(11) of the basic regulation since it follows from the case-law and Article 2.4.2 of the anti-dumping agreement that non-comparable export transactions may be excluded to ensure a fair comparison.

130    In this respect, on the one hand, it should be noted at the outset that, since this part of the plea is based on the claim that the contested methodology led the Commission to exclude de facto transactions relating to non-matching product types, it can only be rejected.

131    Indeed, even if the contested methodology led the Commission to calculate the same or at least a very similar dumping margin for all those product types, the fact remains that the inclusion of those product types had an impact on the overall dumping margin. In this respect, it should be recalled that, as was explained in paragraph 93 above, in order to obtain the overall dumping margin, the Commission adds the dumping amounts obtained for each product type and thus arrives at the total dumping amount for the product concerned as a whole, which it then relates to the total amount of export transactions made by the applicant. Therefore, even if they corresponded, for each of them, to the same or a very similar dumping margin, the dumping amounts obtained for the non-matching product types, which were added to the dumping amounts for the other product types, necessarily had an impact on the overall dumping margin. It cannot be excluded that, as the Commission maintains, the contested methodology even contributed to lowering the level of that overall dumping margin.

132    However, on the other hand, in so far as this part of the plea amounts to arguing that the contested methodology is contrary to the objective of Article 2(11) of the basic regulation, since the result of the application of that methodology does not reflect the full degree of dumping, it must be upheld on the basis of the reasoning set out in paragraphs 83 to 86 above.

133    It follows from all the foregoing that the third plea must be upheld in its two parts, with the result that the contested regulation must be annulled. It is therefore in principle not necessary to examine the other pleas in law.

134    That being said, in the last part of the application, the applicant requests that the Court examine, on the basis of the principle of effective judicial protection, all of the pleas in law in the action even if it were to find that one of those pleas was sufficient to annul the contested regulation. The applicant repeated that request at the hearing.

135    In this respect, it should be noted that the principle of effective judicial protection in no way implies that the Court should examine all the pleas in law in the action, even if it upholds one of them. As Advocate General Léger explained in point 92 of his Opinion in Acerinox v Commission (C‑57/02 P, EU:C:2004:666), by virtue of the principle that the pleas considered should be kept to a minimum, which applies to proceedings concerning the legality of measures, when the EU court decides to uphold a plea and to annul the contested measure, it is not required to examine the other pleas (see, to that effect, judgment of 6 November 2008, Netherlands v Commission, C‑405/07 P, EU:C:2008:613, paragraphs 38 and 75).

136    Nevertheless, it is for the Court to assess whether, in the interests of the proper administration of justice, all the pleas in law and complaints in the action should be examined.

137    In the present case, it is the second time that the applicant has challenged before the Court the imposition of an anti-dumping duty on imports of the product it manufactures. Moreover, the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), did not examine the pleas in law relating to the internal lawfulness of that anti-dumping duty raised by the applicant at the time and which it repeats in the fourth and fifth parts of the second and fourth pleas in law of the present action. In addition, the first plea in law and the first, second and third parts of the second plea, raised for the first time before the Court, also fall within the internal lawfulness of the contested regulation.

138    It cannot therefore be excluded that the outcome given by the Court to the first, second and fourth pleas in law may provide useful information for the possible adoption of measures under Article 266 TFEU, and may, where appropriate, avoid the need for the applicant to bring a new action before the Court.

139    In those circumstances, the Court considers it appropriate, in the interests of the proper administration of justice, to examine the first, second and fourth pleas in law.

B.      The first plea, alleging infringement of Article 2(7)(a) of the basic regulation, on the grounds of errors made by the Commission in the determination of normal value

140    The first plea, which concerns the determination of the normal value of the product concerned, consists of two parts, regarding (i) the erroneous consideration by the Commission of the analogue country producer’s transactions in very small volumes and (ii) the use of a method for calculating specific cost data based on an unreasonable general assumption.

1.      The first part of the plea, based on the Commission’s incorrect consideration of the analogue country producer’s transactions in very small volumes

141    In the first part of the plea, the applicant claims that the determination of normal value is distorted by the Commission’s consideration of domestic sales of certain product types made in very small volumes by the analogue country producer. According to the applicant, the prices of those sales are not representative and are subject to high volatility. It infers from this that the inclusion of those sales results in an unreasonable determination of the normal value, contrary to Article 2(7)(a) of the basic regulation. In addition, it asserts that, because they are unrepresentative, those sales cannot be considered to have been made in the ordinary course of trade within the meaning of the basic regulation and the case-law. In the reply, the applicant further submits that, contrary to what the Commission claims, its suggestion to exclude from the calculation of normal value the analogue country producer’s sales of 200 kg or less per product type on the domestic market does not amount to applying an absolute threshold, but amounts to applying a specific threshold considered appropriate in relation to the sales and sector concerned.

142    The Commission disputes that line of argument.

143    In the present case, as was noted in paragraph 6 above, during the original investigation, the applicant did not obtain the MET and, consequently, normal value was determined, for the purpose of establishing its dumping margin, on the basis of the domestic prices of the analogue country producer.

144    As was explained in paragraph 130 of the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), the sales of the like product on the Indian market had initially been considered to be insufficiently representative of the applicant’s export sales because their total volume did not reach at least 5% of those export sales. In order to establish a provisional anti-dumping duty on those exports, normal value had therefore been calculated on the basis of constructed value.

145    However, in its observations dated 17 December 2012, the applicant claimed that the Commission should have used the domestic prices of the analogue country producer because the 5% representativeness threshold penalises exporting producers that, like the applicant, supply significant and non-absolute volumes, in accordance with Article 2(2) of the basic regulation (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 29).

146    As is stated in recital 17 of Implementing Regulation No 430/2013, the institutions accepted that objection and it was therefore the domestic sales of the analogue country producer which were used to determine the normal value of directly comparable product types, where those sales were made in the ordinary course of trade. Where the Commission found that this was not the case, normal value was calculated on the basis of constructed value (see recital 62 of the provisional regulation and recital 19 of Implementing Regulation No 430/2013).

147    More specifically, the institutions having identified the different types of like product to be taken into account in the determination of normal value, the weighted average normal value of each type of like product was established and compared to the weighted average export price of the corresponding type of product concerned. For non-matching or ‘quasi-matching’ product types, the methods described in paragraphs 57 to 63 above were used.

148    However, as stated in recital 27 of the contested regulation, in the context of the re-opening of the procedure, the applicant requested that sales of 200 kg or less per product type should not be taken into account on the grounds that they were not reliable. The Commission considered that the prices of those product types had been verified and found to be as reliable as the other prices. Recital 28 of that regulation further specifies that one importer contended, in support of the applicant’s argument, that the price volatility for slow-moving product types was high. The Commission replied that that factor could not justify the exclusion of those product types from the calculation of normal value, since it had established that, as far as the analogue country producer was concerned, price volatility was high for all product types, including the best-selling items.

149    As a preliminary point, it should be noted that, as the applicant stated in a written reply to a question put by the Court, it complains, in the present case, that the Commission calculated the normal value of product types with a transaction volume of 200 kg or less on the basis of the actual prices of the analogue country producer instead of using another methodology.

150    This part is based on two separate complaints. First, the applicant claims that sales prices of 200 kg or less per product type result in an unreasonable determination of the dumping margin, in breach of the requirements of Article 2(7)(a) of the basic regulation. Secondly, it claims that those low-volume sales cannot be considered to have been made in the ordinary course of trade within the meaning of Article 2(3) to (6) of the same regulation.

(a)    The first complaint, based on the unreasonableness of the determination of the dumping margin, due to the fact that prices of product types sold in very low volumes were taken into account

151    In support of the first complaint, the applicant relies on an analysis of the Indian producer’s domestic sales, carried out in the light of the detailed calculations of normal value disclosed to it by the Commission in the context of the re-opening of the anti-dumping proceeding. It invokes three factors which, in its view, lead to the conclusion that the taking into account of the product types for which those domestic sales have the lowest volumes leads to an unreasonable determination of normal value, namely the artificially high prices of such sales, their disproportionate impact on the overall dumping margin of the product concerned and, finally, the high volatility of those prices.

152    It is therefore necessary to determine whether the identification of those three factors is sufficiently substantiated.

(1)    The first argument, based on the artificially high prices of the product types sold in very low volumes

153    The applicant claims that the very low transaction volumes found for certain product types result in ‘artificially high’ prices, which would not be ‘relevant’ or ‘representative’. That claim is mainly supported by the finding of price differences between neighbouring product types sold by the analogue country producer in very different volumes and by the statement of one importer maintaining that, in the economic sector concerned, transactions at such a low volume are not representative.

154    By way of illustration, the applicant compares two product types, whose control numbers start with 0092 and which are sold on the Indian market for a volume of less than 200 kg, with two neighbouring product types, whose control numbers start with 0090 and which are sold in much larger quantities. That comparison shows an average unit price difference of more than 60% between product type 0092FF00GN and product type 0090FF00GN, and of almost 100% between product type 0092II00GN and product type 0090II00GN.

155    That being said, even assuming that the comparison of those product types is relevant, which the Commission disputes, those two isolated examples cannot be sufficient to infer a rule or, at least, a general trend that the volume of transactions in the sector concerned would lead, below a certain quantitative threshold, to the taking into account of artificially high prices, regardless of the product type in question. In particular, It cannot be inferred from this, a fortiori, that 200 kg is a reasonable estimate of such a threshold.

156    Indeed, there is no indication that the significant price differences found in those two examples are explained by the allegedly abnormally low volume of both product types 0092FF00GN and 0092II00GN. In the absence of any other concrete evidence concerning the sales of the analogue producer of a volume of 200 kg or less per product type, nothing precludes other parameters from influencing those differences, such as, in particular, the nature of the two product types in question.

157    In this respect, the Commission noted that product types whose control numbers start with 0092 were more expensive to produce than product types whose control numbers start with 0090, the former having an interior thread and an exterior thread (male/female) and the latter having two exterior threads (female/female). It stated that that fact was confirmed by the applicant’s own list of products, which indicates a difference in price between those two categories of product types of 10 to 40%. The applicant’s arguments in the reply and the clarifications it provided in the context of a written answer to a question put by the Court do not call into question the plausibility of the Commission’s explanations.

158    Moreover, it must be noted that, apart from those isolated examples, the applicant does not refer, in support of its arguments, to the detailed table of normal value calculations, the final version of which was communicated to it on 29 May 2017.

159    The 29 May 2017 table does not corroborate the existence of the factor identified by the applicant. The comparative examination of column C of that table ‘quantity sold on the domestic market (in kg)’ and column N ‘normal value’ does not suggest the existence of a price distribution according to transaction volume and does not indicate, in particular, a systematic distribution of the highest prices among sales with a volume of 200 kg or less per product type.

160    Moreover, the importer’s statement referred to in paragraph 153 above does not further support the applicant’s claim.

161    In this respect, in that statement, that importer argues that it is well-known that purchase prices of products available in small quantities generate significant volatility, when those prices result from an ad hoc agreement and not from a long-term agreement. The importer explains in this respect that those prices often seem high because there are not many suppliers producing them and that, consequently, the level of competition is rather low.

162    It must be stated that that argument seems rather to refer to the low volume of transactions of a given product as a whole on a specific market rather than to a low volume of transactions per specific producer, the importer referring to the small number of producers and to a low level of competition. It is neither alleged nor established that the sales volume of 200 kg or less per product type of the sole cooperating Indian producer reflects the low production volume of the product types at issue on the Indian market as a whole. Such an assumption appears all the more unfounded as the applicant did not dispute the Commission’s finding in recital 52 of the provisional regulation that there were no fewer than 300 producers on the Indian market.

163    Furthermore, the relevance of that argument is also greatly reduced by the fact that, as the importer in question indicates, price volatility of a given product sold in small volumes is found when those prices result from an ad hoc agreement and not from a long-term agreement. It is neither claimed nor established that the prices of the product types sold by the Indian producer at a volume of 200 kg or less would result mainly from ad hoc agreements. In addition, it should be noted that the importer does not indicate whether, in its view, sales of a volume of 200 kg or less per product type should be considered as low or very low volume in the sector in question.

164    Therefore, it follows from the above that the applicant has not provided any conclusive evidence to suggest that the product types sold by the Indian producer at a volume of 200 kg or less result in an ‘artificially high’ price determination. It therefore did not call into question the plausibility of the Commission’s conclusions, set out in recital 27 of the contested regulation, that the prices of those product types, which had been verified as well as the prices of the other product types, could be considered as being as reliable as the latter.

(2)    The second argument, based on the impact of taking into account the product types sold in very low volumes on the increase in the dumping margin

165    The applicant claims that taking those sales into account leads to a significant increase in the dumping margin, whereas they represent only 2% of the total sales volume of the Indian producer. On the basis of the difference between the average dumping margins corresponding to sales of a volume of 200 kg or less per product type and the average dumping margins corresponding to sales of other product types, the applicant considers that such an increase distorts the assessment of the overall dumping margin.

166    In this respect, the applicant relies essentially on two comparative tables which it drew up on the basis of the detailed tables of the calculation of the dumping margin submitted to it by the Commission. The first of those tables shows the difference between the dumping margin corresponding to the sales of the analogue country producer of 200 kg or less per product type and the dumping margin corresponding to the sales of other product types. The second of those tables classifies the dumping margins by product type according to the volume sold.

167    On the basis of the first table, the applicant thus submits that the dumping margin corresponding to sales in the first category represents 250% of the dumping margin in the second category. In addition, it relies on the second of those tables to illustrate that the analogue country producer’s sales of 200 kg or less per product type represent only 2% of the total sales volume of that producer, but nevertheless resulted in an ‘important’ increase in the dumping margin due to the fact that those sales were made in low volume.

168    As a preliminary note, it should be noted that, as confirmed by the applicant in a written answer to a question put by the Court, in drawing up those two tables, it used the data provided by the Commission.

169    However, it should be observed that the mere fact that the inclusion of the analogue country producer’s sales of 200 kg or less per product type induces an increase in the applicant’s overall dumping margin, even if important, cannot in itself reveal a manifest error of assessment by the Commission.

170    As the applicant itself explains in the third plea in law and as the Court has found in paragraphs 81 to 86 above, the Commission cannot assume that the dumping margin is the same for all product types. It is precisely in order to properly reflect the extent of dumping, in accordance with Article 2(11) of the basic regulation, that the Commission compares normal value and export prices by product type, which makes it possible to take into account the differences in the dumping margin according to those types and to attach weight to the respective importance of the product types according to their share in the total amount of dumping.

171    In this respect, it is true that, in a situation where there is a significant difference between the normal value and the export price of certain product types, which is significantly higher than the average, and that export sales of those product types are made in comparatively much larger quantities, taking into account the amount of dumping of those product types is likely to have a significant impact on the increase in the dumping margin.

172    However, the applicant has not provided any evidence to seriously question the fact that, in the present case, those results are the logical outcome of the Commission’s method of calculation by product type, which precisely reflects the extent of the distortion caused by dumping on the market in question.

173    In particular, in the absence of any concrete evidence of the unreliability of the normal value of certain product types, an approach that, as the applicant seems to suggest, would automatically preclude the taking into account of such significant differences between normal value and export prices or calculate normal value in such a way as to obtain an amount of dumping closer to the average would deprive that methodology of its useful effect or significantly weaken it.

174    Indeed, such an interpretation of the methodology by product type would in fact amount to authorising the Commission, in the context of the calculation of normal value, to influence the result of the calculation of the dumping margin, solely because certain results obtained from normal value determined on the basis of actual domestic prices would not be in line with its expectations or those of certain interested parties (see, to that effect and by analogy, judgment of 5 April 2017, Changshu City Standard Parts Factory and Ningbo Jinding Fastener v Council, C‑376/15 P and C‑377/15 P, EU:C:2017:269, paragraph 60).

175    In the present case, in paragraph 164 above, the Court concluded that the applicant had not provided any concrete evidence to support its claim that the sales prices of the analogue country producer of 200 kg or less per product type were unreliable because they were artificially high. In those circumstances, the mere fact that those prices induce significantly higher dumping margins than sales of other product types and lead to an increase in the dumping margin cannot reveal a ‘distortion’ of that dumping margin, justifying that those transactions should not be taken into account.

176    In any event, the tables drawn up by the applicant do not show that the analogue producer’s sales of 200 kg or less per product type contribute to the increase in the overall dumping margin because of their lower volume than sales of other product types. It cannot be deduced from those tables that there is a correlation between (i) the volume of those sales and (ii) the significantly higher dumping margin corresponding to the product types in question.

177    In this respect, as was explained in paragraph 171 above, the factors which could explain the relatively more significant impact of certain product types on the increase in the dumping margin are (i) the magnitude of the difference between the normal value and the export price and (ii) the significant volume of export sales of that product type. By contrast, the sales volume of those same product types on the domestic market does not in itself have any impact on the overall dumping margin.

178    Moreover, the data on the product types sold by the analogue country producer at a volume of 200 kg or less do not confirm the applicant’s underlying assumption that those sales would systematically lead to much higher dumping margins than sales of a higher volume.

179    In this respect, the applicant does not dispute the Commission’s assertion that, of the 93 types of the like product in question, only 40 of them correspond to a dumping margin above the average. In addition, it should be noted that, as results from the 29 May 2017 table, for a similar number of them, the dumping margins are negative.

180    It follows from the above that the fact that the taking into account of the domestic sales of the analogue country producer of 200 kg or less per product type leads to an increase in the dumping margin cannot in itself reflect a distortion of the dumping margin.

(3)    The third argument, based on the volatility of transaction prices in very low volumes

181    The applicant claims that the price volatility in the sector in question is likely to aggravate the unrepresentative nature of low-volume sales and the distortion caused by their taking into account.

182    According to the applicant, which relies in particular on paragraph 11 of the Opinion of Advocate General Van Gerven in Goldstar v Council (C‑105/90, EU:C:1991:425), the volatility of the analogue country producer’s prices requires the average price of each product type to be calculated on the basis of a minimum number of sales, reaching a volume sufficient to make it representative. The applicant claims that the level of price volatility is not the main reason why it considers that sales by the analogue country producer of 200 kg or less per product type should be excluded. However, it submits that that volatility is likely to aggravate the infringement of Article 2(7)(a) of the basic regulation by the Commission, which has acknowledged the existence of such a level of price volatility.

183    It should be recalled that, according to its current meaning, price volatility refers to the magnitude of changes in the price of an asset over time. As confirmed by the applicant in a written answer to a question put by the Court, by using that concept, it relates, in the present case, to the extent of the price variation of the same product type during the period considered.

184    In this respect, it should be noted at the outset that the applicant has not provided any concrete evidence as to the extent of the variation in those prices, although, since the original investigation procedure, it has been in possession of the analogue country producer’s accounting data and detailed data relating to prices on the Indian market (judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 205). It is true that, in recital 28 of the contested regulation, the Commission acknowledged the existence of price volatility for all product types sold by the analogue country producer. However, this is a very general observation and cannot be used as a basis for excluding certain specific product types in the absence of any concrete evidence relating to individual transactions by product type.

185    In particular, it should be noted that the applicant seems to assume that the analogue country producer’s sales of 200 kg or less per product type cover an insufficiently large number of transactions. However, there is nothing in the file to support that assumption. Symmetrically, there is nothing in the file to exclude that some of the product types with the highest sales volumes were the subject of a small number of individual transactions made, for each of them, in significant quantities. In those circumstances, Advocate General Van Gerven’s findings that, because of price volatility, the calculation of the normal value of a product or product type must be based on a minimum number of sales in order to be representative (Opinion of Advocate General Van Gerven in Goldstar v Council, C‑105/90, EU:C:1991:425, point 11) cannot justify excluding sales of 200 kg or less per product type in the present case.

186    Moreover, as the Commission pointed out, in the following point of his Opinion in Goldstar v Council (C‑105/90, EU:C:1991:425), Advocate General Van Gerven nevertheless noted that the setting of an absolute minimum threshold of general scope made little sense because the value of an absolute figure depended too closely on the nature of the product. In addition, the Advocate General added that it was necessary to determine, for each product or model, whether there were sufficient sales on the market concerned at stable prices for the average price calculated on that basis to be sufficiently representative and that, apart from a general assertion to the effect that the number of sales on the domestic market was minimal, the applicant in the case in question had not advanced any arguments demonstrating that the average price applied would not be representative (Opinion of Advocate General Van Gerven in Goldstar v Council, C‑105/90, EU:C:1991:425, point 11).

187    That latter reasoning can be applied in this case. Indeed, apart from general considerations on price volatility and low sales volume per product type of 200 kg or less, the applicant did not put forward any concrete arguments concerning the number of sales at stable prices or the total number of transactions made with regard to those product types.

188    In any event, it should be observed that, as the applicant has extensively explained in its submissions, it does not consider the volatility of the analogue producer’s prices as being the determining reason why those prices cannot be considered as representative, but only as a factor aggravating that lack of representativeness. It follows from paragraphs 154 to 164 above and from paragraphs 169 to 179 above that the applicant has not established the existence of the two factors which it identified, primarily, to justify the exclusion of the sales prices of the analogue country producer per product type of 200 kg or less. Therefore, the third of those factors, a fortiori, cannot justify such an exclusion.

189    It follows from all the above that the first complaint must be dismissed.

190    That conclusion is not called into question by the applicant’s argument that, even if a threshold other than that of 200 kg per product type, as proposed by the applicant, were appropriate, the complaint would still be well-founded, since the Commission failed to examine the distortion caused by taking into account the product types sold by the analogue country producer in the lowest volumes.

191    First, according to the case-law, in the context of a complaint challenging the method of determining normal value, the applicant cannot simply rely on a method of determining normal value alternative to that chosen by the Commission, but must provide sufficient evidence to deprive the assessments on which that choice is based of any plausibility, as the EU judicature cannot substitute its own assessments for them (see, to that effect and by analogy, judgments of 4 March 2010, Foshan City Nanhai Golden Step Industrial v Council, T‑410/06, EU:T:2010:70, paragraph 66, and of 11 September 2014, Gold East Paper and Gold Huasheng Paper v Council, T‑443/11, EU:T:2014:774, paragraph 62).

192    In the present case, it follows from the above that the applicant has not demonstrated that the low absolute volume of the analogue country producer’s domestic sales of 200 kg or less per product type was a factor which could affect the reliability and representativeness of the prices of those sales. Consequently, no manifest error of assessment by the Commission, a fortiori, can be found on the basis of the fact that the Commission did not apply an absolute threshold different from that proposed by the applicant to exclude the product types with the lowest sales volume on the Indian market.

193    Secondly, the applicant is wrong to complain that the Commission failed to examine the distortion of the dumping margin caused by taking into account the product types sold in the lowest volumes.

194    The applicant does not dispute that the Commission verified the prices of all product types and considered them to be reliable. In addition, in recital 28 of the contested regulation, the Commission replied to the importer’s argument regarding price volatility for slow-moving product types.

195    Moreover, in the original investigation procedure, the Commission had examined the question of the representativeness of the domestic prices of the analogue producer in relation to its sales volume, since, as was recalled in paragraph 144 above, it had applied the threshold of 5% of the export sales volume in the present case, which led it to completely exclude the taking into account of those prices in the calculation of normal value and to use the alternative constructed value method.

196    It is true that, as the total domestic sales of the analogue country producer represented less than 5% of the applicant’s export sales (3.9% according to the 29 May 2017 table), it had not been necessary for the Commission to examine the representativeness of each product type.

197    That being said, the applicant, which had expressly requested that the 5% representativeness threshold not be applied, does not claim that the sales of the analogue country producer with the lowest volumes per product type correspond to a proportion of its export sales which is too low. In this respect, in the reply, the applicant emphasised that it continued to consider the 5% criterion to be unreasonable, on the grounds that it was meaningless in the context of an analogue country.

198    Finally, the Commission rightly maintains that it cannot be required to apply an absolute threshold of representativeness. Indeed, as was pointed out in paragraph 186 above, in point 11 of his Opinion in Goldstar v Council (C‑105/90, EU:C:1991:425), Advocate General Van Gerven stated that it was not necessary to set an absolute minimum threshold of general scope and that the Commission had to assess whether, for each product or model, there were sufficient sales at stable prices on the market concerned for the average price calculated on that basis to be representative.

199    Moreover, it follows from paragraphs 18 to 21 of the judgment of 13 February 1992, Goldstar v Council (C‑105/90, EU:C:1992:69) that, in the case giving rise to that judgment, the question of the sufficient representativeness of sales on the domestic market in relation to their volume arose in relation to the size of the market as a whole, the applicant in that case claiming that the Korean market for the product concerned was of insufficient size, since total sales of that product had only amounted to 5 000 units. It is in this context that the Court of Justice had examined whether such an overall volume of sales of the product concerned allowed normal patterns of price formation and had found that this was the case, in particular in view of the stable nature of the domestic prices in question.

200    However, such reasoning cannot be applicable to the question of the representativeness of sales of the like product or a given type of that product made by the analogue country producer on the domestic market, but only, where applicable, to the question of the representativeness of the overall sales volume of that product or product type on the Indian market as a whole.

201    In any event, as was found in paragraph 162 above, it is not claimed that the Indian market is too small for the product types sold by the analogue country producer in the smallest volumes.

202    The first complaint must therefore be rejected.

(b)    The second complaint, alleging that the low volume sales were not made in the ordinary course of trade

203    The applicant submits that the concept of ‘ordinary course of trade’ is not limited to a profitability check. According to the applicant, sales volumes such as those of sales by product type of 200 kg or less are so low that they do not reflect normal behaviour on the part of purchasers and do not result from normal patterns of price formation. It relies in this respect on paragraphs 15 to 18 of the judgment of 13 February 1992, Goldstar v Council (C‑105/90, EU:C:1992:69). Furthermore, it submits that, in the light of the case-law, there is a rebuttable presumption that low-volume sales are not made in the ordinary course of trade and that, in any event, the General Court and the Court of Justice considered that the volume of sales was relevant for the analysis of the ordinary course of trade (judgments of 1 October 2014, Council v Alumina, C‑393/13 P, EU:C:2014:2245, paragraphs 28 to 30, and of 30 April 2013, Alumina v Council, T‑304/11, EU:T:2013:224, paragraphs 24 and 25).

204    Without it being necessary to examine the arguments by which the Commission, in essence, asks the General Court to declare the present complaint inadmissible, it must be held that this complaint has no substantive merit.

205    The present complaint is based on a premiss close to that underlying the first complaint, namely that sales by product type of 200 kg or less are made in volumes which are too small to allow normal price formation, such as that resulting from forces normally applied in a market economy. The applicant bases that premiss on the same elements which it relied on to support its first complaint. However, in paragraphs 153 to 201 above, it has been demonstrated to the requisite legal standard that those elements were not convincing and, in particular, that they were not such as to call into question the reliability of the prices of those sales and to justify the determination of an absolute sales volume threshold below which they should be excluded from the calculation of the normal value. Consequently, irrespective of the fact that the concept of ordinary course of trade is distinct from the concept of representativeness of sales, those elements are also not such as to support the second complaint.

206    The case-law cited by the applicant is not such as to call that conclusion into question.

207    On the one hand, with regard to paragraphs 15 to 18 of the judgment of 13 February 1992, Goldstar v Council (C‑105/90, EU:C:1992:69), it was noted in paragraphs 199 to 201 above that those paragraphs dealt with the specific question of the representativeness of domestic prices in relation to the overall volume sold on the whole market concerned, which is not at issue in the present dispute, as the applicant does not call into question the representativeness of prices on the Indian market in relation to the total sales volume of the relevant product types on that market. Moreover, in those same paragraphs, the Court of Justice does not address the question of whether, in view of their volume, sales can be regarded as having been made in the ordinary course of trade.

208    On the other hand, with regard to paragraphs 28 to 30 of the judgment of 1 October 2014, Council v Alumina (C‑393/13 P, EU:C:2014:2245), and paragraphs 24 and 25 of the judgment of 30 April 2013, Alumina v Council (T‑304/11, EU:T:2013:224), it should be noted that the interpretation of paragraph 25 of the above-mentioned judgment of the General Court, defended by the applicant, according to which it should be presumed that sales made in small volumes cannot be regarded as having taken place in the ordinary course of trade, cannot be accepted.

209    In paragraph 24 of the judgment of 30 April 2013, Alumina v Council (T‑304/11, EU:T:2013:224), the Court found that the volume of domestic sales constituted a factor liable to affect price formation and, accordingly, the two criteria could interact, for example, where the domestic market was so limited that the prices were not determined by supply and demand. In paragraph 25 of the same judgment, the Court noted, however, that that possibility of interaction did not mean that, if the representativeness threshold of 5% was not reached, domestic sales were not to be regarded as having been made in the ordinary course of trade. It is in this context that the Court held that it was not wholly inconceivable that, despite their low volume, domestic sales could be made in the ordinary course of trade if they nevertheless reflected the ordinary behaviour of the operators concerned.

210    In those circumstances, the latter consideration, relied on by the applicant, clearly cannot be interpreted as establishing a presumption that the low absolute volume of domestic sales of a product type indicates that those sales did not take place in the ordinary course of trade.

211    First, in paragraph 25 of the judgment of 30 April 2013, Alumina v Council (T‑304/11, EU:T:2013:224), the Court examined the situation where domestic sales did not reach the relative representativeness threshold of 5%, and not the situation where they did not reach a given volume in absolute terms.

212    Secondly, despite the Court’s use of the expression ‘it is not wholly inconceivable’, it follows from the context surrounding the consideration of paragraph 25 of the judgment of 30 April 2013, Alumina v Council (T‑304/11, EU:T:2013:224), relied on by the applicant, that the Court does not consider as an exceptional situation the fact that the low volume of domestic sales may reflect the ordinary behaviour of the operators concerned. On the contrary, it expressly follows from the consideration of the same paragraph 25, which precedes the consideration invoked and is recalled in paragraph 209 above, that the fact that the volume of domestic sales is below the 5% representativeness threshold does not mean that they can be regarded, on that ground alone, not to have been made in the ordinary course of trade.

213    In addition, in paragraphs 28 and 30 of the judgment of 1 October 2014, Council v Alumina (C‑393/13 P, EU:C:2014:2245), cited by the applicant, the Court of Justice merely noted that, with regard to the purposes of the concept of ‘ordinary course of trade’, the question of whether or not sales had been made in the course of such transactions depended not only on the price, but also on the other conditions of the transaction likely to influence it, such as the volume, the additional obligations assumed by the parties or the delivery periods. The fact that the volume of a transaction may be a relevant factor in assessing whether it was carried out in the ordinary course of trade does not mean that transactions carried out at a volume considered to be low in absolute terms should ipso facto be considered as not having taken place in the ordinary course of trade.

214    It follows from all the foregoing that the second complaint and, consequently, the first part of the first plea in law as a whole must be dismissed.

2.      The second part, alleging the use of a methodology for calculating specific cost data based on an unreasonable general presumption

215    By the second part, the applicant asserts that the methodology developed by the Commission to calculate product-type-specific cost data, in the absence of the provision of reliable data in this regard by the analogue country producer, is based on a general presumption of the profitability of individual transactions from the threshold of 92.14% of the average price of the product type concerned. It considers that such a general presumption is unreasonable and leads to an arbitrary exclusion of certain sales transactions and to an unjustified increase of the normal value, in infringement of Article 2(7)(a) of the basic regulation. It adds that Article 2(5) of that regulation does not apply to a normal value determination using the analogue country methodology and that, in any event, the Commission is not obliged in that case to carry out a profitability check such as that provided for by Article 2(4) of the basic regulation.

216    The Commission disputes that line of argument.

217    In this respect, it should be noted that the present part of the first plea concerns (i) an error of law by the Commission in that it wrongly considered that Article 2(5) of the basic regulation was applicable in the context of the market economy third country methodology provided for in Article 2(7)(a) of that regulation and (ii) a manifest error of assessment, in that the application of the provisions of the first of those articles would have led to an arbitrary exclusion of certain transactions, on the basis of a general presumption of profitability of sales above a specified threshold, which was an unreasonable presumption.

(a)    The error of law, in that the Commission misapplied Article 2(5) of the basic regulation

218    First of all, it should be observed that, in the context of Article 2(7)(a) of the basic regulation, which provides for a set of rules for the determination of normal value distinct from that established by Article 2(1) to (6) of the basic regulation, the Commission’s discretion is not expressly limited by requirements similar to those defined in the latter paragraphs (judgments of 27 October 2011, Dongguan Nanzha Leco Stationery v Council, C‑511/09 P, EU:C:2011:696, paragraphs 33 and 34, and of 10 September 2015, Fliesen-Zentrum Deutschland, C‑687/13, EU:C:2015:573, paragraph 65). That being said, as was recalled in paragraph 51 above, when the Commission chooses to determine normal value on the basis of domestic prices in a third country, it follows from the case-law that it has an obligation to ensure that the values and parameters on which it relies are the normal result of market forces.

219    In those circumstances, contrary to what the applicant claims, there is no reason why, in the context of that methodology of determining normal value, the Commission should depart from the principles governing the methodology of the price in the country of origin of exports, as set out in Article 2(1) to (6) of the basic regulation. Those principles aim, in particular, to provide a framework for verifying the reliability of prices used as a basis for determining normal value and the conditions under which, where such reliability cannot be established, the Commission must use the constructed value methodology.

220    Furthermore, the nature and objectives of the normal value determination methodology do not fundamentally change depending on whether that methodology is based on Article 2(1) to (6) or (7)(a) of the basic regulation; that methodology is intended to enable, in both cases, the price level of the product concerned to be determined as it would result from normal market forces in the absence of dumping.

221    Those considerations are not called into question by the provisions of Article 2(7)(b) of the basic regulation, relied on by the applicant, according to which normal value is determined in accordance with paragraphs 1 to 6 of that article, where the producers in question demonstrate that market economy conditions prevail as far as they are concerned. Indeed, those provisions cover cases where, by way of exception to Article 2(7)(a) of the basic regulation, normal value can be determined in relation to data from the country of origin of exports and not in relation to data from a market economy third country. However, it cannot be deduced from those provisions that, where the latter data are used, the methodological principles set out in paragraphs 1 to 6 of the same article with regard to the use of data from the country of origin of exports would not be transposable.

222    Moreover, the applicant’s position adopted in the present part of the first plea is difficult to reconcile with the fact that, in the context of the second complaint in the first part of the first plea, it applies, in the present case, the concept of ‘ordinary course of trade’, invoking case-law relating to the methodology of the price in the country of origin of exports (see paragraphs 203 to 213 above).

223    In the present case, as regards the method of allocating costs, the Commission applied the second sentence of the third subparagraph of Article 2(5) of the basic regulation.

224    In that regard, the third subparagraph of Article 2(5) of the basic regulation provides:

‘Consideration shall be given to evidence submitted on the proper allocation of costs, provided that it is shown that such allocations have been historically utilised. In the absence of a more appropriate method, preference shall be given to the allocation of costs on the basis of turnover. Unless already reflected in the cost allocations under this subparagraph, costs shall be adjusted appropriately for those non-recurring items of cost which benefit future and/or current production.’

225    It is common ground between the parties that the analogue country producer did not provide the Commission with reliable product-type-specific cost data and that therefore only data relating to overall costs and overall profitability of the like product were usable.

226    There was therefore no reason why the methodology referred to in the second sentence of the third subparagraph of Article 2(5) of the basic regulation should not be applied by analogy in the present case, since conditions laid down by the legislature for such application were met.

227    Indeed, it follows from the wording of the first sentence of the third subparagraph of Article 2(5) of the basic regulation that it is for the Commission to give consideration to evidence submitted to it on the proper allocation of costs, provided that it is shown that such allocations have been historically utilised. However, the second sentence of that paragraph specifies that, in the absence of a more appropriate method, preference is to be given to the allocation of costs on the basis of turnover.

228    In the present case, the Commission was therefore in a situation covered by the second sentence of the third subparagraph of Article 2(5) of the basic regulation and therefore did not err in law in applying those provisions.

229    However, the applicant submits that, even if the provisions of the second sentence of the third subparagraph of Article 2(5) of the basic regulation were to apply in this case, contrary to what it claims, the Commission was not required to carry out a profitability check on transactions such as that referred to in Article 2(4) of the basic regulation.

230    In this respect, it should be noted that it follows from Article 2(4) of the basic regulation, and in particular the third subparagraph thereof, that when it is established that, over a period of at least 6 months, the weighted average selling price is below the weighted average unit cost, or that sales below the unit cost represent at least 20% of sales being used to determine normal value, the Commission may exclude such sales from the determination of normal value on the ground that they were not made in the ordinary course of trade.

231    As the Commission explains in recital 26 of the contested regulation, it consistently carries out the ordinary course of trade test in its investigations.

232    Thus, in this case, on the basis of a turnover-based allocation of costs, in accordance with the second sentence of the third subparagraph of Article 2(5) of the basic regulation, the Commission carried out a profitability check on the individual transactions of the analogue country producer by product type on the basis of the third subparagraph of Article 2(4) and consequently excluded from the calculation of normal value those transactions which did not meet the criteria defined by those provisions.

233    On the one hand, nothing in the wording of Article 2(7)(a) of the basic regulation or in the wording of any of its other provisions prevents the Commission from carrying out such a profitability check in the context of determining normal value on the basis of prices in a market economy third country.

234    On the other hand, even if the wording of Article 2(4) of the basic regulation does not provide for an express obligation to carry out such a profitability check, the applicant does not dispute that the Commission has carried it out in this case in accordance with the legal conditions laid down by those provisions. In particular, the applicant does not claim that the Commission excluded individual transactions where, for a given product type, the proportion of profitable sales exceeded 80% of the total sales volume of that product type and the weighted average price was above the cost of production.

235    It follows from the above that the application in this case of the second sentence of the third subparagraph of Article 2(5) in conjunction with the third subparagraph of Article 2(4) of the basic regulation is not vitiated by any error of law.

(b)    The manifest error of assessment, in that the profitability check carried out by the Commission on the basis of the allocation of costs based on turnover would have led to the unreasonable determination of a general profitability threshold and the arbitrary exclusion of certain transactions

236    The applicant submits, first of all, that the approach favoured by the Commission leads to the automatic exclusion of groups of individual transactions by product type as unprofitable, as they were below 92.14% of the average price of the product type in question, instead of checking the profitability for each product type. Such an exclusion would be arbitrary and would lead to an unjustified increase in normal value.

237    In this respect, the applicant explains that, on the basis of the analogue country producer’s data on overall cost and turnover, the Commission determined that the total manufacturing cost of the like product represented 79.5% of turnover and that selling, general and administrative expenses (SG&A expenses) represented 15.9% of the total manufacturing costs, bringing the total cost of production to 92.14% of turnover. According to the applicant, the Commission applied that figure of 92.14% to the average selling price of each of the product types and determined the average unit cost of the product type concerned on the basis of that figure. Where the price of individual transactions of that product type was below that 92.14% threshold, the Commission excluded those transactions and calculated the average price of the relevant product type without taking into account those transactions.

238    In this respect, it should be noted at the outset that the applicant does not dispute that the figures used by the Commission correspond exactly to the overall cost and turnover data for the like product provided by the analogue country producer. Furthermore, the applicant does not dispute that those data are reliable and that, by contrast, the other data submitted by the analogue country producer to the Commission did not allow a unit cost per product type to be determined. In those circumstances, in so far as, moreover, the Commission has correctly applied the conditions set out in the second sentence of the third subparagraph of Article 2(5) and the third subparagraph of Article 2(4) of the basic regulation, a manifest error of assessment cannot be found in respect of a result which is solely the consequence of the application of those provisions to reliable data.

239    The fact that the application of that methodology results in an increase in normal value cannot, in itself, constitute an indication of the unreasonableness or arbitrariness of applying a single profitability threshold to all product types. This is precisely the objective of the application of a profitability check, which aims to eliminate from the calculation of normal value unprofitable transactions leading to an artificial decrease in that normal value. Furthermore, in accordance with the wording of Article 2(4) of the basic regulation, the Commission is entitled to consider that, under certain conditions, sales made at a price below the unit production cost did not take place in the ordinary course of trade, which the applicant does not dispute.

240    In any event, as the Commission essentially states, the applicant does not demonstrate in any way that the alternative methodology which it seems to suggest, namely a methodology consisting in not excluding transactions below the profitability threshold determined by the Commission, is likely to result in a reasonable determination, unlike the methodology it applied.

241    Such an alternative methodology is implicitly, but necessarily, based on the presumption that all individual transactions relating to the like product are profitable. The applicant has not provided any concrete evidence to justify that presumption.

242    Moreover, it is true that the methodology adopted by the Commission only makes it possible to arrive at an approximate value in terms of the profitability of individual transactions. However, it is nevertheless the methodology which it must use in principle where there is no other appropriate possibility of allocating costs, in accordance with the second sentence of the third subparagraph of Article 2(5) of the basic regulation.

243    Next, the applicant’s argument that, in essence, the purpose of monitoring profitability by product type is precisely to take into account the significant differences in profit margins between those product types, as evidenced by the comments of several interested parties, can only be rejected, since the applicant does not dispute that the Commission could not have obtained reliable data on the costs of the analogue producer by product type.

244    Lastly, the elements relied on by the applicant relating to the specificity of the sales and price structure of the analogue country producer are not more persuasive.

245    First, the applicant does not explain in any way how the fact that the analogue country producer sold only small volumes per product type and per customer would justify the Commission’s departure in the present case from its usual practice of carrying out a profitability check. In particular, the applicant does not specify the nature of the impact of those small quantities per product type and per customer on the prices of the analogue producer. Moreover, that line of argument is somewhat inconsistent with that made in the first part of this plea, since, in the context of that part, the applicant seemed to consider that low volumes induced artificially high prices.

246    Secondly, the existence of high price volatility does not seem likely to call into question the justified nature of a profitability check in the light of a single average profit margin for all product types. The existence of volatility in the prices of individual transactions by product type militates in favour of applying the Commission’s methodology, since it allows, at the very least, the exclusion of transactions at abnormally low prices compared to the average unit price of the product type concerned. In any event, the alternative methodology recommended by the applicant would lead to a result that would not be reasonable because it would have no moderating effect on the volatility of individual transactions.

247    It follows from the foregoing that the applicant has not provided any evidence of the unreasonableness or arbitrariness of the methodology adopted by the Commission and, consequently, of the existence of a manifest error of assessment in the choice of that methodology.

248    The second part of the first plea and, accordingly, the first plea in its entirety must therefore be rejected.

C.      The second plea, alleging infringement of Article 2(10) of the basic regulation and Article 2.4 of the anti-dumping agreement, by reason of the Commission’s rejection of the applicant’s requests for adjustment

249    By its second plea, the applicant alleges that the Commission rejected, in part or in whole, its requests for adjustment in order to ensure a fair comparison between the export price and the normal value within the meaning of Article 2(10) and Article 2.4 of the anti-dumping agreement. This plea in law comprises five parts. The first to fourth parts relate to the Commission’s rejection of the applicant’s requests for adjustments relating to level of trade, packing costs, credit conditions and differences in raw material and productivity respectively. The fifth part relates to the unreasonableness of the burden of proof imposed on it by the Commission.

250    As a preliminary remark, it should be recalled that Article 2(10) of the basic regulation implements Article 2.4 of the anti-dumping agreement, the provisions of which it essentially reiterates.

251    However, in the light of the case-law, Article 2.4 of the anti-dumping agreement does not in principle constitute a standard against which the EU judicature may directly review the lawfulness of an anti-dumping regulation, but only incidentally in the context of a plea of illegality challenging the validity of the basic regulation (see, to that effect, judgments of 7 May 1991, Nakajima v Council, C‑69/89, EU:C:1991:186, paragraphs 27 and 28, and of 27 September 2007, Ikea Wholesale, C‑351/04, EU:C:2007:547, paragraphs 30 to 31). It must therefore be considered that the applicant is in fact invoking an infringement of Article 2(10) of the basic regulation, interpreted in the light of Article 2.4 of the anti-dumping agreement (see, to that effect, judgments of 14 May 2009, Internationaal Verhuis- en Transportbedrijf Jan de Lely, C‑161/08, EU:C:2009:308, paragraph 38 and the case-law cited, and of 10 November 2011, X and X BV, C‑319/10 and C‑320/10, not published, EU:C:2011:720, paragraph 44 and the case-law cited).

252    As recalled in paragraph 48 above, it is apparent from both the wording and the scheme of Article 2(10) of the basic regulation that an adjustment to the export price or the normal value may be made only in order to take account of differences in factors which affect the prices and therefore their comparability. Moreover, as was pointed out in paragraphs 50 to 53 above, if the Commission has a significant discretion in the application of those provisions, with particular regard to the vague notion of fairness, it is for the EU judicature to verify whether it has not failed to take essential elements into consideration and whether the material in the file has been examined with all due diligence.

253    In addition, it is settled case-law that, if a party requests adjustments under Article 2(10) of the basic regulation in order to make the normal value and the export price comparable for the purposes of determining the dumping margin, it must prove that its request is justified. Thus, where a producer claims the application of an adjustment, it is for that operator to indicate and demonstrate that the conditions for granting such an adjustment are met. Conversely, where the Commission considers that an adjustment should be applied, it should at least report converging evidence that that condition is met (see, to that effect, judgment of 16 February 2012, Council and Commission v Interpipe Niko Tube and Interpipe NTRP, C‑191/09 P and C‑200/09 P, EU:C:2012:78, paragraphs 58 and 61 and the case-law cited).

254    It is in the light of those considerations that the Court must examine the five parts of the present plea in law.

1.      The first part, based on the Commission’s rejection of the applicant’s requests for adjustment in respect of level of trade

255    The applicant submits that all of its sales were made to traders whereas the analogue country producer sold both to end-users and to traders. It states that it submitted abundant evidence showing that this led to a consistent and distinct price difference.

256    The Commission disputes that line of argument.

257    According to Article 2(10)(d)(i) of the basic regulation, an adjustment for differences in levels of trade, including any differences which may arise in OEM (Original Equipment Manufacturer) sales, is to be made where, in relation to the distribution chain in both markets, it is shown that the export price, including a constructed export price, is at a different level of trade from the normal value and the difference has affected price comparability which is demonstrated by consistent and distinct differences in functions and prices of the seller for the different levels of trade in the domestic market of the exporting country. The amount of the adjustment is to be based on the market value of the difference.

258    In the present case, in recitals 49 and 50 of the contested regulation, the Commission explained the reasons why it refused to make a level of trade adjustment. The Commission considered that a difference in average prices with regard to different categories of customers, relied upon by the applicant, did not amount to the existence of consistent and distinct differences of prices of the seller for the different levels of trade within the meaning of Article 2(10)(d)(i) of the basic regulation. However, it argued that its analysis of the information contained showed that all types of customers were found in all segments of the price range.

259    In the first place, it should be noted that, in support of the present request for adjustment, the applicant submitted a statement from the analogue country producer, made in an email dated 5 January 2017, in which the latter indicated that it applied ‘special prices for end user which [were] normally 15/20% more than the price being offered to trader/reseller’ and that it also granted a ‘3 to 5% special discount on minimum turnover of USD 40 000/50 000 [(between approximately EUR 35 000 and EUR 44 000)] for 1 year to traders/resellers only’.

260    However, as the Commission indicates and as appears from the material in the file, in its reply to the anti-dumping questionnaire, the analogue country producer indicated that there was no price difference between the prices set for end-users and the prices set for traders.

261    In those circumstances, it is difficult to blame the Commission for not having taken into account the analogue country producer’s statement produced by the applicant.

262    The content of the statement relied on by the applicant, which was made almost 5 years after the original anti-dumping investigation, does not make it possible to determine whether it concerns the usual practice of that producer on the date on which that statement was expressed or whether it is also relevant for transactions made during the period considered for that investigation, namely 2011. However, it is plausible that, given the time elapsed since that period, the practices of the analogue country producer have changed.

263    Moreover, according to settled case-law, in order to assess the probative value of a document, regard should be had to the credibility of the account it contains and, in particular, to the person from whom the document originates, the circumstances in which it came into being, the person to whom it was addressed and whether, on its face, the document appears to be sound and reliable (see judgment of 14 April 2011, Visa Europe and Visa International Service v Commission, T‑461/07, EU:T:2011:181, paragraph 182 and the case-law cited).

264    In the present case, it should be observed that the statement relied on by the applicant is hardly compatible with the statement made by the same analogue country producer in response to a question from the Commission in 2012, referred to in paragraph 260 above.

265    It must be noted that that statement was produced in the context of an exchange of emails with the applicant, which is a regular business partner of that producer, and at the applicant’s request. Moreover, it appears from the documents in the file that, in the applicant’s e-mail of 30 December 2016, which the statement relied on follows on from, the context of the re-opening of the anti-dumping investigation was expressly mentioned. In addition, in that email, after having asked the producer if it had a ‘different pricing policy for end users and retailers than for wholesalers’, the applicant observed: ‘because, as you know, sales to users and retailers may normally include additional services and charges’.

266    Therefore, while the links between the applicant and the analogue country producer cannot deprive that producer’s statement to which it refers of any value (see, to that effect and by analogy, judgment of 13 June 2012, Süd-Chemie v OHIM — Byk-Cera (CERATIX), T‑312/11, not published, EU:T:2012:296, paragraph 30 and the case-law cited), the context in which it was produced justifies it being given a lower probative value than that given to the statement made by the same analogue country producer in the context of the anti-dumping investigation.

267    In any event, the fact that the analogue country producer generally has a different pricing policy according to the categories of customers for whom those products are intended does not necessarily result in practice in consistent and distinct differences in the prices used by the Commission within the meaning of Article 2(10)(d)(i) of the basic regulation. Moreover, it should be noted that, in the present case, the prices used by the Commission correspond only to the product types which the analogue country producer produces itself and which it markets on the Indian market. Thus, as is apparent from the documents in the file, those product types constitute only part of the activity of that analogue country producer, which, in addition, sells products imported from China on the same market and also exports some of its products.

268    In the second place, in its comments on the final disclosure, the applicant submitted the results of an analysis covering, in its view, the 131 product types sold in quantities exceeding 200 kg by the analogue country producer, representing 98% of that producer’s sales volume. According to those results, the weighted average price per product type was always higher for end-users than the weighted average price of those product types applied for traders, except for five product types. In particular, the applicant provided figures for three product types representing around 25% of the volume of the analogue country producer’s sales.

269    Furthermore, the applicant points out that the Commission rejected that analysis on the basis of a comparison of the minimum and maximum prices charged to end-users and traders for the three most important product types. It considers that that comparison does not provide any useful information due to the volatility of the prices of the analogue country producer as recognised by the Commission itself.

270    In this respect, first of all, it should be noted that, as confirmed by the applicant in a written reply to a question put by the Court, although it claimed to have carried out an analysis of the 131 product types sold in volumes exceeding 200 kg, it provided the Commission with the data underlying its analysis only with regard to three product types representing 25% of the total domestic sales volume of the analogue country producer. Even if those data were conclusive for those three product types, the Commission did not therefore make a manifest error of assessment by considering that they did not reflect consistent and distinct price differences, within the meaning of Article 2(10)(d)(i) of the basic regulation, for all product types sold by the analogue country producer.

271    Moreover, while the applicant states, without being challenged, that the 131 product types it claims to have analysed represent 98% of the domestic sales volume of the like product of the analogue country producer, it was noted in paragraph 177 above that the sales volume of the like product on the domestic market does not in itself have any impact on the overall dumping margin. By contrast, as the applicant itself claimed in the first part of the first plea, although representing only 2% of the total volume of those domestic sales, the prices of the 93 like product types, each of which did not exceed 200 kg in sales volume, may have had a significant impact on that overall dumping margin. It follows from paragraph 268 above that the applicant’s analysis with regard to price differences linked to level of trade did not cover the latter product types.

272    Next, as was recalled in paragraph 253 above, where a producer claims the application of an adjustment, it is for that operator to demonstrate that the conditions for granting such an adjustment are met. It is therefore for the applicant in the present case to prove that the evidence it had provided to the Commission in support of its request for an adjustment in this case was such as to demonstrate that the conditions for granting it were met. In the context of this complaint, it cannot therefore rely on what the Commission has not demonstrated to the contrary.

273    Finally, it should be noted that, contrary to what the applicant claims, the comparison of the minimum and maximum prices charged to end-users and traders for the three most important product types on which the Commission based its rejection of its request for a level of trade adjustment may provide a useful indication. As the Commission noted in recital 50 of the contested regulation, it can be deduced from this that all types of customers are found in all segments of the price range.

274    Since the evidence provided by the applicant also related only to the same three product types analysed by the Commission and was limited to the finding of a different average price according to the category of customers concerned, it was not necessary for the Commission to further deepen its analysis. Such an average price alone cannot reveal distinct and consistent price differences in level of trade without an analysis of individual transactions. The fact that all types of customers are found in all segments of the price range, as the Commission has found, only confirms the limited scope of that average price parameter.

275    Far from corroborating the applicant’s argument, the alleged volatility of the analogue country producer’s prices is a factor which may weaken the relevance of the elements relating to the average price differences according to the level of trade which it had provided to the Commission. Indeed, in view of such volatility, there is no reason to conclude that those average price differences reflect a general trend which can be observed for most of the individual transactions considered and not a trend concerning a small number of such transactions or customers of the analogue country producer, but which, by the volume and extent of the price differences, are likely to influence the average price of those product types.

276    It follows from all the above that the information provided by the applicant in the context of the anti-dumping investigation is not sufficient to demonstrate that the conditions of Article 2(10)(d)(i) of the basic regulation were met for the application of a normal value adjustment for the level of trade. The additional elements it produced in the context of this action, in particular as an annex to a written response to a question put by the Court, cannot call that conclusion into question. According to settled case-law, the lawfulness of an EU measure must, in principle, be assessed on the basis of the facts and the law as they stood at the time when the measure was adopted (see judgment of 30 June 2016, Jinan Meide Casting v Council, T‑424/13, EU:T:2016:378, paragraph 150 and the case-law cited).

277    The first part of the second plea must therefore be rejected.

2.      The second part, based on the Commission’s rejection of the applicant’s request for adjustment in respect of packing costs

278    The applicant asserts that it submitted evidence that the value of the adjustment for packing costs was erroneous as a result of the use of a wrong allocation key. It maintains that the analogue country producer only packaged the products it produced itself and not the traded goods. However, the Commission allocated the total packing cost over the total turnover rather than over the turnover of the products produced by the analogue country producer itself.

279    The Commission disputes that line of argument.

280    According to Article 2(10)(f) of the basic regulation, an adjustment is to be made for differences in the directly related packing costs for the product concerned.

281    In the present case, it follows from recital 52 of the contested regulation that the Commission considered that the applicant’s argument concerning the incorrect allocation of packing costs and the low and unreasonable nature of the resulting allowance was not sufficiently substantiated and therefore maintained its methodology concerning the allowance for packing costs.

282    In this respect, it appears that the applicant’s request for a recalculation of packing costs is based on the premiss that, since the analogue country producer produces the like product itself, the rate corresponding to those costs determined by the Commission is not appropriate since it is calculated on the total turnover of the analogue country producer, which includes, in large part, sales of products which the analogue country producer does not produce itself and therefore does not have to package. According to the Commission’s unchallenged contentions, the redistribution requested by the applicant would reduce those costs from a rate of 0.52% to 2.68%, which is eight times higher, which should lead to a reduction in normal value and a comparatively lower dumping margin.

283    That being said, the relevance of the calculation requested by the applicant is not demonstrated.

284    Admittedly, the calculation proposed by the applicant is likely to result in a figure closer to the relative value of the packing costs that the analogue country producer actually bears in proportion to the sales of the like product.

285    However, the rate adopted by the Commission corresponds to the proportion of those costs in relation to the overall turnover, in particular because, as the applicant claims, the analogue country producer does not bear any packing costs on a large part of its sales.

286    Thus, the rate adopted by the Commission reflects the fact that the analogue country producer is able to offset the packing costs of the products it produces itself with the savings it makes on the products it sells and does not pack itself. As a result, the analogue country producer has the possibility not to pass on the full packing costs associated with sales of the like product to the price of the like product.

287    The applicant therefore fails to demonstrate that the impact of packing costs on the sales price of the like product is likely to be reflected by the alternative calculation it proposes but not by the calculation adopted by the Commission. Apart from the information referred to in paragraph 278 above, the applicant did not submit any evidence suggesting that the analogue country producer systematically passed on the packing costs of the products it produced itself to the price of those products.

288    Consequently, the Commission’s decision not to change the calculation of the adjustment for packing costs is not vitiated by a manifest error of assessment and the second part of the second plea can only be rejected.

3.      The third part, based on the Commission’s rejection of the applicant’s request for adjustment in respect of credit costs

289    The applicant submits that the Commission’s initial position of not making an adjustment in respect of credit conditions was contradicted by the evidence in the file and that the Commission did not draw the necessary inferences by merely making an adjustment for this purpose for a single specific customer.

290    The Commission disputes that line of argument.

291    Article 2(10)(g) of the basic regulation provides that an adjustment is made on the basis of the differences in credit costs for the relevant sales, provided that that factor is taken into account in determining the prices charged.

292    In the present case, in recitals 30 and 31 of the contested regulation, the Commission indicated that, following the applicant’s requests for an adjustment, it had accepted, in particular, the argument concerning credit costs, since those costs were related to one single customer of the analogue country producer and that it had therefore made the corresponding adjustment.

293    In recital 51 of the contested regulation, the Commission explained why it had maintained that approach despite the applicant’s new comments. First of all, the Commission indicated that, in respect of credit costs, an allowance had been deducted on the basis of the information on the sales documentation, both on the export price and normal value side. Next, the Commission stated that the applicant had argued that the allowance for credit costs on the normal value side should be made on the basis of the actual number of days between the invoice and the payment, since the Commission allegedly deducted an allowance for credit cost on export sales even though no payment term had been made on the sales documentation. The Commission indicated that no evidence in the investigation file had corroborated that argument and that the applicant had also not been able to provide any supporting evidence. In addition, according to the Commission, it had been established that, both on the normal value side and on the export price side, the actual payment was often made after the payment term on the sales documentation. The Commission inferred from this that calculating the allowance in respect of credit costs on the basis of the actual payment date on the normal value side and on the basis of the information contained in the sales documentation for the export price side would therefore not result in a fair comparison.

294    For its part, as set out in the application, the applicant complains that the Commission applied an adjustment for credit costs only in respect of sales to one single customer of the analogue country producer on the basis of an order form submitted by the latter during the on-the-spot verification and showing a payment term of 30 days. According to the applicant, that producer’s ledgers of receivables, which are among the documents obtained during that verification, show an average number of days of credit of 38 days. It therefore concludes from this that an objective and reasonable assessment of the evidence submitted to it should have led the Commission to make an adjustment to all domestic sales of the analogue country producer. The applicant was unable to demonstrate that the sales documentation of the analogue country producer mentioned credit conditions for transactions other than those of the customer for which an order form had been submitted by the analogue country producer. However, it considers that the combination of that evidence and the mention of credit days in the producer’s ledgers of receivables should have been sufficient for the Commission to accept its request for an adjustment. At the very least, the Commission could easily contact the analogue country producer or indicate to the applicant the precise elements it needed.

295    It must be noted that, as the Commission states in the rejoinder, the applicant’s argument is based on a confusion between the sales documentation of the analogue country producer indicating that that producer granted one of its customers, under the commercial conditions governing the transactions in question, a credit of 30 days, and the average number of days elapsing before payment by the customer, which the applicant calculated as 38 days.

296    As the Commission observes, the applicant does not dispute that, in commercial relations, the actual payment term generally exceeds the payment term indicated on the invoice or contract.

297    Moreover, as the Commission rightly argues, Article 2(10)(g) of the basic regulation covers only cases where a credit is granted by the producer to its customers and where the cost of that credit is taken into account in the determination of the prices charged. However, those provisions cannot cover cases where it is only established that customers have observed a certain time limit before paying. There is no evidence, in those cases, that those payment terms, which were not necessarily provided for in the sales contract or invoice, had an influence on the price determination.

298    In addition, the applicant does not establish that, with regard to credit costs applying to export prices, the Commission adopted a different methodology as that adopted for the determination of credit costs with regard to the price of the like product. In those circumstances, as the Commission notes, accepting the applicant’s request for an adjustment would amount to creating an imbalance capable of affecting the comparability between those prices, contrary to the wording of Article 2(1)(g) of the basic regulation.

299    Therefore, the Commission’s obligation to carry out an objective and reasonable assessment of the evidence submitted to it by the applicant could not have the effect of causing it to apply a credit cost adjustment to all transactions concerning the like product. The Commission therefore did not commit any manifest error of assessment in rejecting the applicant’s request.

300    In view of all the foregoing, the third part of the second plea cannot be upheld.

4.      The fourth part, based on the Commission’s rejection of the applicant’s requests for adjustment in respect of raw materials used and productivity

301    By the fourth part of the second plea, the applicant complains that the Commission rejected any adjustments for differences between the raw materials used and productivity, even though it had recognised that there were such differences. In particular, it asserts that the Commission ignored the statements by the analogue country producer itself, which demonstrated that those differences had an impact on price comparability.

302    The Commission disputes that line of argument.

303    As a preliminary remark, it should be noted that Article 2(10) of the basic regulation does not expressly provide for the possibility of making an adjustment for differences in raw material use and productivity. However, Article 2(10)(k) of the basic regulation provides that an adjustment may be made for differences in other factors not provided for in points (a) to (j) if it is demonstrated that they affect the price comparability, and in particular that customers consistently pay different prices on the domestic market because of the differences in such factors.

304    It therefore cannot be excluded that those provisions may be applied in cases where it is demonstrated that differences in raw material use and productivity affect price comparability, in particular that buyers systematically pay different prices on the domestic market because of them.

305    That being said, it follows from the wording of Article 2(10)(k) of the basic regulation, interpreted in the light of the general scheme of Article 2(10), that the requirement to demonstrate a link between the alleged difference and prices is all the more compelling in such a case. If such a factor was not expressly mentioned in Article 2(10)(a) to (j) of the basic regulation, it is because the legislator did not consider that it was one of the factors which are, in most cases, likely to have a direct influence on prices and, consequently, on their comparability, but it cannot be excluded that this may be the case in certain particular circumstances.

306    Furthermore, in recital 32 of the contested regulation, the Commission noted that the applicant had stated that the lower productivity per worker and the use of different raw materials led to a moderate cost increase of the analogue country producer compared to its own. It stated that, while some differences in efficiency or productivity might exist between companies, the guiding principle of anti-dumping calculations was to ensure comparability between export prices and normal value, which did not require that the circumstances of an analogue country producer and an exporter in a non-market economy country were completely aligned. Indeed, only differences for factors affecting prices and price comparability between an analogue country producer and an exporter in a non-market economy country warranted an adjustment.

307    Thus, without disputing the existence of differences in efficiency or productivity, the Commission considers that it has not been demonstrated that those differences had an influence on prices and comparability. It is necessary to determine whether the evidence submitted by the applicant in support of this part is such as to call into question the plausibility of that assessment.

308    As a preliminary point, it should be noted that, while, in recital 32 of the contested regulation, the Commission did not expressly adopt a position on the basis of the applicant’s request for an adjustment in respect of raw materials, by contrast, in the present action, the parties discussed the substantive accuracy of the analysis on which that request was based. In particular, the parties discussed whether the analogue country producer had used only pig iron, a raw material which, according to the applicant, is more expensive than scrap steel, and whether the applicant, as it claims, had used only scrap steel directly or indirectly.

309    That being said, it should be noted that, even if the substantive accuracy of the applicant’s statements concerning the differences in raw materials were to be established, that circumstance cannot be sufficient to demonstrate that the Commission’s conclusions in recital 32 of the contested regulation, concerning the impact of those differences and the alleged differences in labour productivity on price comparability, are vitiated by a manifest error of assessment. It should be noted that the evidence submitted by the applicant is not sufficient to demonstrate such an impact.

310    With regard to the analogue country producer’s statements, the applicant refers to three emails sent by the analogue country producer, the first on 31 January 2013 and the other two on 1 February 2013.

311    In the first place, it should be observed that, in the first email of that exchange, the analogue country producer merely indicated that it would ‘not be surprised if [its] prices were higher on the Indian market, as it has to take into account higher costs in pricing to its customers’. This is a simple assumption, which is not, moreover, supported by any concrete or quantified evidence.

312    In addition, in the same email, the analogue country producer states that ‘it is not possible that the price(s) on the Indian markets would be much higher than [the applicant’s] because … India is a fairly competitive market’. That statement, which confirms the Commission’s finding in recital 52 of the provisional regulation (see paragraph 162 above), suggests that the impact of the analogue country producer’s production costs on its prices is likely to be limited by the competitive pressure on the Indian market.

313    In the second place, the second email contains a number of details on the differences in productivity and efficiency between the analogue country producer’s company and the applicant’s company, relating, on the one hand, to the quality of the industrial equipment and, on the other hand, to the efficiency and training of the workforce. However, that email does not contain any information on the impact that those differences may have on the comparability of the prices of those two companies.

314    In the third place, the last email from the analogue country producer reproduced by the applicant merely concludes, in response to a request for clarification from the applicant, that if it benefited from the same production conditions as the applicant, its costs would, in general, be reduced by at least 25%. In addition to the fact that that very vague estimate is not supported by any concrete evidence, it only concerns the impact of production conditions on the costs of the analogue country producer and not the impact of those production conditions on its prices.

315    Moreover, for reasons similar to those set out in paragraphs 263 to 266 above, the probative value of those three emails is limited with regard to their context. In particular, it should be noted that the content of the applicant’s questions to which those emails reply, which frame or at least guide that exchange, suggests that the claims of the analogue country producer were not very spontaneous. Furthermore, that exchange took place immediately after the adoption of the provisional regulation and the disclosure to the applicant of its provisional dumping margin. The applicant’s email of 29 January 2013, which initiated the exchange, expresses, moreover, the disbelief of the applicant’s representative that the prices of the analogue producer are said to be 40% higher than its own.

316    Those three emails would therefore need to be supported by other evidence (see, to that effect and by analogy, judgment of 13 June 2012, CERATIX, T‑312/11, not published, EU:T:2012:296, paragraph 30 and the case-law cited).

317    However, as regards the sole statement of one importer of the applicant’s products in the European Union, relied on by the applicant, it contains only general considerations but no additional elements in relation to the applicant’s arguments and, moreover, reproduces them very broadly.

318    Those elements therefore do not appear sufficient to establish that differences in the use of raw materials and the efficiency and productivity of the two companies influence the level of their prices and their comparability. Indeed, in view of the wording of Article 2(10)(k) of the basic regulation, interpreted in the light of the general scheme of that article as a whole and the case-law relating thereto, that influence cannot be only presumed but must be demonstrated. In particular, it is necessary to establish that, because of those factors, buyers systematically pay different prices. Otherwise, an adjustment made on the basis of those factors could lead to the creation of an asymmetry between domestic and export prices, which would not be compatible with the objective of the above-mentioned provisions.

319    In the present case, on the one hand, as was recalled in paragraph 312 above, the prices of the analogue country producer are practised in a market where there is significant competitive pressure, in particular due to the large number of producers, and, moreover, it appears from the examination of the first part of the first plea that the applicant had not demonstrated that the formation of those prices was not normal. On the other hand, it would be necessary to carry out an overall analysis of the production costs in order to determine whether the comparative disadvantages of the analogue country producer linked to certain factors are not offset in whole or in part by other factors, for example lower wages and lower equipment costs than in relation to the applicant’s company.

320    It follows from all the above that the Commission did not commit a manifest error of assessment in considering that the applicant had not demonstrated that differences in raw materials and productivity had an influence on prices and price comparability.

321    The fourth part of the second plea must therefore be rejected.

5.      The fifth part, based on the unreasonableness of the burden of proof imposed by the Commission on the applicant

322    The applicant claims that, in rejecting the evidence in support of its requests for adjustments and in failing to indicate to it what was necessary, the Commission imposed an unreasonable burden of proof on it and infringed the principle of sound administration.

323    The Commission disputes the applicant’s arguments.

324    In this respect, it follows from the analysis of the first to fourth parts of this plea in law, in paragraphs 259 to 277, 282 to 288, 295 to 300 and 308 to 321 above, that the Commission considered, without committing an error of law or a manifest error of assessment, that the elements which the applicant had submitted in support of its requests for adjustment concerning the level of trade, packaging costs, credit costs and differences in raw material and productivity were insufficient to establish the need either to modify the adjustment already made by the Commission or to make such an adjustment. The Commission therefore did not impose an unreasonable burden of proof with regard to those adjustment requests and did not fail to fulfil its duties of diligence and impartiality under the principle of sound administration. Moreover, it does not follow from the wording of Article 2(10) as a whole or from the case-law that, where the Commission finds that a request for adjustment submitted by an interested party is not sufficiently substantiated, it would be required to ask the interested party to provide it with additional information. This part of the second plea and, accordingly, the second plea in its entirety must therefore be rejected.

D.      The fourth plea, alleging infringements of Article 3 and Article 9 of the basic regulation, concerning injury and causal link

325    The fourth plea in law comprises two parts. In the first part, the applicant claims that the Commission infringed Article 3(1) to (3) of the basic regulation by using incorrect import data. By the second part, the applicant submits that, unless the Court were to consider that recital 5 of the contested regulation incorporated, by reference, the injury and causation findings of Implementing Regulation No 430/2013, the contested regulation is vitiated either by a breach of Articles 3 and 9(4) and (5) of the basic regulation or by a failure to state reasons.

326    It is appropriate to begin by examining the second part. Indeed, it is only if the Court finds, in the course of that examination, that the contested regulation incorporates, implicitly or explicitly, the findings of Implementing Regulation No 430/2013 relating to injury and causation that it is necessary to examine the first part.

1.      The second part, based on the infringement of Article 3 and Article 9(4) and (5) of the basic regulation, since the Commission failed to establish the existence of injury and a causal link

327    The applicant claims that, following the annulment of Implementing Regulation No 430/2013 as a whole with respect to the applicant, the contested regulation imposes anti-dumping duties on its imports without having regard to the conditions laid down with respect to elements other than dumping. This, in the applicant’s view, constitutes an infringement of Articles 3 and 9(4) and (5) of the basic regulation. Finally, it submits that the reasoning of the contested regulation is also lacking in this respect.

328    The Commission disputes that line of argument.

329    As a preliminary remark, it should be noted that this part of the plea is based on the premiss that, although the annulment of Implementing Regulation No 430/2013 does not oblige the Commission to reassess injury and causation, it necessarily has the consequence of obliging it to ‘readopt’ the findings of that regulation in this respect, even if only by incorporating them by reference in the contested regulation. According to the applicant, in the present case, in the absence of a reference in the recitals of the contested regulation to those conclusions, the contested regulation is vitiated (i) by an error of law, in that the Commission re-imposed anti-dumping duties on its exports without having found the existence of injury and, consequently, a causal link, and (ii) in any event, by a failure to state reasons. In the first place, it is necessary to examine the alleged failure to state reasons.

330    In this respect, it should be recalled that the reasons for an act of the EU institutions must clearly and unequivocally reflect the reasoning of the author of the act, so as to enable the persons concerned to know the justifications for the measure adopted and to defend their rights and so as to enable the judicature to exercise its review. The requirements to be satisfied by the statement of reasons depend on the circumstances of the case, in particular the content of the act, the nature of the grounds invoked and the context of that act and all the legal rules governing the matter concerned. That requirement is satisfied when the author of the contested act sets out the facts and legal considerations of essential importance in the general scheme of that act. In addition, the statement of reasons for a regulation imposing anti-dumping duties must be assessed by taking account, inter alia, of the information which has been communicated by the EU institutions to the interested parties and the submissions made by those parties during the investigation procedure (see, to that effect, judgment of 13 September 2010, Whirlpool Europe v Council, T‑314/06, EU:T:2010:390, paragraphs 111 to 115 and the case-law cited).

331    In the present case, it should be noted that, in recital 133 of the provisional regulation, the Commission had concluded that material injury had been suffered by the EU industry within the meaning of Article 3(5) of the basic regulation on the basis of the findings set out in recitals 94 to 132 of the same provisional regulation. Furthermore, in accordance with Article 3(6) and (7) of the basic regulation, recital 153 of the provisional regulation and on the basis of the findings set out in recitals 135 to 152 of the provisional regulation, the Commission concluded that there was a causal link between imports of the product concerned from China and Thailand and the injury in question. Those conclusions were confirmed, respectively, in recital 64 and recital 79 of Implementing Regulation No 430/2013, on the basis of the findings in recitals 35 to 63 and those in recitals 65 to 78.

332    Moreover, as was recalled in paragraph 137 above, the Court did not consider it necessary to examine all the pleas in law in the action, in particular the applicant’s fifth plea in law, alleging an infringement of Article 3(1), (2) and (3) of the basic regulation, on the ground that the determination of injury caused to the EU industry was based on incorrect data.

333    In this respect, it should be noted that, in recital 6 of the contested regulation, the Commission observed, in essence, that, apart from the finding that it had wrongly rejected Jinan Meide’s request for disclosure, in the judgment of 30 June 2016, Jinan Meide Casting v Council (T‑424/13, EU:T:2016:378), the Court did not question ‘all other findings made in the contested [r]egulation’, which therefore remained valid.

334    It must therefore be noted that that recital of the contested regulation refers, in particular, to the Council’s findings on injury and causation based on recitals 35 to 78 of Implementing Regulation No 430/2013, as it applied to the applicant. It must therefore be inferred from that recital that the Commission considered that those findings remained valid and that, therefore, it could rely on them to determine, once again, the anti-dumping duty on the applicant’s exports.

335    It is true that, as the applicant points out, the Commission did not expressly incorporate the content of recitals 35 to 78 of Implementing Regulation No 430/2013 into the contested regulation and did not refer precisely to those recitals to indicate that it relied on them when adopting the contested regulation. However, the absence of such a precise reference does not affect the legality of the contested regulation, since the reference to ‘all other findings made in the contested [r]egulation’ must be considered as a clear and unequivocal reference, in particular in recitals 35 to 78 of Implementing Regulation No 430/2013, which concerns injury and causation.

336    Moreover, the fact that, in the first part of this plea, the applicant repeated the fifth plea in law of its initial action, which was based on the fact that the determination of injury caused to the EU industry was based on incorrect data, suggests that the reasoning in the contested regulation did not prevent it from defending its rights and therefore did not prevent the judicature from exercising its control with regard to that aspect of the anti-dumping investigation.

337    It follows from the foregoing that, in so far as it is based on a failure to state reasons, this part of the plea must be rejected.

338    For the same reasons, this part of the plea must also be rejected in so far as it is based on an error of law. The latter complaint is based on the premiss that, in the contested regulation, the Commission did not rely on the findings of Implementing Regulation No 430/2013 relating to injury and causation. As was noted in paragraph 334 above, it must be inferred from recital 6 of the contested regulation that the Commission considered that those findings remained valid and that, therefore, it could rely on them to determine, once again, the anti-dumping duty on the applicant’s exports. Contrary to what the applicant claims, the Commission therefore did not re-impose anti-dumping duties on its exports without finding injury and causation. The present complaint therefore has no factual basis.

339    In any event, it follows from the case-law that, where the irregularity in question affects an anti-dumping proceeding not as a whole but only with regard to the establishment of normal value, the Commission may decide, in order to implement Article 266 TFEU, to reopen the proceedings only at the stage of the investigation relating to the establishment of that normal value (see, to that effect, judgment of 28 January 2016, CM Eurologistik and GLS, C‑283/14 and C‑284/14, EU:C:2016:57, paragraph 54).

340    Moreover, it also follows from the case-law that the institution competent to implement Article 266 TFEU is not required to rule again on aspects of the initial decision which were not called into question by the judgment annulling that decision (see, to that effect, judgment of 5 September 2014, Éditions Odile Jacob v Commission, T‑471/11, EU:T:2014:739, paragraph 58 and the case-law cited).

341    In the present case, the Commission could restrict itself to resuming the anti-dumping proceeding at the stage where it had refused the applicant access to the normal value calculations and, as it states in recital 21 of the contested regulation, to examining the possible impact of the disclosure of those calculations on the merits of the investigation, which was, in any event, limited to the determination of the applicant’s dumping margin, in accordance with Article 2 of the basic regulation.

342    The Commission was therefore under no obligation to review the injury and causation assessments in recitals 35 to 78 of Implementing Regulation No 430/2013 or, as the applicant puts it, to ‘re-adopt’ them in the context of the contested regulation by explicitly incorporating them in the grounds of that regulation. Those assessments corresponded to a stage of the investigation distinct from that of the determination of the applicant’s dumping margin and had not been called into question by the partial annulment of that regulation. In this respect, it should be noted that only the operative part of that regulation, in so far as it concerned the applicant, has disappeared from the EU legal order, since it is that part of the regulation and not the assessments made in the recitals thereof, as such, which was the subject matter of the applicant’s initial action (see, to that effect and by analogy, order of 17 September 2014, Afepadi and Others v Commission, T‑354/12, not published, EU:T:2014:798, paragraphs 25 and 27 and the case-law cited).

343    It follows that the second part of the fourth plea must be rejected in its entirety.

2.      The first part, alleging the infringement of Article 3(1) to (3) of the basic regulation, due to the alleged use by the Commission of incorrect import data

344    The applicant submits that the Commission relied on incorrect import data. According to the applicant, it appears from the information at its disposal that those data clearly included imports of products which cannot be considered as related to the product concerned. However, in its view, the Commission failed to take the necessary steps to verify the accuracy of those data and to correct them by excluding those imports. It concluded from this that the Commission infringed Article 3(1) to (3) of the basic regulation.

345    The Commission disputes that line of argument.

346    At the outset it should be borne in mind that, in accordance with Article 3(2)(a) and (b) of the basic regulation, a determination of injury must be based on positive evidence and must involve an objective examination of, first, the volume of the dumped imports and the effect of the dumped imports on prices in the EU market for like products and, secondly, the consequent impact of those imports on the EU industry.

347    The purpose of that examination is to determine, in accordance with Article 3(3) of the basic regulation, whether there has been a significant increase in dumped imports, either in absolute terms or relative to production or consumption in the European Union. Indeed, under paragraph 6 of the same article, it is only if it is demonstrated that the volume or price levels of the dumped imports have a material impact on the EU industry that the existence of injury can be established. In addition, pursuant to Article 6(8) of the basic regulation, except in the event of non-cooperation by interested parties as provided for in Article 18 of that regulation, the information which is supplied by interested parties and upon which findings are based is to be examined for accuracy as far as possible.

348    According to the case-law, the determination of the existence of injury caused to the EU industry requires an appraisal of complex economic situations and the judicial review of such an appraisal must therefore be limited to verifying whether relevant procedural rules have been complied with, whether the facts relied on have been accurately stated, and whether there has been a manifest error in the appraisal of those facts or a misuse of powers (see judgment of 19 December 2013, Transnational Company ‘Kazchrome’ and ENRC Marketing v Council, C‑10/12 P, not published, EU:C:2013:865, paragraph 22 and the case-law cited). Moreover, as was pointed out in paragraph 97 above, an error in the reasoning or method of the author of the contested measure does not lead to the annulment of that measure if, in the specific circumstances of the present case, that error could not have had a decisive influence on the outcome.

349    It is in the light of those considerations that the Court must examine the applicant’s arguments in support of the present part of the plea.

350    As follows from paragraph 344 above, in the context of this plea, the applicant relies, in part, on an error of fact by the institutions, in that they based the determination of the injury caused by Chinese imports on data overestimating the amount of those imports, and in part, on a manifest error of assessment, in that those institutions considered that it was not necessary to carry out checks on those data in the light of the information provided to them, in particular by the China Chamber of Commerce of Metals, Minerals & Chemicals Importers & Exporters (‘the CCCMC’). These two complaints should be examined separately.

(a)    The alleged error of fact

351    According to the applicant, the error of fact committed by the institutions lies in the fact that the data on which the institutions’ determination of the volume of imports for the investigation period was based ‘clearly’ includes imports of products other than those of the product concerned, which are covered, like the latter imports, by CN tariff code ex 7307 19 10. In particular, it follows from the observations of the applicant and of the CCCMC during the investigation procedure that they had doubts as to whether, in the data used by the institutions, the threaded tube or pipe fittings, which are included in the definition of the product concerned, had been properly distinguished from the non-threaded tube or pipe fittings falling within the same tariff code, but which, for their part, are excluded from that definition.

352    The applicant considers that two factors make it possible to establish that overstatement of the volume of imports of the product concerned. The fact that some importers reported very significant errors in the calculation of their own import volumes had to lead the institutions to conclude that the data on the import volumes of the other importers were in all likelihood marred by the same errors. Moreover, the reliability of the import volume data used by the Commission had been questioned by the applicant and the CCCMC throughout the procedure and the data provided by the CCCMC, which were much more reliable, indicated a substantially lower export volume than those used in the provisional regulation and the contested regulation respectively.

353    First, it is appropriate to examine whether the first factor alleged by the applicant justified a review by the institutions of the data relating to imports of the product concerned.

354    In this respect, it should be recalled that the definition of the product concerned, as set out in recital 16 of the provisional regulation, corresponds to threaded tube or pipe cast fittings of malleable cast iron falling within CN tariff code ex 7307 19 10.

355    However, as is noted in recital 49 of Implementing Regulation No 430/2013, it is common ground that the above-mentioned tariff code does not include only threaded tube or pipe cast fittings of malleable cast iron, but also non-threaded tube or pipe cast fittings of malleable cast iron. In addition, it follows from recital 12 of the same regulation that, following observations from certain interested parties, the institutions excluded from the product scope definition two types of products falling within the above-mentioned tariff code, namely bodies of compression fittings using ISO DIN 13 metric thread and threaded circular junction boxes without a lid.

356    However, the documents in the file do not show that the information on imports of the product concerned actually registered in the European Union, which had been used by the Commission, was so deficient that it was not sufficiently reliable to be considered as positive evidence of the volume of imports of that product.

357    Indeed, while throughout the original investigation procedure the applicant and the CCCMC questioned the fact that those data excluded non-threaded fittings falling within the same tariff code as the product concerned, they failed to substantiate those doubts with concrete evidence. While they relied on the figures obtained on the basis of data on exports of the product concerned from the Chinese customs authorities, by contrast, they did not produce any evidence as to the intrinsic reliability of the import data themselves.

358    In addition, it is apparent from recitals 50 and 51 of Implementing Regulation No 430/2013 in particular that, as regards the determination of the volume of imports of the product concerned, the Commission had essentially used four sources of information, namely the data provided in the complaint, the information provided by the European Union’s unrelated importers, Eurostat statistics and data from the Member States’ customs authorities.

359    In this respect, it follows in particular from paragraph 4 of the notice of initiation of the original anti-dumping proceeding and from the applicant’s observations on injury dated 22 March 2012 that the complainants had provided figures relating to the share of imports from, in particular, China, which were based on Eurostat statistics corrected on the basis of certain information indicating errors in the allocation of codes.

360    However, the institutions did not rely only on Eurostat import statistics recorded under the above-mentioned tariff code and the data provided by the complainants which justified the initiation of the investigation.

361    With regard to the information provided by the importers, it should be noted that, as follows from recitals 9 and 31 of the provisional regulation, 33 unrelated importers had provided the information requested in Annex B of the notice of initiation of the proceeding. A sample of 9 of them, representing 67% of imports of the product concerned into the European Union, had been selected and, with the exception of 1, had replied to the anti-dumping questionnaire. Verification visits had also been carried out by the Commission at the premises of 5 of those importers.

362    The Commission therefore had information on the overall volume of imports of the product concerned made by each of the cooperating importers during the investigation period and detailed information on those imports, at the product type level, through the replies to the questionnaires of the importers included in the sample and the verification visits. While one of those importers did not reply to the questionnaire sent to it, it follows, however, from recital 31 of the provisional regulation that that detailed information covered at least 59% of imports of the product concerned. Consequently, with that information provided by the unrelated importers, the Commission was able to rely on a significant range of data from direct sources on imports actually registered in the European Union of the product concerned; moreover, the applicant does not dispute that the Commission verified that information.

363    Subsequently, as follows from the Commission’s letters of 12 December 2012 and 4 January 2013, that institution had determined the volume of imports for the relevant period on the basis of Eurostat statistics corrected with the help of information provided by the national customs authorities on products other than the product concerned that had been imported under the same tariff code. As results from recital 106 of the provisional regulation, that import volume was 30 786 tonnes for the investigation period.

364    While it is true that, in response to the CCCMC’s questions, the Commission indicated, in its letters of 12 December 2012 and 4 January 2013, that (i) it was unable to disclose the information provided by the national customs authorities for reasons of confidentiality and (ii) it was unaware of the methodology used by those authorities to distinguish between threaded and non-threaded fittings, the applicant does not put forward any concrete evidence which would suggest that that methodology is deficient.

365    However, as follows from recital 50 of Implementing Regulation No 430/2013, at a later stage of the proceeding, the institutions used the information provided by certain importers concerning their imports of products other than the product concerned and falling within the above-mentioned tariff code in order to correct downwards the import volumes determined in the provisional regulation for each year of the period relevant for the determination of injury, that is, the period from 2008 to the end of the investigation period (see paragraph 4 above). In particular, the volume of imports from China for the investigation period was set at 28 094 tonnes in Implementing Regulation No 430/2013 as opposed to 30 786 tonnes in the provisional regulation (see table ‘Union import volume (tonnes)’ following recital 51 of Implementing Regulation No 430/2013).

366    The applicant’s argument that the existence of significant errors reported by those importers in the calculation of their import volumes made it likely that similar errors would exist in the import volumes of other importers is a mere presumption which is by no means substantiated and which cannot be used to establish the existence of the alleged overstatement of the overall import volume of the product concerned.

367    In this respect, it should be noted that the example of an actual import volume three times higher than the import volume initially used for one of the importers concerned, to which the applicant refers in support of that argument, corresponds to a very specific situation which cannot be generalised, as the Commission has explained.

368    Indeed, according to the Commission’s explanations which are not contested by the applicant, that example concerns an operator which imports into the European Union a very specific type of product, namely bodies of compression fittings, of which it is the inventor and the main supplier on the market; that product type was excluded from the product scope definition in the contested regulation following that operator’s observations (see recitals 10 and 51 of Implementing Regulation No 430/2013).

369    Accordingly, the fact that the volume of imports corresponding to that product type thereby excluded from the product scope definition constituted the predominant part of the volume of imports attributed to that importer is irrelevant in relation to the existence of possible errors among other importers. The error in question is all the more irrelevant as it did not concern the lack of differentiation between threaded and non-threaded fittings, which was the main reason for the doubts of the CCCMC and the applicant during the investigation procedure. In any event, it is not disputed that, as stated in recital 51 of Implementing Regulation No 430/2013, that error, as well as all those reported by importers on the basis of reliable and documented information, led to a downward correction of the import volume in that regulation.

370    It follows from the above that the applicant has not demonstrated that, in the light of its intrinsic quality, the information on which the Commission relied in determining the volume of imports of the product concerned was not sufficiently reliable.

371    Secondly, it is necessary to examine the second element alleged by the applicant, namely that the data provided by the CCCMC were much more reliable than those used by the institutions and that they indicated a substantially lower volume of exports.

372    It should be noted that, as is apparent from the documents in the file, until the CCCMC’s comments of 17 December 2012, neither that organisation nor the applicant submitted any concrete information making it possible to assess the figure of imports of the product concerned into the European Union from China.

373    In addition, in that letter dated 17 December 2012, the CCCMC estimated the volume of exports from China of the product concerned to the European Union at around 18 000 tonnes and a maximum of 20 000 tonnes, without, however, providing the information on which it relied. Thus, the applicant does not dispute that this was a mere estimate which could not, at that stage and in the absence of evidence, be taken into account by the institutions. Indeed, such an unverifiable estimate could not be considered as positive evidence which could be the subject of an objective examination within the meaning of Article 3(2) of the basic regulation. Consequently, the only evidence provided by the CCCMC on which the applicant can rely in this case relates to the data and explanations provided by the CCCMC in its letter of 6 February 2013.

374    In this respect, following the hearing of 4 February 2013, the CCCMC explained, at the Commission’s request, that it had calculated, on the basis of data from the Chinese customs authorities, that the above-mentioned export volume amounted to 17 231 tonnes. It further argued that that figure constituted the highest possible volume of exports, as it had been determined by adding the figure of the applicant’s export volume for the investigation period, as indicated in its reply to the questionnaire, and the export volume of all products covered by the above-mentioned tariff code exported by the other Chinese producers to the European Union. In addition, in the annex, it provided a table of data on exports of those products to the European Union of 321 Chinese producers, which it indicated it had compiled on the basis of data from the Chinese customs authorities.

375    With regard to the latter data and explanations, in the first place, it should be observed that the mere fact that the figure of 17 231 tonnes of export volume of the product concerned from China to the European Union during the investigation period is significantly lower than the figure adopted by the institutions in the contested regulation for the corresponding import volume is not decisive in this case. It follows from paragraphs 356 to 369 above that no evidence was provided during the original investigation procedure to establish the unreliability of the import volume which had accordingly been determined by the institutions. Thus, it is for the applicant to prove that, despite the absence of that evidence, the figure relating to exports of the product concerned on which it relies would nevertheless be more reliable.

376    In the second place, it should be noted that that figure of 17 231 tonnes and the supporting data provided in the CCCMC’s letter of 6 February 2013 corresponded to the investigation period, that is, the year 2011.

377    As argued by the Commission, that evidence therefore did not cover the rest of the period relevant for the injury determination, which, as was pointed out in paragraph 364 above, was between 2008 and the end of the investigation period. That figure thus constituted a partial figure which did not reflect the trend of the imports in question, which, as set out in Article 3(3) of the basic regulation, is decisive for the establishment of injury. Furthermore, it should be observed that the use of data on exports of the product concerned to the European Union for only 1 year of the period relevant for the determination of injury, and not for the whole period, could in any event lead to inconsistencies distorting the examination of import trends.

378    In the third place, it follows from the CCCMC’s letter of 6 February 2013 that the figure of 17 231 tonnes of exports of the product concerned, on which the applicant relies, is not a figure which results directly from official data but a figure which the CCCMC itself determined on the basis of an addition of two figures of a different nature and source.

379    On the one hand, the CCCMC used the actual amount of exports to the European Union of the product concerned made by the applicant, which was included in the anti-dumping questionnaire completed by the applicant and which amounted to 11 603 tonnes.

380    On the other hand, on the basis of data from the Chinese customs authorities, it identified certain producers of all products falling within CN tariff code ex 7307 19 10 as being producers of the product concerned and used the total exports to the European Union made by those producers, namely 5 628 tonnes.

381    That latter calculation was supported by a table annexed to the letter of 6 February 2013 containing data presented as originating from the Chinese customs authorities, apparently relating to all exports to the European Union of products falling under the above-mentioned tariff code. For each of the 321 companies listed therein, that table included, inter alia, a column showing their respective export volumes. The total sum of those export volumes, as indicated in the table, was 40 517.46 tonnes. That table also included an additional column, apparently added by the CCCMC, which was entitled ‘Exporter of the product concerned or not’ and which was completed with the word ‘yes’ for 22 of the companies listed in the table, the rest of that column being empty.

382    It must be observed that the amount thus obtained by the CCCMC does not correspond to the total exports of the product concerned actually recorded from China to the European Union, but, in part, as regards the amount relating to exports by producers other than the applicant, referred to in paragraph 381 above, to an estimate by the CCCMC itself.

383    However, it is not apparent from the documents in the file that the CCCMC had informed the Commission of the methodology used to identify, among the 321 producers registered by the Chinese customs authorities as exporting products to the European Union under CN tariff code ex 7307 19 10, the 22 producers exporting the product concerned and, a fortiori, that it had sent it supporting documents relating to that identification.

384    In this respect, the parties seem to agree on the fact that, despite what is indicated by recital 49 of the contested regulation, not all the producers thus identified by the CCCMC were cooperating producers. This is confirmed by recital 30 of the provisional regulation, according to which only 12 Chinese exporting producers cooperated.

385    Consequently, the 10 non-cooperating exporting producers mentioned by the CCCMC as having exported the product concerned to the European Union could not be identified as such on the basis of the elements of the investigation procedure, as those exporting producers did not make themselves known to the institutions.

386    They also could not be identified on the basis of the information provided by the Chinese customs authorities, since, as follows from paragraph 380 above and as explained by the CCCMC in its letter of 6 February 2013, that information did not make it possible to distinguish, among the exports covered by CN tariff code ex 7307 19 10, threaded tube and pipe fittings falling within the product concerned from non-threaded tube and pipe fittings not falling within it.

387    For the same reasons, the elements of the investigation procedure and the information provided by the Chinese customs authorities also failed to ensure that the remaining 299 companies, for which no mention was made in the column entitled ‘Exporter of the product concerned or not’, did not export the product concerned, that is, threaded tube and pipe fittings, to the European Union.

388    Therefore, in the absence of evidence produced by the CCCMC to support its identification of 10 non-cooperating exporting producers as being the only ones among that category of exporting producers to have exported the product concerned to the European Union, the data thus established were not verifiable.

389    Invited by the Court to specify, in a written reply, the methodology used by the CCCMC to identify the 22 Chinese producers exporting the product concerned, which are mentioned in the table annexed to its letter of 6 February 2013, the applicant indicated that that body had (i) asked its members to provide it with information on their own activities and on the activities of their competitors mentioned in the relevant table and (ii) that it had carried out an examination of the websites of the companies mentioned in that table in order to verify whether those companies had produced or sold the product concerned.

390    In addition to the fact that it is not apparent from the documents in the file that such clarifications were made to the Commission in the context of the anti-dumping investigation, it should be noted that they contribute to raising further questions about the reliability of the estimate provided by the CCCMC.

391    On the one hand, it does not follow from those clarifications that all the 321 companies listed in the table annexed to the letter of 6 February 2013 were members of the CCCMC, the latter having sought to supplement the information provided by its members on their own activities with information held by them on their competitors and online research.

392    On the other hand, the information held by members of the CCCMC on the products manufactured by their competitors and that which can be found on the websites of those companies, where such a website exists, is clearly not sufficient to ensure that the CCCMC can accurately determine which of those companies manufactured the product concerned and provide a precise estimate of the volume of the product concerned exported to the European Union.

393    Moreover, as confirmed by the Commission, in a written reply to a question put by the Court, a number of companies exporting products covered by CN tariff code ex 7307 19 10 were mentioned by the Chinese customs authorities not using their name, but using a Chinese language term meaning ‘unknown company’. It is therefore relevant to note, as the Commission did, that most of those companies were excluded by the CCCMC from its list of exporters of the product concerned. Indeed, while that circumstance could not be sufficient in itself to exclude the data invoked by the CCCMC, it confirmed the existing uncertainty regarding the information available to the CCCMC regarding the exports of the 299 companies which it had not identified as exporting the product concerned to the European Union.

394    Finally, the applicant does not dispute that the total volume of exports falling under CN tariff code ex 7307 19 10, which appeared in the table provided by the CCCMC, that is, 40 517 tonnes, was higher than the total of 35 986 tonnes corresponding to the volume of imports falling under the same tariff code, as indicated by the European Union’s customs statistics for the same period. That factor could be considered by the institutions, during the original anti-dumping proceeding, to be an indication that the allegations of the applicant and the CCCMC that the volume of exports of the product concerned was significantly lower than that of imports indicated in the provisional regulation were not sufficiently substantiated.

395    The institutions therefore correctly considered that the data provided by the CCCMC in its letter of 6 February 2013 were not sufficiently reliable and complete for the Commission to consider that the estimate of exports of the product concerned from China to the European Union, which was included therein and which was based on those data, correctly reflected the corresponding import volume or at least reflected it more reliably than the figure of 28 094 tonnes, as determined by Implementing Regulation No 430/2013 in this respect.

396    The fact that the Council wrongly stated in recital 50 of Implementing Regulation No 430/2013, in order to justify the refusal to take into account the evidence submitted by the CCCMC, that the CCCMC relied on estimates relating to exports by the cooperating companies is irrelevant in this case. Indeed, in accordance with the principles recalled in paragraphs 97 and 348 above, that error could not have had a decisive influence on the result. In this respect, on the one hand, the Commission is right to note that, although the CCCMC did take into account, in its estimate, some non-cooperating companies, it did in fact exclude the vast majority of them from that estimate and, consequently, the vast majority of the total volume of exports made by all those non-cooperating companies. On the other hand, as concluded in paragraph 395 above, the data produced by the CCCMC could not be considered to be sufficiently reliable. Consequently, despite the incorrectness of the Council’s assertion in recital 50 of Implementing Regulation No 430/2013, the institutions did not err on that issue of the reliability of the data relied on by the applicant.

397    The complaint based on the error of fact must therefore be dismissed.

(b)    The manifest error of assessment

398    First of all, it should be observed at the outset that, as follows from paragraphs 356 to 396 above, the data collected by the Commission on imports of the product concerned could rightly be considered to be reliable by the institutions, whereas that was not the case with the data provided by the CCCMC, which were partial, uncertain and unverifiable.

399    Consequently, the applicant’s argument that the data provided by the CCCMC came from an official source and, if not verifiable by the institutions, should have led them, at the very least, to question their own data, cannot be accepted.

400    Indeed, regardless of the issue in relation to the delay in submitting that evidence, the institutions did not commit any manifest error of assessment in considering that they could not use that evidence to correct the data at their disposal without having verified that data.

401    Moreover, the institutions did not commit a manifest error of assessment by considering that it was impossible for them to verify that evidence, in particular because it was submitted at an excessively late stage of the anti-dumping proceeding.

402    The applicant submits, in essence, that, prior to the expiry of the deadline for the submission of comments on the provisional regulation, it and the CCCMC warned the Commission of the need to ensure the accuracy of the import data and that that body provided an initial estimate of the volume of imports.

403    However, as was pointed out in paragraphs 372 and 373 above, it is not disputed that no information from the applicant or the CCCMC on the volume of imports of the product concerned had been submitted to the Commission before the CCCMC’s letter of 17 December 2012 and that the unsubstantiated estimate contained in that letter could not be regarded as positive evidence within the meaning of Article 3(2) of the basic regulation.

404    With regard to the figure of 17 231 tonnes of export volume of the product concerned to the European Union in 2011 and the supporting data submitted by the CCCMC in its letter of 6 February 2013, it should be observed that, in view of the strict time limits for the various stages of an anti-dumping proceeding, the Council’s assertion in recital 49 of the contested regulation that it was impossible to verify that information in an objective examination process without extending the investigation period beyond the maximum time limit for the investigation period is not manifestly erroneous.

405    It followed from Article 9(4) of the basic regulation, read in conjunction with Article 9(2) of the same regulation, in its version in force on the date of adoption of Implementing Regulation No 430/2013, that the adoption by the Council of a proposal for definitive measures by the Commission terminated the anti-dumping investigation. Moreover, pursuant to Article 6(9) of that regulation, the investigation had to be terminated within a maximum period of 15 months following its initiation. In addition, again pursuant to Article 9(4) of the basic regulation, in conjunction with Article 9(2) of the same regulation, the Council had 1 month from the transmission of the Commission’s proposal to reject that proposal, which, in the absence of such a rejection decision, was deemed adopted. Moreover, pursuant to Article 20(4) of that regulation, final disclosure on the essential facts and considerations on the basis of which it was intended to recommend the imposition of definitive measures was forwarded to interested parties no later than 1 month before the Commission submitted a proposal for final action.

406    In the present case, since the investigation was initiated on 16 February 2012, the definitive regulation had to take effect, in any event, by 16 May 2013 at the latest. In addition, under the provisions referred to in paragraph 405 above, the Commission was in any event required to submit a proposal for definitive measures by 16 April 2013 at the latest and to disclose the final information document to interested parties by 16 March 2013 at the latest. Thus, the Commission had a maximum period of time between 6 February and 16 March 2013 to verify the information provided by the CCCMC and, if necessary, to use it in order to correct the figure determined for the volume of imports.

407    If such a verification had been possible, it would have implied that the CCCMC provided explanations on the information enabling it to consider that the Chinese companies which were not mentioned as exporting the product concerned to the European Union were indeed in this case, as well as information making it possible to distinguish threaded fittings from non-threaded fittings in the export volumes to the European Union of the producers, mentioned in that table.

408    Given the uncertainties about the accuracy of the information available to the CCCMC, as noted in paragraphs 384 to 393 above, the above verification would likely have required further investigation by the CCCMC. Even if those investigations had been successful, it would still have been necessary for the Commission to review the data at its disposal in the light of that information, which would necessarily have involved further investigations with the customs authorities of the Member States and importers. Indeed, in order to comply with the obligation resulting from Article 3(2) of the basic regulation to establish the volume of imports on the basis of positive evidence and an objective examination, the institutions cannot carry out less detailed checks. Thus, by considering that the time available to them, taking into account the procedural time limits recalled in paragraph 406 above, was not sufficient to carry out such checks and investigations, the institutions did not commit a manifest error of assessment.

409    In any event, there was a significant risk that the necessary checks and investigations, mentioned in paragraph 408 above, would not culminate in positive evidence.

410    Therefore, in this case, the institutions were not required to carry out additional checks in the light of the evidence provided by the CCCMC. Such verifications were neither necessary nor likely to culminate, with a sufficiently high probability, in more reliable results than those already obtained within the applicable time limits.

411    Those considerations are not called into question by the argument that the institutions took into account information submitted by certain importers subsequent to the information provided by the CCCMC in its letter of 6 February 2013. The applicant does not dispute that, as the Commission argued, the information it received from those importers was verifiable, that is to say that it was supported either by customs documents showing that they imported goods other than the product concerned, or by serious evidence on their commercial activity indicating that they imported goods for electrical installations not covered by that product. It was therefore reliable information which could be considered to be positive evidence within the meaning of Article 3(2) of the basic regulation and which could be the subject of an objective examination within the meaning of the same provision, without the need for further investigations. Therefore, such information could be taken into account by the institutions when determining the volume of imports of the product concerned. Consequently, the fact that the institutions had treated that information differently from the information produced by the CCCMC cannot be considered discriminatory, since that treatment is justified by an objective difference between that information and the latter evidence.

412    Therefore, the first part of the fourth plea and, consequently, the plea in its entirety must be rejected.

413    It follows from all the foregoing that, although the first, second and fourth pleas in law must be rejected, the third plea in law, taken in its two parts, is well founded. The contested regulation must therefore be annulled.

 Costs

414    Under Article 134(1) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings.

415    Since the Commission has been unsuccessful, it must therefore, in accordance with the form of order sought by the applicant, be ordered to pay the costs.

On those grounds,

THE GENERAL COURT (Fifth Chamber)

hereby:

1.      Annuls Commission Implementing Regulation (EU) 2017/1146 of 28 June 2017 re-imposing a definitive anti-dumping duty on imports of threaded tube or pipe cast fittings, of malleable cast iron, originating in the People’s Republic of China, manufactured by Jinan Meide Castings Co., Ltd;

2.      Orders the European Commission to pay the costs.

Gratsias Labucka Ulloa Rubio

Delivered in open court in Luxembourg on 20 September 2019.

E. Coulon       D. Gratsias

Registrar President

Table of contents

I. Background to the dispute

A. Background to the dispute in Case T424/13

B. The action in Case T424/13

C. Background to the dispute subsequent to the judgment of 30 June 2016, Jinan Meide Casting v Council (T424/13)

II. Procedure and forms of order sought

III. Law

A. The third plea, alleging infringement of Article 2(7)(a), Article 2(10) ab initio and (a) and Article 2(11) of the basic regulation, owing to errors made by the Commission in determining the normal value of non-matching product types

1. Preliminary observations

2. The first part, based on the fact that the Commission adopted an unreasonable methodology to determine the normal value for non-matching product types

3. The second part, based on the fact that the Commission infringed Article 2(11) of the basic regulation by adopting a methodology which leads to the de facto exclusion of transactions relating to non-matching product types

B. The first plea, alleging infringement of Article 2(7)(a) of the basic regulation, on the grounds of errors made by the Commission in the determination of normal value

1. The first part of the plea, based on the Commission’s incorrect consideration of the analogue country producer’s transactions in very small volumes

(a) The first complaint, based on the unreasonableness of the determination of the dumping margin, due to the fact that prices of product types sold in very low volumes were taken into account

(1) The first argument, based on the artificially high prices of the product types sold in very low volumes

(2) The second argument, based on the impact of taking into account the product types sold in very low volumes on the increase in the dumping margin

(3) The third argument, based on the volatility of transaction prices in very low volumes

(b) The second complaint, alleging that the low volume sales were not made in the ordinary course of trade

2. The second part, alleging the use of a methodology for calculating specific cost data based on an unreasonable general presumption

(a) The error of law, in that the Commission misapplied Article 2(5) of the basic regulation

(b) The manifest error of assessment, in that the profitability check carried out by the Commission on the basis of the allocation of costs based on turnover would have led to the unreasonable determination of a general profitability threshold and the arbitrary exclusion of certain transactions

C. The second plea, alleging infringement of Article 2(10) of the basic regulation and Article 2.4 of the anti-dumping agreement, by reason of the Commission’s rejection of the applicant’s requests for adjustment

1. The first part, based on the Commission’s rejection of the applicant’s requests for adjustment in respect of level of trade

2. The second part, based on the Commission’s rejection of the applicant’s request for adjustment in respect of packing costs

3. The third part, based on the Commission’s rejection of the applicant’s request for adjustment in respect of credit costs

4. The fourth part, based on the Commission’s rejection of the applicant’s requests for adjustment in respect of raw materials used and productivity

5. The fifth part, based on the unreasonableness of the burden of proof imposed by the Commission on the applicant

D. The fourth plea, alleging infringements of Article 3 and Article 9 of the basic regulation, concerning injury and causal link

1. The second part, based on the infringement of Article 3 and Article 9(4) and (5) of the basic regulation, since the Commission failed to establish the existence of injury and a causal link

2. The first part, alleging the infringement of Article 3(1) to (3) of the basic regulation, due to the alleged use by the Commission of incorrect import data

(a) The alleged error of fact

(b) The manifest error of assessment

Costs

*      Language of the case: English.

1      This judgment is to be published in extracts.