National Stock Exchange Of India Ltd. v. Competition Commission Of India

PER MR. JUSTICE V.S. SIRPURKAR, CHAIRMAN An order passed by the majority Members of the Competition Commission of India (for short the CCI) holding Appellant-National Stock Exchange of India Ltd. (for short NSE) to be a dominant player in the relevant market and further holding that it had abused its dominance therein and on that count inflicting a penalty of Rs. 55.50 crores (@ 5% of the average turnover) falls for consideration in this Appeal. The other order passed by the minority two Members namely - Shri Anurag Goel and Dr. Geeta Gouri, however, exonerating the Appellant-NSE, holding that there was no violation of Section 4 of the Competition Act, 2002 (for short the Act) on the part of NSE does not.

2. Information was led before the CCI at the instance of the respondent MCX Stock Exchange Ltd. (MCX-SX for short) against the National Stock Exchange of India Ltd. (the appellant herein) and DotEx International Ltd. (DotEx for short). The MCX-SX is now Respondent No. 2 in this Appeal while DotEx has not been joined as a party to this Appeal which was registered as Appeal No. 15 of 2011.

3. In the said information, it was alleged that the appellant had abused its dominance under Section 4 of the Act by introducing predatory pricing by waiving transaction fee altogether in the newly established Currency Derivatives Segment (CD Segment for short). It was also urged that for this NSE was using its dominance in the non-CD segments to enter into and protect its position in the CD Segment and was also causing denial of market access by promoting MCX SX, Financial Technologies of India Ltd. (FTIL) from offering its software ODIN for the use of its CD Segment. The information which was filed under Section 19(1)(a) of the Act was also against the Omnesys Technologies Pvt. Ltd. (OMNESYS). It was urged that MCX-SX and NSE were providing currency futures exchange services. It was pointed out that NSE through its circular dated 26.8.2008 announced a transaction fee waiver in respect of all currency future trade (Currency Derivatives) executed on its platform and then it continued to extend its waiver programme from time to time and even on that date when Section 4 of the Act came on the anvil on 20.5.2009. Further information was provided that due to this transaction fee waiver by the NSE, MCX-SX which was the only other player in the field in respect of currency derivatives had also to waive transaction fee on its platform for CD Segment from the date of its entry into the stock exchange business and which was somewhere in the month of October, 2009 and thus MCX-SX was suffering huge losses as it had no income through its CD Segment and the MCX SX had the license only for dealing in the CD Segment. This CD Segment seems to have been introduced by the recommendations of RBI and SEBI in August, 2008 the date from which the NSE started its operation in CD Segment. It must be noted that the MCX-SX got the license for operating only in the CD segment. It did not have the license to operate in any other segments like Stocks Future and Options (F&O), W.D.M. etc. It was pointed out in the information that NSE was charging no admission fee for membership in the CD Segment though it was so charging in the equity, F&O and debt segments. It also did not collect the annual subscription charges and an advance minimum transaction charges only in respect of CD Segment. It was urged that the cash deposits to be maintained by a member in the CD segments were also kept at a very low level as compared to the other segments.

4. It was further urged that NSE was not charging any fee for providing the data feed only in respect of CD segment ever since its commencement the segment on 26.8.2008 and continued not to charge any fees in respect of the CD Segment. It was, therefore, urged that MCX-SX was unable to charge anything on account of the transaction fees, admission fees for membership, annual subscription charges, advance minimum transaction charges and also fee for providing data fee and thus it had no income from the CD segment whatsoever and further CD segment was only segment in which the MCX SX was given.

5. A complaint was also made in respect of OMNESYS which was a software provider for financial and security market in which the NSE had taken 26% stake through DotEx, which is a 100% subsidiary of NSE. It was urged that the DotEx/OMNESYS had introduced a new software known as "NOW" to substitute a software called "ODIN" developed by Financial Technologies India Ltd. (FTIL), which was the promoter of the MCX-SX and the market leader in the brokerage solution sector.

6. It was further urged that after taking stake in OMNESYS, DotEx which was 100% subsidiary of NSE had written individually to the NSE members offering them the technology of "NOW" free of cost for the next year. Simultaneously, NSE had refused to share its CD Segment Application Programme Interface Code (APIC) with FTIL and thus disabling the ODIN users from connecting to the NSE CD segment trading platform through their preferred mode. The product was thus thrust upon the consumers desirous of the NSE CD Segment, was the product "NOW" developed by DotEx and OMNESYS, in place of ODIN. It was urged that NSE was using "NOW" on a separate computer terminal for accessing its CD Segment. It was further urged that the main advantage of the ODIN software was that a trader could view multiple markets using same terminal and take appropriate calls. The shifting between different terminals (NOW and ODIN) severely hampered the traders ability to do so, thus the expected response from a common trader was to confine to one terminal which connected to the dominant player only i.e. to use the "NOW" terminal (free of cost) and confine itself to the NSE CD Segment. It was therefore urged that on this count also the NSE had abused its dominance and so had OMNESYS.

7. It was further urged that the losses suffered by informant in the CD Segment were much higher than the loss suffered by the NSE (due to the waiver of the transaction fees and other fees) as the NSE enjoyed the economies of scale and has the ability to cross-finance the losses from the profits made in other segments wherein it was dealing and thus it has the financial strength to fund its predatory practices based on massive reserves built through accumulation of monopoly profits over the years. In contrast, MCX-SX was dependent solely on the revenues from the CD Segment and its losses were mounting in view of its transaction fee waiver, the continuation of which was compelled by the NSE's decision to continue with the total fee waiver.

8. It was also urged that NSE's fee waiver would not only eliminate the business of the MCX-SX in CD segment but also eliminate the potential and efficient competitors from the entire stock exchange services. It was urged that the policy of fee waiver was adopted by NSE as an exclusionary device to kill competition and competitors, and to eliminate MCX-SX from the market as a supplier of stock exchange services (CD Segment) and thus NSE had, therefore, used its dominant position in the relevant market to eliminate competition and competitors. It was also urged that NSE along with DotEx and OMNESYS had violated provisions of section 4 of the Act by denying the integrated market watch facility to the consumers by denying access of Application Programme Interface Code (APIC) to the promoter of MCX-SX.

9. It was further urged that the various fee waivers and the low level of deposit requirements only with respect to the CD segment of NSE were completely at a variance with its conduct in other segments and were aimed at eliminating competition and discouraging potential entrants and amounted to the tactics for excluding the other competitors from entering into the field. On this basis number of reliefs were prayed : (a) To investigate infringement of Section 4 of the Act by NSE; (b) To direct the NSE to discontinue transaction fee, data- feed fee and the admission fee waivers in respect of the CD segment and to impose transaction fees, data-feed fee and admission fee in the said segment equal to that in the other segments of NSE; (c) To order NSE to require its members to maintain deposits for the CD segment at a level that is consistent with the levels of other segments; (d) To grant an injunction restraining the NSE from continuing the transaction fee, data-feed and admission fee in respect of the CD segment in line with those in other segments; and (iii) mandate NSE to collect deposits from members at a level on par with those in its other segments, pending final disposal of the complaint; (e) To order NSE to pay all of the complainant' costs and impose the highest level of penalties on the NSE in accordance with the Act, so as to have deterrent effect and ensure free and fair competition in the relevant market; and (f) To pass such other order as the Commission may deem fit to ensure free and fair competition in stock exchange services market.

10. On this basis the CCI entertained this information and took a prima facie view that this information deserves to be investigated by the Director General (D.G. for short) and ordered accordingly.

11. MCX-SX had also filed an application dated 6.7.2010 for interim relief under Section 33, According to which it was complained that if NSE continued to offer its services in the CD segment free of cost despite a significant increase in the turn over, the MCX-XS could suffer combined loss of around Rs. 100 crores. It was urged that since the CCI had already formed a prima-facie opinion in this matter and had sent the matter for investigation to the Director General, MCX-XS would be required to exit the market. The CCI, however, refused to pass any order under Section 33 particularly in view of the fact that the investigation by the D.G. ordered by it, was near completion. We need not go on that issue whether the CCI was right in refusing the interim relief. We also need not consider the question about the APIC and the alleged tactics played by DotEx and OMNESYS in respect of ODIN and NOW for the reasons which we would elaborate at the end of this judgment. The parties also did not address us on that issue for the obvious reasons that that question was already closed between the parties by a compromise affected before the Bombay High Court.

12. The matter was investigated by the DG which investigation included examination of the financial statement of NSE, details of fees and charges, and other costs incurred in different segments.

13. In addition, the D.G. also studied several reports of regulators and circulars of Expert Committee as also regulations/circulars issued by SEBI to understand the mechanics of various charges imposed by the stock exchange services.

14. The D.G. while considering the relevant market came to the conclusion that the stock exchange business as a whole constituted the relevant market. He took this view as according to him the product differentiation was not of much practical consequence and the demands - supply structure was similar across all the segments and there was obvious co-relation between the segments which were limited in number. It was also noted by him that from the demand side, majority of the stock brokers are the members of all the segments and the users were also almost common. He deduced that each product was used with a common objective of profiteering of investment and trading.

16. This view was obviously opposed by the NSE according to which the stock exchange services could not be a relevant market in this case. It argued before the D.G. that each segment of the capital market and the debt market is a distinct market by itself as there were separate trades at stock exchange in respect of different segments. It was argued by the NSE before the DG that the CD market was of recent origin and could not be said to be interchangeable or substitutable from the demand side. Further, it pointed out that the CD segment was essentially for the importers and exporters who desired to hedge the currency fluctuation risk which was not in case of equities/debts/F&O segments. Without prejudice to this contention, NSE further argued that if at all there was the question of interchangeability or substitutability arose the CD market could be seen as a substitute of OTC segments. Thus according to the NSE the CD market and the OTC market was the relevant market. This position was taken by the NSE also before us also. The DG, however, held the related market to be the stock exchange business as a whole. In that he considered the following five segments to arrive at a relevant product. They were :-

(i) Equity segment

(ii) Equity F&O segment

(iii) Debt segment

(iv) CD segment; and

(v) OTC market for trades in foreign currency. The D.G. noted the provisions of the Securities Contracts (Regulation) Act, 1956 (SCRA) and after the issue of regulatory framework, both Bombay Stock Exchange and NSE could commence the trading in CD segment immediately, which fact indicated that CD segment was a part of the stock exchange market services. According to the DG report, since any exchange could easily start operations in any of the segments of capital market, there was supply side substitutability between the segments. Therefore, according to the DG report, the entire stock exchange market service was a single relevant product.

17. The D.G. also came to the conclusion that it was not possible to ascertain substitutability between CD and the other segments of stock exchange services. The D.G. relied on several cases from international jurisdictions such as Case Nos. 351 US 377 (1956), ECR 1973 0215, ECR 1980 page 03775, ECR 1983 page 03461, ECR 1991 page I - 03359, ECR 1994 page II - 00755, ECR 1996 page I - 05951, ECR 1998 page I - 0779 and others. Hence he held that in all these judgments, the courts have relied on the requirement of interchangeability in contrast with substitutability. It was also noted that the courts had placed greater reliance on the characteristics of the product for the purpose of satisfying constant needs.

18. He concluded that from the very definition of futures contract, it was clear that the basic characteristics of the product were similar to the equity futures contract and therefore CD and equity derivative segments had common characteristics. He held that "equity segment including equity F&O and CD segment which mainly comprise the stock exchange services market are substitutable on the product characteristics basis." He also observed that F&O market and CD market are used by similar type of participant, namely speculators and hedgers. He, however, came to the conclusion that this could not be said about the OTC market for which he firstly relied on the provisions of SCRA, RBI Internal Working Group Report, RBI - SEBI report on CD Market, FEMA etc. He thus concluded on the similarity of operations of stock exchange services in relation to different segments traded in exchanges, that they were substitutable.

19. He also took into account the membership patterns of MCX-SX and NSE and found a very high commonality of members at NSE as well as MCX-SX with the membership of other segments. According to him, this clearly established that the existing members of other segments were primary traders in the CD segment, which further implied that actual hedgers of foreign exchange did not see substitutability or interchangeability in the CD market as against the OTC market.

20. The D.G. also considered the SSNIP (Small but significant and non- transitory increase in price) test and held that it was not possible to rely on that test.

21. Thus the D.G. finally took the view that stock exchange services in India including equity F&O, WDM and CD was the relevant market but not the OTC market.

22. While assessing the dominant position, the D.G. examined the status of the NSE on account of : (a) Position of strength (b) Ability to operate independently of competitive forces prevailing in the relevant market; and (c) Ability to affect its competitors of consumers or the relevant market in its favour.

23. The D.G. also examined the market power along with the lines indicated in Section 19(4) of the Act. While considering the market share in the relevant market of stock exchange services, the D.G. found that when NSE had started its operation in November, 1994 there were 21 stock exchanges in India with Bombay Stock Exchange (BSE) commanding a market share of 41.5% in the equity segment. By 2008-09, NSE had acquired 71.43% of the equity segment as against the vastly reduced share of 28.55% of BSE. He found that in the F&O segment, NSE commenced trading in June, 2000 and had risen to over 99% market share since then. In the WDM segment, NSE commenced trading in June, 1984 NSE has consistently maintained market share of over 90% since 2001- 2002. It also found that insofar as the CD segment was concerned the NSE had a market share of 47 - 48% as against 52 - 53% of MCX SX. In view of this the DG found that NSE was a dominant player.

24. While considering the size and resources of the enterprise also on the basis of the statistics available, the DG came to the conclusion that NSE had total income of Rs. 1042 crores with profit before tax of Rs. 689 crores which indicated a sound financial position of the NSE. While considering the third factor of size and importance of the competitors, the D.G. concluded that in comparison with NSE which had commenced the trading in November, 1994 the remaining 19 original exchanges started collapsing due to intense competition from NSE. This included even the BSE which was one sound player found that though MCX-SX entered the arena only on 14.8.2008, it ended the first year with the carry forward loss of Rs. 298.7 million. The D.G. also took into consideration the economic power of the enterprise and found that there also NSE had presence in 1486 cities and towns and majority of investors, brokers etc. were connected with NSE with its extensive infrastructure and also found that unlike MCX-SX, NSE could raise equity and debts to funds its requirements. On this count also the D.G. considered NSE as a dominant player.

25. On the next count of vertical integration of enterprises or sale or services net work of such enterprises, the D.G. clearly held that the NSE came clearly as a leader.

26. So also on the other factor of dependence on consumers the D.G. held that the there was a far greater number of buyers and sellers to NSE and it enjoyed the benefits of network effects resulting from higher liquidity and lower transaction costs and thus it emerged as a leader. On the last two aspects on the countervailing buying power it was held that the users of the stock exchange services were individually too small to countervail buying power.

27. On entry barriers, it was noted by the D.G. that stock exchange services was an area of high regulatory barriers. He also considered the high capital cost of entry, financial risk, marketing and technical entry barriers further strengthens the already dominant position of NSE. Thus the D.G. concluded that the NSE was a dominant player in the market.

28. On the question of abuse of dominant position, the D.G. examined the abusive behavior on account of four factors :- A. Transaction fee waiver; B. Admission fee and deposit level waivers; C. Data feed fee waiver; and D. Exclusionary denial of "integrated market watch" facility.

29. The D.G. noted on the first aspect of transaction fee waiver that the NSE had issued a Circular No. NSE/CD/11188 dated 26.08.2008 whereby it announced transaction fee waiver in respect of currency futures trades executed on its platform. It thereafter issued three circulars (1) on 26.09.2008, which was to be valid upto 30.09.2008; (2) a circular dated 28.11.2008, valid upto 31.03.2009; and (3) a circular dated 30.03.2009, which was to be valid upto 30.06.2009. It is obvious that out of these four circulars, three related to the pre 20.05.2009 period. Considering that section 4 was activated w.e.f. 20.05.2009, it is the last circular dated 30.03.2009, which would be a relevant circular, as it covers the date 20.05.2009. Thereafter, no circulars came to be issued right till the date when the impugned order of CCI was passed. Thus, right w.e.f. 20.05.2009 the NSE continued not to charge any fees in respect of the CD segment.

30. The D.G. has referred to the defence of NSE that the waiver was done to encourage larger participation as the CD segment was at a nascent stage. According to the NSE, this policy was influenced by report of the High Powered Study Group on Establishment of New Stock Exchanges. It was also stated that the Pricing Committee was constituted by the Board of NSE to guide and decide all pricing matters and the waiver of transactions fee was the decision of that Pricing Committee. (We shall have the occasion to refer to the decisions of this Pricing Committee in the subsequent part of this judgment)

31. The D.G. then examined the transaction charges levied by NSE in other segments and noted that turnover of NSE was Rs. 1078 crores in comparison to the BSE turnover of Rs. 2.52 crores. The D.G. noted that NSE did not have the historical philosophy of waiving fee to develop a nascent market for which he based his findings on the transaction fees levied on WDM segment and the other segments. The D.G. also gave additional reasons to refute the theory of NSE of nascent market. The D.G. found that in Gold ETF segment the transaction charges were levied from March 2007 till August 2009, when it was the only exchange trading in Gold ETF segment. However, it was only after February, 2010 that NSE waived/ reduced transaction fee in Gold ETF segment. From this the D.G. came to the conclusion, after noting the entry of BSE into Gold ETF market, that the NSE introduced waivers/ reductions in this sub segment from March, 2010 with the obvious view of maintaining its superiority in the market. After examining various board minutes and agenda items of NSE, the D.G. concluded that Pricing Committee never went into the factors such as cost of infrastructure, man-power, and risk containment measures etc. while deciding upon the fee structure or waivers. (We must here itself note that the Pricing Committee which was deciding upon the policy of pricing for the CD segment does not seem to have taken into consideration the advent of the Act. In fact, after 30th of March, 2009 when it had issued the last circular, when section 4 was promulgated on 20th May, 2009, it was expected to take into account the effect of the zero pricing, particularly because it was then not the only player in the market and the only other player in the market was the MCS-SX, which had no other business to do excepting the CD segment. Very strangely the Pricing Committee does not seem to have taken this into consideration).

32. The D.G. also examined the pattern of the fees charged by way of admission fee and deposit level waivers. It is already noted that for CD segment there was no admission fee or deposit level waivers or requirement of making any deposit. The D.G. noted that the NSE was charging this admission fee for all other segments. As regards the deposit level waivers, the D.G. noted the arguments by NSE that the requirement of deposit levels was made keeping in line the nature of the segment in terms of the risk associated and the other factors. The D.G. noted that earlier deposit required for CD segment could not be said to be unjustifiably low. It was found by the D.G. and observed that NSE had reduced deposit structure w.e.f 28.11.2008, which was of necessity followed by MCS-SX from January 13, 2009. Thus, it was in this sector also that the NSE had initiated the lowering down of the deposit levels.

33. As regards data feed fee waiver, the D.G. noted that this was waived right from the beginning. Consequently, MCX-SX was also not in a position to charge the fee. The D.G. noted that the same reasons were forwarded by the NSE in respect of this waiver also. The NSE had tried to justify that its subsidiary DotEx was providing data feed service in various forms and was not charging any fee for the CD segment and therefore NSE did not charge its clients. The D.G. noted that in respect of other segments, the NSE was charging a substantial fee for data feed. On this backdrop, the D.G. concluded that DotEx had waived the fee with the purpose of capturing the market. It was noted by the D.G. that the DotEx was 100% subsidiary of NSE. The D.G. also noted that the waiver of data feed was not discussed during any of the board meetings over the initial 16 months from the date of commencement of trading in CD segment and it was first time discussed only later on.

34. The D.G. went on to analyze the predatory pricing by NSE. In this the D.G. took into consideration the definition of predatory pricing and the 2009 Regulations for determination of cost production, which can be referred to as cost regulations. After considering the implication of various terms like costs in Regulation 3(1), the DG took stock of the argument by the NSE that it was not incurring any variable cost for running the CD segment and therefore, the zero pricing could not amount to predatory pricing within the meaning of section 4 of the Act. The D.G. asked a very relevant question, that being, if NSE was not having any other segment to support income, could it survive with this zero pricing policy in respect of the CD segment and noted that answer would be obviously in the negative. The D.G. also considered the argument from the NSE that this policy was in the nature of introductory or penetration pricing, which has no objective of ousting or reducing the competition. The D.G. however, observed that even in the introductory/ penetration pricing, there had to be an element of pricing. The NSE argued before the D.G. that the variable cost under the circumstances was zero and since this cost was zero (approximately) therefore, no pricing policy, could not be said to be a predatory pricing policy. The D.G. observed that the NSE could run operations in the CD segment only due to substantial fixed cost, which it has already incurred for all the segments. If the pricing of any segment is to be linked only to the variable cost, NSE would have zero pricing for all the segments, because none of them would have any variable costs. The D.G. held that the investigation had already established that this claim of NSE was not substantiated by the facts. The D.G. had also referred to the report of the RBI SEBI. The DG also took into account the statement of Director (Finance and Legal) that additional expenditure was incurred for machinery, manpower, IT support, disaster recovery etc. in respect of the CD segment system. It was also admitted that surveillance system for the CD segment was also set up. It was also admitted that there were many dedicated employees for the CD segment and NSE paid substantial amount to these employees and therefore, the D.G. came to the conclusion that the contention of NSE that none of these costs constitute variable costs could not be accepted. Various views taken by the international jurisdictions were considered by the D.G. including the US Department of Justice, DG Competition of European Union etc. Various other discussion papers on EC Exclusionary Abuses were also taken into consideration. All the concepts like average variable costs, AAC, long run average incremental cost were taken into consideration by the D.G. The D.G. relied on the following observation of the European Commission :- If the price are below average total costs but above average variable costs, those prices must be regarded as all abuses are determined as a part of the plan for eliminating a competitor. The D.G. then went on to hold that there was a strong justification for following ATC or at least AIC in the instant case for determining predatory pricing in the relevant market of stock exchange.

35. Undoubtedly, the NSE had been asked about the details of allocation of all fixed and variable costs for the CD segment for the last two years. However, very significantly NSE submitted that it did not prepare accounts in which separate profit and loss account statements are provided for either the CD segment or any other segments. The NSE tried to justify this stand by saying that there were difficulties in allocating common costs across a multiple products firm. However, the D.G. examined certain trends in the balance sheets and provided a profit and loss accounts of NSE. From the investigation, it was pointed out that there was a quantum increase in fixed assets in general and IT hardware/ software, since the CD segments started in particular after financial year 2007-08. Previous to that, the increase in the fixed assets was only Rs.31.472 crores, however, there was an increase of Rs.133.671 crores. During 2008-09 a further increase was of Rs.93.475 crores and further in the following year it was Rs.90.1 crores. In respect of refusal by the NSE to provide segmented costs, the DG considered the details of overall capital costs, expenses, segment-wise long run incremental cost (LAIC) and established the effect on the costs subsequent to the start of CD segment. The D.G. observed that the total cost for 2008-09 worked out to Rs.4.42 crores and 2009-10 it came to Rs.31.07 crores. The D.G. distributed the total cost of NSE on a pro-rata basis for all the segments that the NSE was dealing with. The D.G. also estimated the depreciation of Rs.5.63 crores during 2009-10. The D.G. also noted that NSE had conducted several seminars, workshops and road shows for promoting operations in CD segment including 1163 promotional activities in 103 locations across India, the expenditure of which was not provided in the details of expenditure.

37. The D.G. also examined the pattern of clearing and settlement charges incurred by NSE. These activities were executed by the NSE through NSCCL, which is wholly-owned subsidiary of NSE. It was found that for other segments like F&O and equity, the NSCCL was charging NSE at 15% of the transaction charges in equity charges. The D.G. therefore, held that transaction charges amounted to a variable cost linked to the volume of transaction. The D.G. also observed that the NSE Board by a resolution in June 2010 enhanced clearing of settlement charges in the F&O segment, showing the clear strategy for loading the settlement charges for the CD segment on to the F&O segment. After considering the issue of notional clearing and settlement charges for the CD segment at 15% of transaction charge, the D.G. came to the conclusion that the expenditure could be notionally Rs.13.74 crores, payable to the NSCCL for the periods from August 2008 to April, 2010. The DG also came to the conclusion that MCX-SX was operating only in the CD segment and its operating expenses could be no different from the expenses of NSE. The D.G. on this basis rejected the theory of NSE that it did not incur any variable costs. The D.G. also relied on the judgment of Ontario Supreme Court of Canada in Regina vs. Hoff Mann La Roche Ltd. On this basis, the D.G. concluded that the waiver of transaction charges, data feed charges, admission fees and the reduction of deposit levels by NSE amounted to the actions violating section 4(2)(a)(ii) of the Competition Act.

38. In addition to above, the D.G. held that NSE had used its dominant position for leveraging and thereby it is guilty of contravention of section 4(2)(e) of the Act. For this purpose, the D.G. took into consideration the share of NSE in F&O, the equity and WDM segment, which were 100%, 75% and 90% respectively. The D.G. held that NSE was using this profit, which it earned to leverage this position in the CD segment, wherein MCX- SX was competing with it and this it was doing by not charging transaction fee, data feed fee etc. The D.G. seems to have relied on few cases like Tetrapak II case and Deutsche Post AG (DPAG)/ United Parcel Service (UPS) case, where the strategy of cross subsidies from other business activities was found to be anti-competitive by the European Commission. On these accounts, the D.G. found NSE guilty of contravention of section 4(2)(e) of the Act.

39. Ultimately the D.G. concluded that the acts on the part of NSE have harmed competition in the Indian Capital Market particularly in the CD segment. The behaviour of NSE is clearly exclusionary and the facts indicated that such acts were done with intent to impede future market access for potential competitors and to foreclose existing competition. The D.G. also held that this anti-competitive conduct enhanced the harm as the relevant market of the stock exchange services is a network effect of market.

40. This report was forwarded to the CCI, which directed the service of the report to the Opposite Parties No.1 and 2 for filing their reply/ objections. Some additional submissions were also made by the Informant, which were also forwarded to the Opposite Parties. Some other applications filed by the Informant were also directed to be served to the opposite parties.

41. The Opposite Parties No.1 and 2 filed their main reply along with annexures. Thereafter several letters and submissions were filed on various dates. The Informant also filed their preliminary submissions as regards to the D.G. report. Further written submissions were also filed by the Informant, while the Opposite Parties No.1 and 2 also filed additional written submissions.

42. The Opposite Parties relied on the reports submitted by the Genesis Economics Consulting Pvt. Ltd. (Genesis) and Prof. Richard Whish, Professor of Law at Kings College. Similarly, the Informant relied on the reports of their economic consultants, LECG Ltd.

43. The CCI has neatly and in great details noted the contentions raised by the Opposite Parties No.1 and 2 as also the objections to D.G.s report. So also it noted the legal and economic objections raised by Opposite Parties No.1 and 2. It took into account its details and noted the same in the impugned order. The CCI also noted the counter submissions of the Informant on D.G.s report and the submissions of NSE on the D.G. report. We need not deal with them as the further part of the judgment would be devoted to consider those objections, which were not only raised before the CCI, but before us also. Shortly stated on the basis of the rival contentions, the CCI came to frame the following issues :- (a) What is the relevant market, in the context of section 4 read with section 2(r) and section 19(5) of the Competition Act, 2002? (b) Is any of the Ops dominant in the above relevant market, in the context of section 4 read with section 19(4) of the Competition Act? (c) If so, is there any abuse of its dominant position in the relevant market by the above party?

44. In its final judgment, however, there was a division of opinions. The Chair-Person with three other learned Members has passed the final order holding NSE guilty of the breach of section 4(2)(a)(ii) and 4(2)(e) and has inflicted the penalty @ 5% of the average turnover of Rs.1110 crores amounting to Rs.55.50 crores. On the other hand, the two learned Members have written a separate order disagreeing with the conclusion drawn in the majority order and have held that no violation of any of the provisions of section 4 has been established against the NSE. Thereby, the two learned members have completely exonerated the NSE. In addition to inflicting of the penalties, the majority order has also issued certain directions under section 27(a) as also under section 27(g). These directions are :- (1) to cease and desist from unfair pricing, exclusionary conduct and unfairly using its dominant position in other markets to protect the relevant CD market; (2) to maintain separate accounts for each segment with effect from 01.04.2012; (3) to modify its zero price policy in the relevant market and to ensure that the appropriate transaction costs are levied, which action was directed to be taken within 60 days; (4) The NSE was directed to put in place system that would allow NSE members free choice to select NOW, ODIN or any other market watch software for trading on the CD segment of NSE. This was directed to be done under the overall supervision of SEBI, if necessary. For this NSE was also directed to ensure all cooperation from DotEx or Omnesys. Before we proceed, we must put here that this fourth direction is of no consequence, as there has been a compromise in this behalf before the Honble Bombay High Court. The parties also did not address us in respect of this aspect.

45. Marathon arguments went on before us by the learned counsel who appeared in this matter and possibly every view point was canvassed vociferously before us. It is on these rival contentions that we now proceed to decide the matter. Relevant Market

46. According to the D.G. the geographical relevant market was India, the product market was the services offered by the stock exchange. There is no difficulty about the geographical market being of India, as both the sides, as also the two deferring judgments by the CCI agreed on that proposition. The question is about the product market. According to the NSE, this market should be the market of currency futures as also the Over The Counter (OTC) market. It is a common knowledge that the OTC market is used by the hedgers, who want to cover their risk. The hedgers include those who have to satisfy the claims in foreign currency immediately or in future. Even the banks cover their risk by hedging on the OTC market. Shri Sibal, arguing for the appellant very forcefully submitted that in order to start a new segment of capital market in India, namely exchange traded currency derivatives segment, a report was brought into existence, which is RBI-SEBI Standing Technical Committee Exchange Trade Currency Futures report (RBI-SEBI report) in 2008. Shri Sibal very heavily relied on the following excerpt in this report :- Exchange traded futures as compared to OTC forwards serve the same economic purpose, yet differ in fundamental ways. The counter party risk in a future contract is further eliminated by the presence of Clearing Cooperation. Further in an exchange traded scenario where the market lot is fixed at a much lesser size than the OTC market, equitable opportunity is provided to all classes of investors whether large or small to participate in the futures market. Shri Sibal therefore, argued that since the exchange traded currency futures and OTC, serve the same economic purpose, the relevant market should be the market of the currency future along with OTC. The CCI in both the judgments did not agree with this contention. In so far as the OTC market is concerned, the CCI discussed it thoroughly. The CCI pointed out that the CD market was futures derivative market where underlying securities are the currencies. OTC market, however, includes various products such as forwards, swaps and options for hedging the currency risks. Functionally, the products may be considered as similar, but according to CCI they are quite different in terms of characteristics as well as participants. The CCI found that there was differentiation from the OTC market in terms of settlement on maturity, settlement period, counter party risk, size of market lot and participation, amongst other things. The CCI also noted the major difference that the CD segment had maximum maturity of 12 months, whereas OTC forwards could be for much longer durations. While considering the participation, the CCI noted that the equity and equity derivative segments or WDM segment were essentially for investors or speculators who seek to gain from price movements of equities. It noted, however, that the OTC segment was basically for importers and exporters having contractual exposures and who try to hedge their risks emanating from fluctuations of exchange rates. The CCI also noted that OTC products are not traded on exchanges and only specified entities can participate in this market and since the CCI was looking at a case where the Informant and the Opposite Parties are both providing stock exchange services, a product that is not being traded, cannot be said to be a part of any market the two are operating in. The CCI also considered the SSNIP tests and found it to be unnecessary in the circumstances. For this purpose, the CCI relied on the US Horizontal Merger Guidelines 2010, which has held that the SSNIP test was solely a methodological tool for performing hypothetical monopolist test for the analysis of mergers. The CCI also referred to the Official Journal by European Commission and came to the conclusion that the reliance on the test was unnecessary. In that the CCI also referred to the small and insignificant transaction fees and other fees. The CCI also refused to go into the interchangeability or substitutability of the products. The CCI therefore, rejected the plea that the OTC market should be included in the relevant market.

47. Shri Sibal, the learned counsel appearing for the appellant, very seriously urged that if the SSNIP test is to be considered as the applicable test, then 5-10% increase in price for the service of CD is unlikely to drive purchasers of CD contracts to purchase equity. However, in the very next breath Shri Sibal urges that such a price increase could, however, drive users of the CD segment to the OTC segment. Now these are the contradictory arguments. Particularly when all the speculators and the players in the CD market are not interested in the OTC market in which only the hedgers and arbitrageurs are interested and all the speculators cannot be imagined to be the hedgers and arbitrageurs. Shri Sibal, therefore, urged that the CD and OTC segment form part of the same relevant market. When we consider the Genesis report as also the opinion of Prof. Richard Whish, which were heavily relied upon by him, we find the thrust of that report is not so much on OTC market, as it is to canvass that all the stock exchange services like CD futures, F&O, WDM and securities cannot be covered under the relevant market. The Genesis report as well as Prof. Richard Whish have laid considerable stress only on that point. There does not appear to be any relevant discussion about the OTC market being the part of the CD market. We therefore, reject these reports at least in so far as the first question is concerned, as to whether the OTC market could be included in the CD market. It has already been considered by the CCI and we also consider that the two have a complete different complexion apart from the platform where the OTC and CD are traded. The OTC market essentially comes under the regime of the banking laws and would be restricted to the hedgers and arbitrageurs. There would be no scope from the speculators in that market. We therefore, uphold the finding of the CCI that OTC market cannot be a part of the market for CD segment.

48. Now we consider the other finding on which both the judgments are unanimous. The CCI seems to have relied on the further part of the aforementioned report, particularly in para 5.2 of Chapter 5 where a clear separation of CD segment from other segments in any recognized stock exchange where other securities are also traded is given. It also relied on the further stipulation that the trading and the order driven platform of the CD segment must be separate, as also the membership of the segment must also be separate and the CD segment must have a separate governing council. It also recommended a rigid arrangement to the effect that no trading/ clearing member should be allowed simultaneously to be on the governing council of the CD segment and the cash/ equity derivatives segment. The CCI also referred to Chapter 7, where it was mentioned to begin with, FIIs and NRIs would not be permitted to participate in currency futures markets. After mentioning about the entry of MCX-SX in the market and the fact that MCX-SX was only permitted to operate in the CD segment, the CCI deduced three factors (1) that in the minds of policymakers, the CD segment was not only completely different from other segments but also differed from OTC in fundamental ways, and therefore the policy recommended strict segregation of the CD segment; (2) till 2008 the exchange capital market in India did not have exchange traded currency forwards segments; and (3) competition concerns, if any, have to be examined in the segregated and new market where the Informant is operating. The CCI therefore, held that the exchange traded CD market was fundamentally distinct from other segments of the capital market. In fact, it did not exist prior to August 2008. The CCI therefore, deduced that a market which earlier did not exist and which was consciously created by the policy makers as a new and distinct market, cannot be said to be part of a market that existed. Ultimately, it came to the conclusion that the CD segment being a distinctive and separate market, the relevant market in this case should be the services offered in CD segment.

49. When we consider the findings of the minority order on this issue, we find that the minority order has referred to section 2(r), 2(s) and 2(t). The minority order has also in its determination and more particularly at para 7.6.1 noted that the exchanges only provide the infrastructure (platform) for such products to be traded subject to regulations, rules, by-laws and operative procedures. It concluded, therefore technological support and the facilities provided by the exchanges, which results into easy execution, lower cost of transaction, efficient risk of management, fail-proof settlement mechanism. It mentions that a robust infrastructure mechanism with enhanced technological support definitely adds volumes necessary for the development of the market. It however, mentions that merely because several products are traded in different segments of the same stock exchange and are categorized as exchange traded products, they do not lose their product differentiating features or their identity as representing different asset classes with different target customers/ consumers. It, therefore, held that both the exchanges and the securities traded are the external trappings while, the real substances lies in the classes of assets that are traded as underlying. It then jumped to the conclusion that the relevant market in the present case is currency derivatives segment of stock exchange services. It is, therefore, clear that we would have to concentrate on the majority judgment for examining the correctness of the ultimate finding that the relevant market was related to the CD segment. We have already referred to the three deductions, first being about the policy; second being about the CD segment being introduced only in 2008; and thirdly the Informant operating only in the market of CD segment. In our view, the third deduction about the Informant operating only in the CD market is irrelevant. After all when the judgment was written in June 2011, a third player had also been added right from September 2010, that was United Stock Exchange (USE). Though it had lesser market share as compared to NSE and MCX-SX, in fact USE has started in September 2010 with highest market share of 45.53%. It started losing its market share gradually with sporadic gaining the market share upto June 2011. Again nothing depended upon the Informants being engaged only in CD segment. In our view, the CCI committed an error in relying on this factor. There was after all no guarantee that the other exchanges would not step into and it actually happened much latter when even Bombay Stock Exchange also joined the CD segment, somewhere from November 2013, during the pendency of this Appeal. When we consider the second factor that the CD segment started only in 2008, that in our view again would be an irrelevant factor. Merely because the CD segment started in 2008, would not make it a distinct market. Lastly, even the first factor about policy, to say the least is inconsequential factor. The policy did not show that CD segment was totally and completely different. All that it says in para 5.2 that a recognised stock exchange where other securities are also being traded may set up a separate currency futures segment. It has then suggested the three factors (1) the trading and the order driven platform of currency futures should be separate from the trading platforms of the other segments; (2) the membership of the currency futures segment should be separate from the membership of the other segments; (3) the currency futures segment should have a separate Governing Council on which the representation of Trading/ Clearing Members of the currency futures segment should not exceed 25%. Further 50% of the public representatives on the Governing Council of the currency futures segment can be common with the Governing Council of the cash/ equity derivatives segments and the Exchange; (4) The Chairman of the Governing Council of the currency futures segment shall be a member of the Governing Council. If the Chairman is a Trading Member/ Clearing Member, then he shall not carry on any trading/ clearing business on any Exchange during his tenure as a Chairman; (5) No trading/ clearing member should be allowed simultaneously to be on the Governing Council of the currency futures segment and the cash/ equity derivatives segments We have gone through practically the reports supplied to us, para 5.3 describes the eligibility criteria for Clearing Corporation of the currency futures segments, while para 5.4 speaks about the separation from other segments of the Clearing Corporation. In our view this report merely considers the safety aspect and the insulation of possible disputes due to the interlinked interests of the exchanges and/ or the officials. In fact, the report and more particularly para 1.2 specifically hints at the difference between OTC and the currency futures. The report is a complete mechanism on as to how the CD segment would work and the separation procedure provided in para 5.4 is more or the less is a safety mechanism. We fail to understand, as to how this report by itself could be relied upon by the CCI to hold that currency futures is a different product. The Informant MCX-SX has filed before us the report of internal working group on currency futures. The findings in their internal group do suggest that OTC segment could not form a part of the same relevant market as the CD segment.

50. We have now to consider whether the CD market by itself could be a separate and distinct market. We must here note that a fundamental error committed in treating the CD segment as a product by itself. In fact, at one place the majority order had defined the relevant product market as stock exchange services in respect of the CD segment. Now, if the stock exchange services were common, then there was no need to restrict these stock exchange services in respect of CD segment alone. The fundamental error that was committed by the majority and minority order was that it says that it assess the relevant market focused on the products being traded on stock exchange as opposed to the services, which are offered by the stock exchanges. It must be understood here that a stock exchange does not manufacture, offer or sale any product. It simply offers a trading platform and associated services for brokers to use. The market for assessment therefore, has to be the services offered by stock exchange independent of the product being traded on that exchange because a stock exchange does not sell a product. It must be borne in mind that a stock exchange does not create products like WDM, F&O, securities and currency derivatives. It merely offers the services. The competition assessment has to be therefore, only in respect of the services offered by the stock exchanges irrespective and independent of the products traded on the stock exchange. The learned counsel for the MCX-SX rightly argues that SEBI allowed the trading on stock exchanges of (1) equity; (2) debts; (3) futures; (4) options; and (5) currency derivatives. All the stock exchanges provide trading services in respect of these products, though at the relevant time, the MCX-SX was providing the service only in CD segment. When the competition question comes, it would have to be understood, as to in what manner and what conditions these services were offered by various stock exchanges including NSE and MCX-SX. The very existence of the institution of stock exchange is for providing services to the speculators, brokers and all those interested in those products. Therefore, what is important is a service not the segments in which the stock exchanges deal. A beautiful example came to be cited by Shri Haksar. He compared the stock exchanges with the firm doing the business of dry- cleaning. He pointed out that in a competition between the two dry- cleaners, the only relevant factor would be the services given by the dry- cleaners in dry-cleaning the clothes, whether it be shirts or coats or pants. According to him, it cannot be imagined that one dry-cleaning firm is cleaning only shirts, or only pants, or only coats. The example is extremely apt. It must be realized that the nature of the product does not affect the services, and the competition law assessment can and should be done only with respect to the services being offered, especially when the enterprises concerned do not have any control over the products being traded because the products do not belong to them or nor are they created by them. Shri Haksar points out that in the present case, the NSE did not waive the price of the product being traded on its platform, but simply waived the fee for the services offered by NSE. He also gave another example of a card room. A person operating a card room simply provides the premises, a pack of cards, and various tables for the players to play. All these items would be the services offered by the proprietor of the card room. Whether the players play poker at one table and bridge at another, does not take away the fact that proprietor of the card room is simply providing certain services to the card players. A competition law assessment between the two proprietors of the card room would, therefore, be based only on the services offered by each of them and not based on the card games that the players playing inside each of the card rooms. What the players wish to play at any time is determined by the players, not by the proprietor of the card room, and the similar things take place at the stock exchanges. A stock exchange provides certain services to the participants (i.e. broker) on its platforms. Whether a broker uses its services for trading in shares or currency derivative, does not affect the nature of services provided by a stock exchange for competition law assessment. In our opinion the argument is infallible.

51. It was also heavily argued by NSE that considering the definition of section 2(t), the relevant product market must comprise of all those products/ services, which are viewed as an interchangeable/ substitutable by the consumer. According to him, this implies that only demand side substitutability, namely what consumers consider as interchangeable, that according to him is the only relevant consideration for determining the relevant product market. He further argues that what suppliers found to be substitutable (i.e. supply side substitutability) is not a factor to be considered in determining relevant market for the purpose of the Competition Act. He has given the example that a person going to Mandi to purchase onions will not find wheat seller at the Mandi to provide an adequate substitute. He therefore, argues that as long as onions are not substitutable for equity/ F&O/ WDM, the products (and consequently their service) cannot form part of the same relevant market. According to him in paragraph 10.24 of the majority order seems to have agreed with this argument. Before proceeding further, we must hold that what the consumer wants from the Mandi seller, is a definite product like onion or potatoes or grains. He does not go to the shopkeeper for his services. He has a definite product in mind. Such product is absent in the present scenario. If a person wants to purchase or deal in shares, he only uses the services of the stock exchange. He may in the process purchase few shares, but those shares are not the products. The product is the service offered by the stock exchange for getting either shares or F&O or WDM or CD derivatives. Therefore, the example is incorrect. There is tangible product in this example in shape of onions or potatoes or grains. Such tangible product is absent here. This is apart from the fact that a person purchasing CD need not restrict himself only to CD, he may have a choice to deal with the securities or WDM or F&O. It is always a broker, who deals and broker need not restrict himself only to the product of CDs. He can deal with any other product, provided he has the license to that effect. It is, therefore, that we say that the argument about supply side and demand side is irrelevant in this matter for the simple reason that it is a question of service being offered to the customer in this case either the broker or the speculator or anybody. Merely because a broker could get a service of CD from a separate platform that does not become a whole relevant market. It was tried to be argued by Shri Sibal that of the twenty largest trading members by volume in NSE CD segment, only three are also amongst the top twenty traders in the equity and F&O segments. Very strangely, the learned counsel relied on this data for canvassing that the two are separate markets. We fail to follow. Even if the three persons out of the top twenty persons are dealing both in CD segment, equity, F&O segments that is enough to buttress the point that it is the service of the stock exchange in all the sectors, which would be a relevant market. The learned counsel has also relied on the percentage, the 7.6% of the trading volumes of the CD segment. In our opinion, this argument must be rejected as inconsequential. The learned counsel also argued on the basis of SSNIP test and contended that if there was non-transitory increase in price of 5-10%, it was unlikely to drive the purchasers of the CD contracts to purchase equity. In the same breath, however, the learned counsel urges that such a price increase could, however, drive users of the CD segment to the OTC segment. We do not agree. It may be that if a transaction fee was charged by the NSE in CD sector, the concerned broker might stop dealing in the CD sector altogether and might turn to the other segments as he has to remain in the business. Therefore, in our opinion, SSNIP test would be of no consequence. We must again realize that in section 2(t) there is a separate mention of the products or the services. Therefore, the two concepts cannot be confused with each other. What are we concerned here, are not the CDs, futures derivatives for CD or the shares or the WDM or F&O. We are concerned with the price of a service, which is offered by NSE. The real issue was as to whether the NSE in offering the service had abused its dominance. While interpreting section 2(t), we cannot ignore section 2(u), the definition of service, which suggests as under :- 2(u) service means service of any description which is made available to the potential users and includes the provision of services in connection with business of any industrial or commercial matters such as banking, communication, education, financing, insurance, chit funds, real estate, transport, storage, material treatment, processing, supply of electrical or other energy, boarding, lodging, entertainment, amusement, construction, repair, conveying of news or information and advertising. The latter part of the definition though restricts itself to the factors mentioned, it cannot be ignored that basically service means 'service' of any description. The only condition is that 'service' must have been made available to the potential users. In this case the 'service' is used by the brokers, speculators and other players of the stock exchanges, which the stock exchanges offer. It provides the platform to such person for giving the service. Now, it is not as if a person dealing in the stock exchange service for CD, he would not utilize that service in the other sectors, that only depends upon on his will. Therefore, what is relevant here is the service. It cannot be further restricted to what the CCI has done by treating service only to the CD segment. That is clearly impermissible.

52. Our attention was invited to the 2nd Genesis report and more particularly to para 2.1.4 of that report. It was tried to be shown that the international case precedent had consistently found that equities are in a separate market to currency derivative based on differences from a demand-side perspective. The first such example was merger of TSX Group Inc and Bourse de Montreal (2009). A quotation is used in that case equities, derivatives and commodities had distinct risk profiles; as such, demand substitutability was limited. The Bureau concluded that these instruments were not competitive substitutes with one another and examined these three grouping separately.

53. Second example given was Australian Stock Exchange and SFE Corporation Ltd. (2006), where it is suggested that the merger between an equities exchange and derivatives exchange was approved on the basis that there was no product substitutability, implying a lack of demand substitutability. The Australian Competition and Consumer Commission (ACCC) also found that supply side substitutability was unlikely in practice due to network effects and that could arise as a result of liquidity requirements. While giving the third example the report takes into consideration the Deutsche Borse AG, Euronext NV and London Stock Exchange. A quotation of the Competition Commission of UK was quoted there to the following effect derivatives, equities and bonds are really substitutable from the purchasers point of view. The UK Competition Commission also found that there was no room for supply side substitutes based on the facts that a platform would incur non-trivial cost over a year or more in order to start trading in another product. Furthermore, they noted the limitation on economies of scope arising from power in adjacent markets. Equities derivatives and bonds are typically traded on separate platforms, suggesting that economies of scope are not strong enough to warrant inclusion in the same market. The fourth example is that of Eurex. In this case the European Commission observed the following, despite the connections to the markets for listing and trading services for securities, on which Parties operate, and the derivatives market, on which Eurex operates, there are substantial differences between the two markets.

54. All these cases are merger cases apart from the fact that the law laid down by foreign competition authorities, are not binding on this Tribunal. It must be said in respect of these cases that, being cases relating to merger, the law laid down will not be applicable to the present case, where the relevant product is a service offered by the security exchange. After all a service offered by NSE was complained of, as they refused to charge anything for that service and thereby had tried to bleed MCX-SX slowly, since MCX-SX did not have a license for any other segment except the CD segment. If the services offered by NSE without charging anything was the cause of the complaint by MCX-SX, then we would strictly have to take into consideration the service aspect. The service offered by NSE in the matter of currency derivatives, would be no different than the service offered in the other segments in which it was operating. It has already been shown that it was not necessary that a person taking the service in currency derivatives would not take that service in the other segments. On the other hand, it was clear from the statistics that there were number of persons, who were utilizing the service in the other segments also. In fact, it is an admitted position that out of the biggest twenty players, as many as three players were utilizing the service of NSE in the other segments. Therefore, while dwelling upon to decide the relevant market, this aspect of service alone cannot be ignored and it would have to be held that the service offered by the security exchanges would be the relevant market.

55. We must not ignore the fact that for the purpose of defining relevant market in a case relating to abuse of dominance, there is no international precedent in construing different services offered by the stock exchanges as separate relevant market. Secondly, even if a separate relevant market is found, as it has been found in the aforementioned four cases, that has been found only in the cases of mergers and joint-ventures. Thirdly, the present case is not the case of merger approval, where for defining relevant market, each service has to be compared with the competitors service, so that a single player does not start to dominate after getting merged with another entity. In our opinion, therefore, it would be irrelevant to rely on the decision holding the relevant market for merger cases for being used in the case under abuse of dominance. A holistic picture would have to be taken into consideration and in our opinion the D.G. has correctly held the relevant product market to be the services. It must also be noted that the merger analysis is an ex-ante review of the proposed merger and to examine whether the proposed merger will significantly alter the structure of the market and impact the participants in the market in the futures. It is an assessment frequently based on an assessment of probabilities and likelihood of certain types of behavior arising from the merger. Therefore, it is natural that the definition of relevant market for assessment becomes as narrow as possible to evaluate the impact of merger in the future. The consideration is that if the proposed merger does not significantly and adversely alter the structure of the narrowest possible relevant market, then it can be justifiably concluded that the proposed merger may not have an impact on competition in future. This is in sharp contra distinction with a review in connection with abuse of dominance case, which is an ex-post facto review. The competition authorities in an abuse of dominance case review and assess the conduct of an enterprise in the past of an event, which has taken place already or is continuing as on the date of investigation/ order. This, therefore, becomes a static analysis of an event, which has occurred between the two points in time. Thus, a merger analysis is a prospective assessment, while the abuse of dominance position assessment is effectively an assessment of the conduct that has already taken place. Shri Haksar very earnestly urged this proposition for which he relied on extracts from Bellamy and Child. Shri Haksar also analyzed the aforementioned cases cited by NSE and contended that in the case of merger of TSX Group Inc. and Bourse de Montreal Inc., the Canadian Competition Bureau had to define the relevant market as services offered in respect of each segment to assess any potential overlaps that may raise competition concerns. This depended on the specialization agreement, which played critical role in this assessment. He also commented that in Deutsche Borse AG, Euronext NV and London Stock Exchange, the assessment focused on services offered in the equity segment because the target enterprise i.e. London Stock Exchange operated primarily on the equity segment and the derivatives segment constituted only 3% of the business transacted on the London Stock Exchange. Shri Haksar explained that due to insignificant business of the target company i.e. London Stock Exchange in the derivatives segment, the UK Competition Commission confined its assessment of the potential impact of the merger to the equity trading segment. That was because the business of the derivatives segment was insignificant and the proposed acquisition of the London Stock Exchange could not raise any competition concerns in UK. He also pointed out that the parties in this case were ad idem on the narrow market definition.

56. Alternatively, Shri Haksar contended that there are various similarities between F&O and CD segments. He brought out as many as ten similarities, they being :-

(i) All derivatives markets in India are cash settled no one gets an actual share or foreign currency by entering into derivatives contract. He only gets the difference in cash which is common for F&O and CD. F&O and CD contracts are very similar as borne out by following.

(ii) A consumer in either segment does not have any ownership right to an underlying security or currency, but only gets a contractual right to the difference in prices.

(iii) CD & F&O Contracts are unique to the exchange which have launched it and can be closed out only on the same exchange; unlike securities purchased in Equity Segment, which can be sold in any other exchange.

(iv) Nature of derivatives contracts available for trading in both segments are very similar futures and European style options.

(v) F&O segment regulations adopted as it is for CD segment by NSE, pursuant to a specific permissive clause in Chapter VII of RBI-SEBI Technical Committee Report.

(vi) SEBIs Master Circulars issued on December 31, 2010 clearly club F&O and CD segment contracts in single circular, while Equity is segregated in a separate Master Circular of same date.

(viii) Consumer view: Statement of a very sophisticated global consumer i.e., Bloomberg, as extracted in page 81 and 82 of DG Report.

(ix) Trading member, clearing member classification only relevant for F&O and CD segments.

(x) Trading parameters, margins and settlement modes for derivatives (similar for both F&O and CD) are very different from those for cash trades.

57. Shri Haksar, therefore, urged that if at all CD market is to be considered as a separate relevant market then it should also be considered along with F&O market, in which obviously the NSE has a forceful presence. It is not necessary for us to go to this aspect, as we are convinced that the relevant market in this case is that of the services offered by the stock exchanges. In that view, we confirm the finding of the D.G. on this issue. We would not, however, comment on both the majority and minority orders, which has accepted the CD segment as the relevant market, for the simple reason that even if that market is to be held as a relevant market, in our opinion, the majority order was correct in deducing that the NSE is a dominant player in that market.

58. It need not detain us, if the relevant market is taken to be the services offered by the security exchanges then there would be no question of the NSE not being a dominant player. We would separately consider the arguments of the parties on the aspect that the relevant market was the CD segment alone and the NSE was a dominant player, even if that market alone is considered as the relevant market.

59. This takes us to the issue of dominance. We have to consider whether the NSE is a dominant player in the relevant market. We have already clarified that even when the relevant market was defined narrowly as being the market of currency derivatives alone, by both the judgments of the CCI, the majority order held NSE to be a dominant player. We would, therefore, consider as to whether the NSE was rightly held to be dominant in that judgment of the CCI, even when the relevant market was construed narrowly to be the market only for currency derivatives. We must at this juncture point out that before holding the CD market to be the relevant market, the CCI separately considered the various aspects of equity market, F&O market and WDM market. Lastly, it also considered the CD market and the OTC market. The CCI then went on to record that equity and equity derivatives segments or WDM segment were essentially for the investors or speculators, who seek to gain from price movements of equities, while the OTC segment was basically for importers and exporters having contractual exposures and who try to hedge their risk emanating from fluctuations of exchange rates. The CCI then went on to record that the CD segment primarily for speculators of currency values and short term hedgers, who want to cover their economic exposure, but require greater liquidity. Then the CCI in its majority order went on to reject the SSNIP test, holding that it was merely a methodological tool for performing hypothetical monopolist test for the analysis of mergers. It then referred to the Cellophane Fallacy. The majority order also refused to go into the extended debate to distinguish the words interchangeable or substitutable, considering the facts of the case and different aspects of capital market in India. The CCI held this to be unnecessary and not useful. It acknowledged that equities and currencies were entirely different and consequently related derivatives were also different. It therefore, went on to hold that the currencies and equities were related to the different market. It acknowledged that from any practical point of view, a product over CD segment exchange could not be said to be either interchangeable or substitutable by a product in segments like equity and F&O. We have also indicated above that we do not subscribe to this view because, this view is predominantly a view treating equities and currency derivatives to be the products, which itself is not a correct notion. The majority judgment then went on record to say that the stock exchange services in respect of the CD segment is clearly an independent and distinct relevant market. Though, the majority order in para 10.24 agreed that the DG had found a high degree of commonality amongst the members of the MCX-SE and NSE and held that this itself had no bearing on interchangeability or substitutability between various segments of stock exchange services. It wrongly gave an example of wholesale traders of grains and wholesale trading of vegetables, completely ignoring that in that example, the products were tangible products of grains and vegetables. It is on this basis that the majority order came to the conclusion that relevant market was CD segment in India.

60. Though, we do not see this as an absolutely correct finding, all the same we will have to consider, as to whether the majority judgment was correct in holding the NSE as a dominant player in the narrow market of CD segment. The following factors, which were found by the DG or mentioned in the majority judgment are stated as under :- a. In the equity segment of stock exchange services in India, NSE has continuously held high market share for the past 8 years going beyond 71% in 2008-09. b. In the F&O segment, NSE has almost 100% market share. c. In WDM segment, NSE has maintained more than 90% market share for the past 6 - 7 years. d. Putting together equity, F&O, WDM and CD segments, NSE have garnered 92% market share as of 2008-09. e. In CD segment itself, NSE has a market share of 48% according to the DG report. f. NSE has been in existence since 1994 as against incorporation of MCX-SX IN August, 2008. g. As at 31.3.2009, reserves and surplus of NSE stood at Rs. 18.64 million, deposits at Rs. 9.17 billion and profit before tax at Rs. 6.89 billion. h. In comparison, BSE had a net profit of Rs. 2.6 billion only and MCX-SX carried forward net loss of Rs. 298.7 million for the period ending 31.3.2009.

i. NSE has presence in 1486 cities and towns across India. BSE has presence mainly in Maharashtra and Gujarat and is now reduced to mostly operating in equity segment. MCX-SX has only about 450 centres and operates only in CD segment. j. NSE has high degree of vertical integration ranging from trading platform, front-end information technology, data information products, index services etc. k. Stock exchange services in India are highly regulated and require approvals of SEBI to start a new exchange.

61. According to the majority judgment, these factors which were undisputed created a complete picture of players in the capital market in general and in the relevant market of currency derivatives. The order then touches upon history relating to the first half century of independent India, in which BSE was way ahead of all the regional stock exchanges, but only before the entry of NSE on the scene, which soon became the market leader. The CCI then referred to the entry of MCX-SX and USE in the CD segment and noted that by the time these two players made entries, NSE had already occupied an overall position of strength. The order then notes that there were only three players in the market, i.e. NSE, MCX-SE and USE and referred to their current percentage, to be 34% MCX-SX, 30% NSE and 36% with the latest entrant USE as of October 2010. The judgment then mentions that it is these three players which would have at least some ability to affect its competitors or the relevant market in its favour, even if it is not capable of operating completely independent of competitive forces or affecting consumers in the relevant market. The CCI majority judgment then went on to refer to section 19(4). It mentioned in this behalf that the position of strength has to be arrived at after rational consideration of relevant facts, holistic interpretation of seemingly unconnected statistics or information and application of several aspects of the Indian economy. It mentions that what has to be seen is whether a particular player in a relevant market has clear comparative advantages in terms of financial resources, technical capabilities, brand value, historical legacy etc. to be able to do things which would affect its competitors, who in turn would be unable to do so or would find it extremely difficult to do so on a sustained basis. The reason is that such an enterprise can force its competitors into taking a certain position in the market which would make the market and consumers respond or react in a certain manner, which is beneficial to the dominant enterprise but detrimental to the competitors. On this backdrop, it referred to in Explanation (a) to section 4 of the Act and then proceeded to consider whether NSE had a position of strength, which enabled it to affect MCX-SX as a competitor in its favour.

62. It firstly asked itself a question (1) can NSE sustain zero pricing policy in the relevant market long enough to outlive effective competition? It answered this question holding looking at the financial statements of NSE, its reserves and surplus or its profits after tax, it cannot be argued that the capacity of NSE to defer profits or to bear long-term risk of possible market failure is lesser than that of MCX-SX in the relevant market. According to it this was clearly a position of strength.

63. The second aspect that came for consideration was whether there was any indication that the conduct of NSE showed that it was aware of its capability? The CCI noted that NSE had not followed Accounts Standard 17 (AS17), which stipulated the segment reporting. The CCI in the majority order rejected the facile explanation that the so called detachment of profit motive was with the desire to develop the CD segment for the larger good of the capital market in India. The CCI rejected this explanation as unpalatable. It then mentioned it is unthinkable that a professionally managed modern enterprise can afford such financial complacency in the face of competition unless it is part of a bigger strategy of waiting for the competition to die out. This complacence can only point to awareness of its own strength and the realisation that sooner or later, it would be possible to start generating profits from the business, once the competition is sufficiently reduced.

64. The third aspect, which was considered was whether in the absence of above strengths, would NSE be able to or want to continue with zero pricing indefinitely? The majority order answered this question holding that had NSE not got the undeniable advantages arising out of its operations in other markets, it would not have been able to or wanted to charge nothing for providing stock exchange services for the cash derivatives forwards market. It also noted that in this behalf MCX-SX or any other current or future competitor did not have similar advantages. From this, the majority order deduced that NSE enjoyed a position of strength in the relevant market, which enabled it to affect its competitors in its favour. It mentions that for arriving at this conclusion, the CCI have taken into account relevant aspect of the financial statements of the parties concerned, HHI index of more than 5000 in the CD segment (2009-10), ICR3 of more than 99 and other key indicators. The majority order said that it had also given consideration to some important cases from international jurisdiction, such as AKZO, [1978] ECR 207, United Brands [1991] ECR I-3339, Du Pont. From this, the majority order came to the conclusion that NSE had the position of strength and therefore, enjoyed dominant position in the relevant market.

65. In this behalf, when we consider the minority order, reference is made to the finding of D.G. that NSE held absolute dominance, even if CD market is assessed in isolation with other segments, on account of its incomparable economic power, size, resources, higher degree of vertical integration, absolute dependence of consumers and large degree of economies of scale in operating different segments with adequate scale in each of those segments. The minority order then went on to consider the argument by NSE that while it started with 100% of the CD segment market in October 2008, it did not enjoy the first position in terms of market share and this indicated that NSE was not able to operate independent of competitive forces, nor could it affect its competitors. It also noted the argument by NSE that MCX-SX had continued to increase its market share after entry, pushing NSE to second position and its market share had gone up to 60.47% by August 2010, when USE entered the market in September 2010. The respective share of NSE, MCX-SX and USE was 32.48%, 42.77% and 24.75%. The minority order then considered the position upto October 2010 of the respective market shares, which was taken from Genesis report dated 30th October, 2010. The minority order then went on to analyze the market shares. It deduced from this that the NSE at least on the basis of the market share could not be said to be a number one player. It disagreed with the deduction of D.G. as also the majority that NSE enjoyed economic power, which was reflected in its ability to maintain zero price over the long run and to sustain losses in the CD segment from other segments. It rejected the argument that all of these could be perhaps under different circumstances translated into a competitive advantage. It mentioned that however, in a networked industry, a new comer could have easily overcome the competitive advantage of the incumbent by offering innovative product with value added services. It also mentions that the CD segment did not drive any specific benefit from other segments of the stock exchanges. It also took stock of the further deduction by the D.G. about NSEs dominance on the basis of vertical integration of the enterprises. It also rejected the other deductions of D.G. about entry barrier.

66. The minority order then went on to consider the definition of dominance in section 4 and stressed at the words .in its favour. It noted that the market share of NSE, which was 100% in August 2008, had gone down to 33.17% nearly in two years time and this fact by itself according to the minority order was sufficient to show NSEs inability to influence the market or its competitors in its favour. It then went on to mention we are not aware of any case in the history of competition jurisprudence, where a firms market share has been reduced drastically (to less than one third in this case) in a relatively short period (two years in this case), and yet it has been found to be dominant by a competition regulator or a court.

67. In our opinion, the analysis by the majority order is correct and the analysis by the minority order, more particularly about the market share and the so called reduction thereof is not satisfactory. First, we will take up for analysis the market share aspect. It seems that the minority order has considered the position upto October 2010. This judgment is dated 03.06.2011. It is obvious that the minority order has restricted its findings to the whole period of two years. Even if we consider the market share for upto October 2010, as has been done by the minority order, it will be seen that right from August 2008 upto November, the NSE was a clear market leader. Again from December (2008), January and February (2009), its market share was almost equal to the market share of MCX-SX. From there again, its market share rose and it became number one from March 2009 to August 2009 i.e. the next six months. Then right upto December 2009, it was almost hand-in-hand with MCX-SX. In 2010, however, it started declining. Therefore, right upto the month of August or September 2009, in majority of months the market share of NSE was more than the market share of MCX-SX and when it was not so, it was fully comparable to the market share of MCX-SX. Shri Sibal pointed out that the period of 1st Circular when NSE charged zero price for the transaction fee was issued on 26th August, 2008 and was valid for about one month i.e. upto 30th September, 2008. The second Circular was dated 26th September, 2008 and was valid upto 30th November 2008. The third Circular was issued on 28th November, 2008 and was valid upto 31st March, 2009. All these Circulars undoubtedly are prior to 20th of May, 2009, when section 4 was for the first time activated. The culprit Circular is dated 30th of March, 2009, which is valid upto 30th of June 2009. When the Act came on 20th May, 2009, NSE was a clear leader, having 53.19% of market share as compared to the market share of MCX-SX being 46.81%. This position continued upto August 2009 i.e. for four months including May, when the market share of NSE was more than the market share of MCX-SX. Again for the next four months the market shares are almost similar with MCX- SX. We have it on record that the Circular dated 30th March, 2009, which was valid upto 30th June, 2009, was the last circular issued and thereafter it was continued right till August 2011, when for the first time, the NSE started charging the transaction fees. It is very significant to note that thereafter it was number two only upto December 2011 and from January 2012 it bounced back and continued to have the number one position excepting for the months of March and April 2012. Right till March 2014, its number one position continues in so far as market share is concerned. We are really concerned with not only the market share, but the other relevant factors considered by the majority order and more or the less ignored by the minority order. The most relevant question was as to whether the NSE could continue with its no transaction fee policy, as compared to its competitor MCX-SX. It is absolutely clear that while NSE could continue because of its relative strength in other segments, MCX-SX could not have continued with that policy and per force it had to adopt the same policy of no transaction fees, as otherwise it could not have even entered the market, forget about its sustenance in that market. In this behalf, therefore, we feel that the consideration by the majority judgment is more realistic, objective and covers wider spectrum.

68. Shri Sibal while addressing us on the question of dominance invited our attention to Explanation (a) of section 4, as also the various provisions in section 19(4) of the Act. We must at this juncture note that the duty of the CCI while considering the position of dominance is to take into consideration all the factors under section 19(4). However, sub-section (4) also empowers the CCI to take any of the factors, considering that market share is but one factor, size and resources of the enterprise, economic power of the enterprise, vertical integration of enterprises or sale or service network of such enterprises are also some of the other relevant factors. It is obvious that the size and resources of NSE were tremendous. We need not go into the details thereof. In so far as vertical integration is concerned, NSE was the richest amongst all the stock exchanges. Its economic power can also not be disputed and indeed it was not disputed excepting that it was argued that MCX-SX was also a very big company. In our opinion, though Shri Sibal argued on all these aspects, the leading position of NSE cannot be disputed. Same thing can be said about the market structure and the size of the market as covered in clause (j) of section 19(4) of the Act. Our attention was invited by Shri Sibal to LECG Report and the alleged admission therein at page 2370 that NSE would probably not be dominant in the market that the Genesis Reports define, whether or not that market includes OTC products. However, further contents in the same paragraph are worth noting :- We agree that NSE would probably not be dominant in the market that the Genesis Reports define, whether or not that market includes OTC products. However, this does not rule out the possibility of abuse. NSE could be dominant in other markets from which it is leveraging market power onto a market for trading services in currency derivatives. The evidence shows that NSE is indeed dominant on such markets. We find that if separate markets are defined for stock exchange services in the trading of each of cash equities, equity derivatives and wholesale debt products, then these markets will be closely associated with the provision of stock exchange services in currency derivatives. NSE would be dominant in these markets for cash equities, equity derivates and wholesale debt products, and it may then be able to leverage its dominance from these markets onto the market for currency derivatives. This strategy would allow NSE to protect its existing dominant positions by preventing MCX-SX from growing to a size where it could challenge NSEs existing dominant positions.

69. We have already shown that firstly, the market could not be CD segment alone and even if it is CD segment, even then NSE would be a leader. Shri Sibal is undoubtedly right when he argues that while applying the factors listed under section 19(4) of the Act, a check-the-box approach should not be followed and the factors in that section should only be considered as an aid in assessing dominance. However, even if we consider those factors, as is suggested by the learned counsel, the inevitable result would be that NSE clearly emerges as a leader. Shri Sibal argues giving the examples of Ashok Leyland and TATA Motors in the truck market or Coca Cola and Pepsi in the soft drinks market. He urges that in both these markets, the players have high market shares and immense resources at their disposal and that they cannot be considered dominant only on those factors. The argument is undoubtedly correct. However, we hasten to add that in this present scenario MCX-SX or the other players did not have the advantage of having the complete domination over the other segments like securities, WDM and F&O. It was also argued by the learned counsel that the finding of dominance could not be lightly given since that finding imposes a special responsibility on enterprises to ensure undistorted competition in the market. The learned counsel argues that by that finding one hand of the dominant enterprises remains is tied behind its back and it has to compete with such restraint in the relevant market. He is undoubtedly right that a dominant player in the market has a responsibility, an added one, because of its dominance. However, if that responsibility were to be voiced against NSE, the results were bound to be adverse to NSE. The learned counsel again reiterated his argument that the NSE could not be said to be in a position of strength, since it could not operate independently of the competitive forces. The learned counsel pointed out that even when NSE did not charge for the transaction fee, its competitors were able to match and successfully outmanoeuvre NSE and as a result the market share of NSE fell from 100% to 33.6%. We have already commented on the market share percentage aspect to suggest that, that is not be-all and end-all of the matter, apart from the fact that the overall picture suggests that even if the CD segment is held to be the relevant market, NSE still was the leader for the major part of the time span. It is further urged that NSE could not affect its competitors in its favour, as BSE, MCX-SX and USE successfully entered the CD segment and were able to sustain zero pricing for periods longer than NSE. We do not agree with this contention, as firstly it is misnomer to say that MCX-SX successfully entered the market and were able to sustain themselves. It was because MCX-SX has suffered huge losses when the transaction fees was not being charged at all by NSE and consequently by themselves. Shri Sibal also pointed out that now because of the introduction of the transaction fee pursuant to the impugned order by the CCI, the consumers are now paying for the services, which they were getting free. He also pointed out that the trading volumes also halved leading to lower chances of consumers finding a counterparty at the price they wished to trade at. This is also not a complete argument. There may be and we are sure there are many other reasons for the decrease in the trading volumes. The introduction of the transaction fees by itself cannot be the only reason for that purpose. Shri Sibal also criticized the observation of the CCI in paragraph 10.34 of the majority order and contended that the majority should have found none of the players in the CD segment to be dominant. We are unable to agree with this argument for the reasons already quoted above. Shri Sibal further argues that in the majority order the strength of NSE was considered and that should not have been done. He gave an example that any rich person entering any market would be considered to be dominant. He gave an example of Google entering into bicycle manufacturing market and pricing its bicycles below those of its competitors. The argument is defective, as the size of NSE and its strength is not considered in isolation, it is only one of the factors to make it as a dominant player in the relevant market.

70. Shri Sibal then relies on the observation of the minority particularly in para 8.2.6. We have already pointed out that the consideration by minority of the position upto October 2010 was not correct and it should have taken an overall picture into account. The learned counsel then again argued on all the factors including the market share. We have already pointed out and analyzed the market share aspect in this case. In our opinion, the argument of market share is clearly flawed. Shri Sibal takes an exception on the historical aspect considered in majority order. In our opinion the majority was absolutely right in considering that aspect. He has also addressed us on the profitability aspect, national presence of NSE, degree of vertical integration and entry barriers etc. The argument about the profitability aspect considered in majority order is that the USE and MCX-SX were new ventures and therefore expected to have a lower level of profitability. Then the argument is put about the strength of USE and MCX-SX and their values. All these details in our opinion are unnecessary. We have nothing to do with the strength of BSE and FTIL and the worth of MCX-SX. Even if MCX is a big company, in the relevant market as we have held, the strength of NSE is unmatched. The vertical integration was also commented upon by the learned counsel and he contended that the competitors in CD segment, namely USE and MCX-SX are closely linked to the extensive infrastructures and capabilities of their promoters. He urges that MCX-SX promoter MCX operates a clearing house, while its other promoter, FTIL markets a trading software ODIN. In our opinion, all these arguments are of no consequence, particularly when it is clear that in so far as the infrastructure is concerned, NSE is clearly a leader. There is again no point in describing the strength of MCX, which is only a promoter company of MCX-SX, nor is the strength of FTIL in any manner is relevant in the present controversy. The learned counsel further argues that the fact that MCX-SX and USE have been able to enter and operate in the CD segment suggests that there are no entry barriers. The fact that MCX-SX and USE have entered the market cannot be a relevant fact by itself for giving a finding on entry barrier. That would have to be tested on the policy of NSE of not charging any transaction fees and it will have to be considered as to whether in such a situation would any new player chose to enter the CD segment. The answer of which would always be in negative. This argument is opposed by the MCX-SX by pointing out that for entering into the CD segment there are some onerous conditions put in by SEBI, which alone is authorized to allow for starting of stock exchange. In short, to start a stock exchange is not a small exercise. That also goes in favour of the argument about the entry barriers. The learned counsel also severely criticized the three questions asked by the majority order to itself and contended that these questions lacked economic logic. In our opinion, the majority order committed no error in asking the three questions. It has to be considered that it is not only on the basis of answers to those three questions that finding about the dominance of the NSE has been arrived at.

71. Shri Sibal then contends that the majority order focused only on one competitor, rather than the category of competitors. He secondly argued that the financial resources of all the players were not correctly ascertained. Thirdly, it was urged that it was wrongly considered that NSE had a better capacity to defer the profits in comparison to MCX-SX. Lastly it was urged that the test of the deepest pockets was of no consequence. In our opinion, all these arguments must be rejected. It is incorrect to rely on a stray observation in a order. The stray observation was in respect of the MCX-SX and that leads to the argument that majority order considered only MCX-SX. In fact for first few months there was only one competitor and probably that is why the majority took into consideration the position of MCX-SX vis--vis NSE. We have already commented upon the so called financial power of the MCX company, which was the promoter company of MCX-SX. It is irrelevant as to who backs the company. What is required to be considered is the company by itself and its strengths. Lastly, it cannot be said that the MCX-SX was in better position than NSE to sustain the losses and to keep on sustaining them.

72. Shri Sibal criticized the reliance placed by MCX-SX on the judgment of the European Court of Justice in Michelin vs. Commission ([1983] ECR, 3461). In this case, it was held by the European Court of Justice that Michelin had a worldwide strength while ascertaining whether it was in a dominant position in the relevant market of Netherlands. Shri Haksar criticized this reliance by saying that in that case, the worldwide strength of Michelin was taken into consideration as in the opinion of the Court that strength gave the distinct advantage over its competitors in the relevant market i.e. Netherlands. Shri Sibal then urged that this judgment is in apposite for the present controversy. In our opinion, the judgment applies on all fours. In fact, the overall strength of NSE clearly gives it a leverage to be benefited in the CD segment. The other observations in Michelins case are extremely relevant. They being about the temporary losses suffered by the dominant player.

73. Shri Sibal argued that the present figures regarding the market shares should not have been taken into consideration and only the market shares at the time when the alleged abuse took place should be considered. He is undoubtedly right and hence we have seen that in first few months, the NSE clearly emerged as a leader on account of its policy of zero transaction fees. Shri Sibal from time to time relied on the minority order and quoted various paragraphs, they include paragraphs 8.6.2 (vii), 8.2.8 as also paragraphs 8.2.7 of the minority order. We have already considered these paragraphs and have given our reasons why we do agree with the views expressed in those paragraphs. We are not impressed by the other arguments regarding the speech of Shri Narayansami, on which heavy reliance is placed by Shri Sibal. He has also raked up the controversy about the alleged fraud in the MCX-SX by Mr. Jignesh Shah, Vice-Chairman and Mr. Joseph Massey, Director. We find all these references unnecessary in the present controversy. Shri Sibal also commented on the judgment in United Brands vs. Commission [1978] ECR

207 on which MCX-SX had relied, which suggested that even if United Brands had 40-45% market share, it was held to be a dominant player in the market, because of its other strengths and advantages like ability to control production, know-how, access to raw materials, large capital, brand loyalty, transportation facilities etc. In our opinion, the judgment is rightly relied upon by the MCX-SX, as it is already pointed out that NSE enjoys many more factors as its strength, which have already been discussed in the earlier paragraphs. We are also not impressed by the interview of Shri T. Narayansami. We are also not impressed by the arguments about the strength and the worth of MCX and FTIL. We have already rejected that argument in the earlier part of the judgment. In short, we are convinced with the finding of the majority order of the CCI that NSE was in a dominant position.

74. That leaves us to decide as to whether in the present circumstances, it has abused its dominant position. Lengthy arguments were addressed on this question. Shri Sibal argued that in order to find the abuse of dominance three tests have to be satisfied, they being - (a) dominance; (b) pricing below cost; and (c) intent. We have already expressed that first aspect about the dominance. Considering the second aspect of pricing below cost, according to Shri Sibal, the term cost must not be confused as used in the general parlance. He says that the manner of calculating costs under the Competition Act is governed by the Determination of Cost of Production Regulations, 2009 and more particularly Regulation No.3 of it. He points out that under this Regulation, the concept of Average Variable Cost (in short AVC) has to be considered. According to the learned counsel, the language of explanation (b) to section 4 of the Act itself suggests that in case of predatory pricing, cost would be as determined by the regulations. He, therefore invites our attention to the Determination of Cost of Production Regulations, 2009 and relies on Regulation 3, which suggests that the Cost in the Explanation to section 4 of the Act shall, generally, be taken as average variable cost, as a proxy for marginal cost. There is a proviso to it, which is as under :- Provided that in specific cases, for reasons to be recorded in writing, the Commission may, depending on the nature of the industry, market and technology used, consider any other relevant cost concept such as avoidable cost, long run average incremental cost, market value. He further points out that the marginal cost is a cost for producing one additional unit. He gave an example of a manufacturing concern that in a factory manufacturing steel, there would be fixed costs, like rent paid for the use of the factory, salaries of employees, etc. and these costs have to be incurred irrespective of the amount of steel that is produced. While variable costs for the same factory would be iron ore and other raw materials, electricity etc., that is a costs which would vary with each additional unit of steel produced. He also takes an example of an ice cream and suggests that the additional costs of milk, flavouring, sugar etc. required to produce an additional scoop of ice cream. According to him, the test for predatory pricing through the Cost Regulations ensure that parties who are pricing low (a consumer benefit) are doing so in a way that an equally efficient competitor could also provide the same product at such low cost. According to him in the present case, this would amount to the resources required to facilitate one extra trade on the CD segment. According to him, since the marginal cost, is not always readily observable, therefore, AVC is to be used as a proxy. The learned counsel further argues that since in case of the stock exchanges, it is nearly impossible to determine the costs involved in facilitating one extra trade, hence AVC is a generally accepted cost benchmark to determine whether an enterprise has engaged in predatory pricing. According to the learned counsel in case of NSE, the AVC was zero and hence in not charging any transaction fees or in other words charging zero price for transactions did not amount to predatory pricing. We will have to test this claim of the average variable cost being zero in case of CD segment. In our opinion, it is impossible to support this conclusion.

75. The average variable cost is defined in the Regulations at 2(1)(b), which is as under :- 2(1)(b) average variable cost means total variable cost divided by total output during referred period;

76. A reference may be made to the definition of total cost, which is defined in the Regulations at 2(1)(c)(i), it is as under :- 2(1)(c)(i) total cost means the actual cost of production including items, such as cost of material consumed, direct wages and salaries, direct expenses, work overheads, quality control cost, research and development cost, packaging cost, finance and administrative overheads attributable to the product during the reference period;

77. The definition of total variable cost is as under - 2(1)(c)(ii) total variable cost means the total cost referred to in clause (i) minus the fixed cost and share of fixed overheads, if any, during the referred period;

78. These definitions would suggest that for arriving at the 'average variable cost', first, one would have to arrive at a total variable cost. Now total variable cost in the present case would be the total cost as defined in Regulation 2(1)(c)(i) minus the fixed cost and share of fixed overheads, if any, during the referred period. We must appreciate that in this case, the average variable cost is claimed to be zero. This situation can come only by first arriving at the total variable cost being zero. In our opinion and there is good evidence available to suggest that the total variable cost cannot be zero in order to justify the zero available variable cost. It is only when the total variable cost is zero, then that zero cost is divided by total output cost. It need not be explained that when zero is divided by some figure, the result has to be necessarily a zero. Such is certainly not a case here. This situation could have come only and only, if total variable cost comes to a zero. It has not been shown by NSE that the total variable cost in this case was a zero or in other words, there was no total variable cost. The figures of the total variable cost have been provided by NSE in their confidential version. They had to show that their total variable cost was zero as their total cost minus the fixed cost and share of fixed overheads came to zero. It is only and only in this situation that their average variable cost could be zero. The most persuasive argument of Shri Sibal as well as Shri Naval Chopra Satarawala could not convince us on this aspect. In our opinion, it is not necessary for us to go to the concepts of total avoidable cost and average avoidable cost or even long run average incremental cost, in the facts of this particular case, where the appellant NSE has not been able to convince us that the figure of total variable cost comes to zero. In fact, it is on this short ground that one can lead to the conclusion that NSEs claim that it did not at all have any total variable cost, in the sense that its total cost minus the fixed cost and share of fixed overheads came to zero. We, therefore, straightway reject this claim of the NSE that since it had zero average variable cost, therefore, they were justified in introducing a zero transaction fees policy.

79. If this justification of introducing zero transaction policy fails, as we have shown it must, fail, then we will have to examine the conduct of NSE in introducing the zero transaction fee policy in August 2008 and then despite the advent of section 4 of the Act, continuing with the same to an indefinite period. In this behalf, we would have to consider as to whether this policy of zero transaction fee was an outcome of the genuine concern on the part of NSE to collect liquidity and also whether the NSE had taken notice of the advent of section 4 of the Act, which NSE must have known. It cannot happen that a player like NSE would not know of the activated section 4. In this behalf, we must refer to the Pricing Committee meetings of the NSE. As per the records, on 10th June, 2008, 2nd meeting of the Pricing Committee was held in which a proposal was mooted to setup CD segment and waiver of levy of transaction charges for a period of three months, from the date of commencement of trading. We are prepared to accept the position that this was obviously done in order to encourage the liquidity in the market, after all the market was nascent at that time. We do not have the advantage to know as to what transpired in the third meeting of the Pricing Committee and it was only on 26th November, 2008 that the 4th meeting of the Pricing Committee was held. Minutes of this meeting also speak that the NSE wanted to encourage the participation in trading in currency futures and it had paid special attention to the fact that the exchange traded currency future was still in the nascent stage. In that view, it decided to extend the waiver of fees upto 31st March, 2009. On 3rd March, 2009 again the Pricing Committee sat to consider its pricing policy, where it specifically considered the extension of waiver of admission fees for the new members in CD segment, as also considered the waiver of levy of transaction charges in respect of the trades done in currency futures. By this, it considered the proposal to extend the transaction fee waiver upto 30th September, 2009. There is also a 6th meeting of the Pricing Committee alleged to have been held on 27th August, 2009, in which again there was an extension of waiver of levy of transaction fees in respect of the trades in currency futures and it agreed to continue it upto 31st March, 2010. Strangely enough the Circulars issued by the NSE, however, do not tally with these minutes, because by the first circular, the valid period for the zero transaction fees was only upto 30th November, 2008. Second Circular came on 28th November, 2008, in which the NSE declared that it would not charge transaction fees upto 31st March, 2009. The last Circular seems to have come on 30th March, 2009 and in this Circular the zero transaction fees policy was to be continued upto 30th June, 2009, though, the minutes suggest that they were to continue the zero transaction fees policy upto 30th September, 2009. Very strangely, thereafter, in spite of the fact that there was another meeting held on 27th August, 2009, in which the Pricing Committee decided to extend the zero transaction fee policy upto 31st March, 2010, no such Circular seems to have been issued. At least that is not the claim made before us. In short the Circular dated 30th March, 2009, in which zero transaction fees was to be continued upto to 30th June 2009 (i.e. after 20th of May, 2009 when section 4 was introduced) merrily continued to rule the scenario and it seems that NSE in spite of its meeting on 27th August, 2009 of Pricing Committee, did not bother to issue any other Circular.

80. We must at this juncture point out that the minutes of both these meetings are completely silent about the effect of section 4 having been taken into consideration by NSE. We can understand about the meeting dated 3rd March, 2009, on that day the section 4 had not come. But what completely perplexes us is the studied(?) silence of the Pricing Committee in the meeting dated 27th August, 2009. In fact, anybody would expect that after the advent of section 4 the Pricing Committee would immediately sit up and take the notice of section 4, as on 20th May, 2009 there were only two players in the market, one was NSE itself and the second was MCX-SX. The NSE was well aware of the fact that MCX-SX did not have license for any other segment and were dealing exclusively in the CD segment. They were actually the competitors of NSE, even if it is considered that the relevant market was only currency derivatives. It was, therefore, clear that if the zero transaction fees was continued by the NSE, MCX-SX would have no other alternative but to also follow the suit, and that is precisely what happened. The MCX-SX had to continue with the zero transaction fees policy, if not anything else at least to meet the competition and in the process the MCX-SX in the very first year suffered the losses of Rs.29.81 crores, in comparison to NSE, which registered profit of Rs.515.54. It was known all over the world that MCX-SX do not have any other license for running the stock exchange in any other segments excepting CD segment and therefore, if they persisted with their zero transaction fees policy, which was a direct outcome of NSE's zero transaction fees policy, they would certainly bleed to death. That it did not happen, cannot be a justification for NSE to, firstly, totally ignore section 4 and to remain complacent with their own policy by firstly introducing the same and secondly continuing the same in spite of the advent of section 4. In our view, this was the best example of abuse. In this, one can analyse that MCX-SX could not have effectively competed with NSE on the basis of this zero pricing conduct. The data clearly suggests that the prices charged by NSE had the potential to foreclose MCX-SX, which was the only competitor in the field then, or for that matter any other competitor, who did not have the strengths of NSE. In our opinion, there is enough evidence to support this in as much as the losses suffered by MCX-SX kept on increasing. It was a well known fact that it did not have any other segments to deal with and it ultimately got the same somewhere after about two years, after a litigation.

81. On this factual backdrop it would be interesting to see D.G.s findings on abuse of dominance by NSE. The DG had considered the following four factors :- a. Transaction fee waiver; b. Admission fee and deposit level waivers; c. Data feed fee waiver; and d. Exclusionary denial of "integrated market watch" facility. While considering the transaction fee waiver, the DG took into consideration some previous history in view of the argument of NSE that the step of waivers were taken to encourage larger participation in the nascent CD segment. It was also argued before the DG that this policy was influenced by a report of the High Powered Study Group on Establishment of New Stock Exchanges and lastly it was pointed out by the NSE that it had constituted a pricing committee to guide and decide all pricing matters.

82. The D.G. considered the earlier such waivers by NSE in various other segments like equity, F&O and had also noted that the NSE had outstripped BSE by 2001. It was also noted by the D.G. that after outstripping BSE, NSE did not extend the waiver and the waiver in options in segments of F&O had continued till 2005. Even in respect of WDM segment, while the NSE had started trading in June, 1994, for one year, NSE had levied charges of Re. 1/- per Rs. 1 Lakh. From this the D.G. came to the conclusion that NSE did not have a historical philosophy of waiving fee to develop a nascent market. The D.G. also considered the conduct of NSE relating to Gold ETF segment where the transaction charges were levied from March, 2007 till August, 2009 when NSE was the only exchange trading in Gold ETF segment. However, it started waiving/reducing transaction fee only after February, 2010. This was probably to score over BSE which had entered into Gold ETF segment in September, 2009 and had grabbed a market share of about 5% which rose up to 19% by February, 2010. According to the DG this explained as to why NSE had introduced waivers/reductions from March, 2010 onwards. The D.G. then observed that the management of competition was the prime factor which influenced the transaction policy. The D.G. also considered the fact that NSE collected admission fee of Rs.5,61,800 in equity, F&O and debt segment but did not charge any admission fee in its CD segment and also did not collect any subscription charges and advance fee in the CD segment. The reasons for not charging given by NSE were also referred to. The other factors in respect of admission fee and deposit level waivers were also considered by the D.G. The D.G. also observed in his report that NSE had reduced deposit structure w.e.f. November, 28, 2008 which was subsequently followed by MCX-SX from January, 2009.

83. Same such observation was made by the D.G. in respect of the data feed fee waiver. It also considered the argument of NSE that the waiver in data feed fee was for the same reason for not charging transaction fee for the CD segment. The D.G. observed in this behalf that the issue of data feed fee was never discussed during any board meetings over the initial 16 months from the date of commencement of trading in CD segment.

84. As regards exclusionary denial of integrated market watch facility, we need not express anything here about the D.G.s observation in view of settlement of the concerned parties.

85. While analyzing the predatory pricing by NSE, the D.G. considered Regulation 3(1) of Cost Regulations and also further considered the concepts of variable cost. The D.G. also posed a question whether in a hypothetical situation of NSE not having any other segment to support its income, could NSE have survived in the wake of its waiver policy? The answer was obviously in negative. The D.G. observed the variable costs being zero and rejected this argument. The DG referred to the additional expenditure incurred by NSE for machinery, manpower, IT support, disaster recovery etc. in respect of the CD segment along with other factors like number of dedicated employees for the CD segment who were engaged by NSE and were paid substantial amount. The D.G. came to the conclusion that these costs did constitute variable costs. A reference was made by the D.G. to examine the views taken by the international jurisdictions while determining the appropriate cost and further observed that the relevant product market was on the basis of high level of network externalities. A reference was made by the D.G. to the case of European Commission namely Wanadoo Interactive SA (WIN).

86. The D.G. then noted that though NSE was asked to provide comprehensive details of allocation of all fixed and variable costs for the CD segment for the last two years, NSE had failed to do it on the ground that it did not prepare accounts in which separate profit & loss account statements are provided separately for either the CD segment or any of the other four segments. The NSE had also argued before the D.G. on the basis of the observations made by UK Competition Commission investigation wherein it was concluded that the allocation of common costs down to product level was impossible and would be misleading. The D.G. countered these arguments on the basis of the balance sheet and profit & loss accounts of NSE. The D.G. had noted that there was a quantum increase in fixed assets of NSE in general and IT hardware/software, since the CD segment started. It was noted that during 2006-07, the increase in fixed assets was only Rs. 31.472 crores, in comparison, the increase was Rs. 133.671 crores during 2007-08 that is after the starting of CD segment in 2008-09, it was Rs. 93.475 crores and during 2009-10, it was again Rs. 90.1 crores. The D.G. report had looked at the details of overall capital cost expenses, segment wise long run incremental cost etc. The D.G. also made estimation to the total costs for the CD segment working out to Rs. 4.42 crores for the year 2008-09 and Rs. 37.07 crores for the year 2009-10 which had been arrived at by the D.G. on the basis of pro rata assumption.

87. We are deliberately making a detailed reference to the D.G. report as even the CCI has completely relied on the same. It was argued before the CCI, firstly about the NSE not being dominant, secondly the transaction fee waivers were done in order to develop the nascent market, thirdly that NSE had historical philosophy of waiving fee in developing nascent market and fourthly that there was no element of predatory pricing as there was no variable cost and lastly that the charge of leveraging could not apply as NSE was not dominant in the CD Segment. The CCI firstly rejected the argument about the dominance and then considered the theory of development of nascent market. It discussed in detail the various stages of nascency of a market and came to the conclusion that the market could be nascent for the first few months but certainly not for ever or for indefinite period. According to the CCI, this waiver policy was a strategy and not a bona fide step for preserving or developing an otherwise nascent market. It considered the various occasions on which the notifications for waiving transaction fees was issued. The CCI also considered the historical background of waivers in case of equity segment and F&O segment and upheld the observation of the D.G. that NSE only after outstripping BSE, re- imposed transaction charges after it had surpassed BSE. Similarly, CCI also confirmed the D.G.s findings about WDM segment where after commencing the trading on 30.6.1994, it levied transaction charges for a full year till June, 1995. It confirmed the D.G.s observation that this conduct of NSE contradicted the claim of consistent policy of fee waivers to develop nascent market. On these grounds it rejected the defence regarding the development of nascent market.

88. It further noted the pattern of NSE behavior in respect of F&O segment and WDM segment as also Gold ETF segment and concluded that the historical conduct of NSE suffered from inconsistency and nothing could be derived from this behaviour patterns which would lead to the conclusion that NSE had consistently followed the philosophy of fee waivers in the nascent market. It, therefore, confirmed the D.G.s observation in this behalf. Similarly, the CCI had found no merit in the justification given by the NSE regarding data feed fee waiver.

89. It also noted about the NSEs failure to give separate segment figures to the DG.

90. It then referred to the fact that MCX-SX was operating only in the CD segment as also its financial statements of 2008-09 and 2009-10 showed that it was incurring variable costs. It also referred to the total expenses made in 2008-09 and 2009-10. The CCI, therefore, confirmed the finding that the CD segment did require some variable costs to be incurred by the stock exchange and it cannot be zero as claimed by the NSE. (We have already in the earlier part of the judgment shown as to how the theory of the average variable cost being zero was palpably wrong).

91. Thereafter, the CCI referred to the exercise made by the DG for allocating and estimating the costs. Thereafter, it referred to the zero pricing.

92. After considering the concept of unfair pricing, the CCI came to the conclusion that the D.G.s observations about the variable costs were correct. The CCI, therefore, came to the conclusion that the zero pricing was unfair as far as MCX-SX was concerned. The CCI then referred to the fact that NSE was in a position of strength which enabled NSE to resort to zero pricing since August, 2008 while MCX-SX did not have such strength. The CCI then observed - There is practically no justifiable reason for NSE to continue offering its services free of charge for such a long duration when it is paying for manpower and other resources for running the business. It is also a fact that no enterprise would have the intention to engage in a profit-less venture for eternity. The CCI then referred to the fact that MCX-SX had only CD segment and that was a major constraint and therefore the zero price policy of NSE was unfair. On this basis, the CCI came to the conclusion that the zero pricing policy was unfair. It then observed in para 10.77 that - In this case the conduct of NSE is beyond the parameters of promotional or penetrative pricing. It can, in fact, be termed as annihilating or destructive pricing.

93. We generally agree with the finding of abuse of dominance given by the D.G. as well as the CCI. We find no justification on the part of NSE to continue with the predatory pricing for unspecified period after 20th May, 2009. Much of the discussion has come as regards the predatory pricing. We need not repeat that again.

94. Shri Sibal, however, calls the pricing as unfairly low pricing. His submission appears to be that a zero price cannot be said to be predatory pricing at all. We have given our own reasons in the earlier part of the judgment to point out as to why a zero price would amount to predatory pricing. As a matter of fact, there is a lot of difference in the concepts of unfairly low price and zero price. In the former, there is at least some price (which is more than zero), which is charged by the enterprise, while in the latter there is absolutely no price and all the services required are given free of cost. While discussing the issue of cost in the earlier part of the judgment, we have already pointed out that there was no justification, whatsoever in the claim of NSE that it did not have any average variable cost, must fall to the ground. We have also pointed out, as has been done by the D.G. and the CCI that there was no justification for zero pricing. Shri Sibal argued that CCI has termed zero price as unfairly low price. We do not agree. For unfairly low price, there has to be some price somewhere, which is certainly more than zero. We, therefore, reiterate our finding that this was a clear cut example of predatory pricing. Without arguing as to what are the stringent tests for predatory pricing, the learned counsel merely said that they have not been met. Further arguments on these aspect regarding the concept of dominance, pricing below cost, intent are merely the repetitions of the earlier arguments. The learned counsel of the Appellant has given us a short written submission consisting of 211 pages, most of which are repetitions.

95. In the common parlance, predatory pricing means the pricing which is essentially below the cost and has an intent of destroying the competition in the market. We see no different meaning of this term. The whole argument centers around the point that this was an example of unfairly low price. We do not agree. In fact, it is a blanket examination of predatory pricing. A comment was made about the four pieces of evidence relied upon by the CCI for giving finding of immediate and demonstrable harm to competition, they being :- (a) NSE has deeper pockets that MCX-SX; (b) There is no justifiable reason for NSE to continue charging zero price for a service that costs money to run; (c) As a result of zero price, MCX-SX is slowly bleeding to death; (d) The unfairness of price is accentuated by the inequality in available resources between NSE and MCX-SX.

96. The counsel again repeats about the theory of deeper pockets. We have already shown that the financial power of MCX, the promoter of MCX-SX is irrelevant in this situation. It is again repeated that the so called zero pricing was with an idea to increase liquidity in the market. We have already pointed out that, that may be so in the beginning, but there was no justification to continue with the same policy after 20th May, 2009. Frequent references in these written submissions have been made to the entry of BSE and USE. In our opinion, USEs entry and exit and BSEs entry in the market would go to prove nothing. Very strangely, a reference is made that the MCX-SX has followed the zero price policy in the other segments it has entered. Thereby, the learned counsel wanted to draw our attention at the facts, which were much posterior to the relevant period. The learned counsel relied on a singular factor that USE decided to enter the market and ensured that the zero transaction fee policy did not have the exclusionary effect. We do not agree. We have already given our reasons in the earlier part of the judgment. The learned counsel then relied on the Genesis exclusionary conduct note dated 19.02.2011 to say that by their calculations MCX-SX would have taken 80 years to quit the market. This was very seriously doubted by the other side. Be that as it may, it is not a question as to when the competitor would have exited the market, once it is proved that the competitor was suffering losses year after year. The learned counsel tried to convince us that the zero price had its own benefits. It is then argued that the unfair pricing was liable to be considered qua all the concerned parties namely competitors, consumers, the relevant market and NSE itself and then it should have been balanced against the alleged unfairness qua MCX-SX. We do not agree with this proposition, so long as we are convinced that the zero transaction fee policy was a well intentioned policy to oust the competitors from the market. Much has been stated about the sentence that this policy was unfair to MCX-SX. Shri Balbir Singh on behalf of CCI argued that even if a statement is made only about MCX-SX, firstly, it was not only against the MCX-SX that the matter was being considered, the reference to MCX-SX came because the MCX-SX being the competitor was an Informant also. He also pointed out that when NSE was proved to be a dominant player in the market, it was not supposed to be abusive and if it targeted the only competitor in the market for first few months, that too consciously after the advent of the section 4 would still be an abuse of breach of section 4. We do not consider that the finding of the CCI was only with regard to MCX-SX and that any comparison as such was made only with MCX-SX. Though, the learned counsel made much of this so called unfair only vis--vis MCX-SX, we are of the clear opinion that the finding of the CCI is general in nature and not only vis--vis MCX-SX. Much was said about the financial power of MCX-SX and its expenditures even in the other segments. We do not think all that is relevant in the present controversy.

97. The learned counsel also urged that this could not amount to an unfair conduct vis--vis the consumer. It has been repeated again and again that in applying predatory pricing, though initially the enterprise suffers the losses, it then can start charging higher prices vis--vis the consumers. Much was said about the effect of the transaction fee being charged. We do not think that it is in any way relevant to the present controversy. Much was said about the fact that selling at cost was not predatory, but was merely an example of forging profits and that other competitors also could have chosen to follow the same price policy by forgoing their profits. We have already shown that NSE had nothing to lose in this, whereas MCX-SX had no other segment except the CD segment and would have incurred tremendous losses, which it actually did. The learned counsel urged that there was no evidence to state that NSE planned to continue the zero price strategy for eternity. We do not think the learned counsel is right in that behalf. After the last Circular in which zero transaction fee policy was to last only upto 30th June, 2009, thereafter no Circular came and the zero price policy continued till the impugned order of CCI. The learned counsel also urged about the actus reus and the cause of action. In our opinion, the cause of action started right on 20th May, 2009, when the zero transaction fee policy was in operation and thereafter it continued everyday during the continuation of that policy.

98. The learned counsel criticized the reliance by MCX-SX at AKZO Chemie vs. Commission C-62/86. All that we say is that in AKZO, there was at least an allegation that the price was below AVC. Here is an example where there is no price whatsoever. In this behalf Shri Balbir Singh for CCI and Shri Haksar for MCX-SX has rightly relied on AKZO and the observations made therein.

99. In paragraph 174 of the written submissions made by NSE, there are repetitions of the earlier argument that NSE had not priced below cost. The learned counsel tried to point out that NSE had provided extensive financial cost data for its CD segments for the financial year 2008-09 and 2009-10 in the First Genesis Report and in the Second Genesis Report. We have already given our reasons why we are unable to rely on these reports, particularly on the claim that there was no variable cost. We have given our reasons during our discussion on the Cost Regulations. In that behalf, we fully agree with the finding of the D.G. as well as the CCI. Much was said about the management accounts. We are not impressed by that argument at all. The fact of the matter is that NSE had failed to submit its statutory audit in respect of the CD segment on the special plea that it did not maintain the same. The other arguments are merely repetitions of the earlier arguments on this aspect. The other arguments about the costs are again repetitions of the earlier arguments. Once we have rejected the theory of Genesis that the variable cost was zero, in fact the argument should end there itself.

100. It is urged that NSE did not intend to reduce the competition or eliminate competitors. The learned counsel argued that the standard of proof should be the civil standard of preponderance or balance of probabilities. We agree with the learned counsel. We also agree with the learned counsel that the finding of predatory pricing should be on the basis of strong evidence. Once it is accepted that the civil law standards would apply, then the D.G. as well as the CCI were quite justified in handing out that finding on the basis of the circumstances discussed earlier. The facts in this case are telling themselves. In fact, it was for NSE to justify the zero transaction fees policy. It has miserably failed to do so. Two orders of Securities Appellate Tribunal (SAT) and one of SC were relied upon. They being Sterlite Industries (India) Ltd. vs. SEBI, Appeal No.20/2011 (Order dated 22.10.2001) and Dilip Pendse vs. SEBI, Appeal No.80 of 2009 (Order dated 19.11.2009) as also Mousam Singha Roy vs. State of West Bengal, (2003) 12 SCC 377. The two judgments of the SAT speak about the standard of proof, while the Supreme Court judgment deals with the murder case. We have nothing to say about the laid down standards, but we fail to see, as to how the judgments apply here. A judgment of Competition Appeal Tribunal (CAT) of UK was also referred to, where the UK Tribunal held that the infringements were not classified as criminal offences and had further expressed that the issues regarding relevant market, dominance, abuse etc. have to be decided on the basis of economic data and also the conflicting expert evidence. That Tribunal also held that the burden of proof in civil matters applies in such matters. We have already shown earlier that in this case the facts were plain and simple. We have also made reference to the minutes of the Pricing Committee. We have also seen that a player like NSE chose to ignore the advent of section 4 on the legal scenario. We are left with no doubt that the whole exercise to continue with zero price policy was deliberate. Very interestingly, an argument was made on the basis of a decision by the Ontario Supreme Court in R vs. Hoffman La Roche Limited (1981) 33 OR (2d) 694 that while Hoffman La Roche engaged in zero pricing in response to a new entrant, unlike in the CD Segment where NSE imposed a zero price when it was the only player for pro-competitive and profit maximizing motives. We fail to see any rationale behind this argument. Question was not on the date of imposition. The relevant question is of continuation of such policy indefinitely. It is further argued that there was no foreclosure from the market as was in the case of Hoffman La Roche. We do not have to wait for the actual foreclosure to happen, even if there is any possibility of the competition being affected that is sufficient for the purpose of section 4(2). In this case, it is quite clear that the idea was to oust MCX-SX from the market, which was the only competitor. Very strangely a reference has been made to suggest the bona fides on the part of NSE to NSEs Pricing Committees regular assessment of the nascency of the market after for the alleged period of 3 years was vindicated on the levy of the fee. We have already made extensive comments on the minutes of the Pricing Committee and we need not say anything more. We do not know on what basis it is then contended that the market continued to be nascent. In short, we do not accept the six points, which are as under :- a. All documentary evidence available on the record evidence good intent; b. NSEs Pricing Committees regular assessment of the nascency of the market after for the alleged period of 3 years was vindicated on the levy of a fee; c. Economic analysis also shows that the market continued to be nascent; d. Promotional or penetration pricing is a legitimate business strategy, especially in stock exchanges; e. NSE was meeting its competition; and f. NSEs past conduct in other segments have been grossly misinterpreted and in fact demonstrates that NSE did not have any anti-competitive intent. In our opinion, the concept of nascency is extremely fluid.

101. As regards the documentary evidence available, the learned counsel relies on the Pricing Committee agenda and minutes. We have already made our extensive comments on the relevant meetings. Indeed if Pricing Committee was and we believe it comprised of experts, we would have expected it to take notice of the advent of section 4 of the Act before deciding to continue with the zero transaction fees policy. We see an admission to that effect, which is as under :- NSEs Pricing Committee could not have predicted a brightline test as to when the CD Segment would be mature as such a consideration requires an assessment of a host of factors including the size of the CD Segment, the growth of the CD Segment, the current and projected volatility of the Rupee, state of the Indian economy, etc. What the Pricing Committee can do, and the evidence on record shows that they did so, was consider the status of the CD Segment when they made their decisions and determined that, at that time, the CD Segment was not mature enough to levy a fee; Strangely enough, it does not even distinctly mention about the effect of promulgation of section 4 of the Act. The other contentions on these Circulars are mere repetitions and we have already considered them earlier.

102. A reliance has been placed on so called letters by Mr. S.B. Mathur and Mr. Vijay Kelkar. We have mentioned these letters only to be rejected as initially the introduction of zero transaction fee policy may be justified, but its continuance after 20.05.2009 was certainly not justified.

103. The learned counsel argued on the good intention of NSE and suggested that NSE had not destroyed any evidence. That may be so, however, the intent of NSE was apparent enough and we have given our reasons for holding so.

104. The learned counsel argued that since the market had not become mature, they continued with the transaction fee waiver policy. This is wholly incorrect. In this, there is no evidence to show that NSE ever took the promulgation of section 4 of the Act into account as also the other factors like MCX-SX not having any other segment than the CD segment. The learned counsel tried to argue about the facts which are posterior to the passing of the order of the CCI about the MCX-SX entry into Equity and F&O segment. In our opinion, this argument is irrelevant. The written submissions mention about some other stock exchanges offering transaction fee waiver to develop the market like Chicago Mercantile Exchange and NYSE Matchpoint. In our opinion, these facts are irrelevant, as the facts in this case are entirely different. It was then urged that the transaction fee waivers are a legitimate new market entry strategy. That may be so, but the continuance thereof in the wake of the available facts was wholly incorrect.

105. The submissions then referred to the aspect of nascent market and the claim made on behalf of NSE that the market throughout remained nascent, which justified the continuance of zero transaction fee policy. We are not impressed by this argument at all, as the nascency of the market does not appear to be the only reason for indefinite continuance of zero transaction fee policy.

106. We are also not impressed by the five grounds given in support of nascency of the market. In any circumstance, in our opinion, the nascency could not continue forever, justifying the zero transaction fees policy for an indefinite period and particularly after 20.05.2009. It was at least bound to be considered afresh. The other contentions raised are mere repetitions of the arguments for nascency of the market.

107. The impugned order of the majority was criticized on the ground that a distinction made by majority in nascent, infancy, immature and mature market is devoid of logic. Again this is nothing but repetition of earlier plea about the nascency.

108. An argument was raised about the Bolton Test and the following three factors were suggested as the conditions thereof :- (a) Enhanced efficiency gains : meaning falling unit costs, and increase in the hedging activity; (b) No less restrictive means of gaining efficiency : includes the low prices to attract users. (c) Recoupment due to efficiency : refers to lower per unit cost to recoup all cost, to recoup all costs.

109. The submission also refers to the Test proposed by the European Commission in the ECs Guidelines on the Commission enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings, they being :- (a) Efficiencies brought about the conduct; (b) The conduct is indispensable to realize those efficiencies; (c) Efficiencies outweigh way any likely negative effects on competition and consumer welfare; (d) Conduct does not eliminate effective competition.

110. Speaking strictly about Bolton test, we agree that the lower prices are key to attract users. We also agree that the falling unit costs would increase the hedging activity on the CD segment. However, that is not the be-all and end-all of the matter. In fact, the last two factors in European Commissions Guidelines speak clearly about the negative effect on the competition as also elimination of effective competition. This is precisely what has happened because of the indefinite continuance of the so called policy. In our opinion, even if we accept these tests as valid tests, NSE has clearly failed in the same, in view of the reasons given above. It was very interestingly suggested in para 244 of the written submissions as under :- NSE submits that even dominant enterprises should be allowed to compete on merits and NSE was constrained to continue its zero pricing to meet the competition posed by both, BSE and MCX-SX.

111. It is very interesting to note that NSE describes itself as a dominant enterprise and then speak about its constraints for continuing the competition posed by BSE and MCX-SX. BSE was nowhere in the picture and MCX-SX was the only competitor, whom they wished to eliminate altogether by introducing zero transaction fees policy, fully knowing that MCX-SX did not have any other segment to deal with. The written submissions mention about the schedule of the Circulars right upto 26th March, 2012. Very significantly, we do not see any Circular after 30th March, 2009 even to cover the period after 20th May, 2009. All the other contentions are repetitions of the earlier arguments, which have been considered earlier.

112. A reliance placed by MCX-SX on the decision in Napp Pharmaceuticals ([2012] CAT 1) was criticized in the written submissions and it was suggested that Napp was selectively pricing below AVC to see off competitors and hence it was found guilty. It was then urged that NSE did not price below its AVC, nor did it had any intention to see off its competitors. We hasten to add that the contention is about the price being below AVC is completely incorrect for the reasons given earlier. In our opinion, MCX-SX correctly relied on the above mentioned decision. Lengthy submissions were made about the questions posed by the D.G. and the answers thereto, as also about the Equity, F&O and WDM segments of NSE, so also Gold ETF. We find all these contentions to be irrelevant. It was further suggested that the argument that NSE was aware of MCX-SX potential entry and hence in the very first decision it decided to waive transaction fees was incorrect and flawed. This argument is in fact termed as being too remote. We do not think so. The action on the part of NSE in continuing with the zero transaction fees policy, according to us is a classic example of exclusionary conduct. Same argument are repeated in respect of data feed waivers and the same plea of nascent market etc. are repeated in the written submissions. We have already dealt with data feed waivers. We have no doubt that, it was also a part of the strategy on the part of NSE. Some facts about the DotEx were also pleaded. We have already pointed out that, this issue was no more open. We would therefore not go into that issue.

113. Similar arguments appear to have been addressed in the written submissions in respect of admission fees and deposit fees. Admittedly, the same arguments in respect of transaction fees waiver are also repeated in respect of both these aspects. We endorse the finding of the majority that these waivers too were anti-competitive in nature. We are therefore, fully convinced that the conduct on the part of NSE in waiving various fees, such as transaction fees, data feed fees, admission fees etc. was absolutely anti-competitive. We also confirm the finding of the majority that this caused the breach of section 4(2) of the Act and was a classic example of abuse of dominating position by NSE. This takes us to the issue of leveraging covered under section 4(2)(e).

114. We are fully convinced that majority order was incorrect in holding the NSE guilty of the breach of section 4(2)(e). We have gone through the order and the acrobatics therein for seeing two markets. We wonder as to how a second market could be found out. We will discuss this in the further paragraphs. However, Shri Haksar very fairly submits that if we hold that the market, as we have held, being the security exchange services, there would be no question of section 4(2)(e) being breached. As it is, he did not very seriously press this issue, nor did he support the finding of the majority on the question of the breach of section 4(2)(e), i.e. leveraging for the same.

115. The language of section 4(2)(e) of the Act itself suggest that there have to be two markets, one in which the enterprise has a dominant position and the other in which it intends to enter or protect. However, both the markets must be relevant markets distinct from each other. In the wake of our finding, as also the finding by the D.G. that the relevant market in this case is the services by the stock exchange, there is no question of two markets and on that short ground itself the allegation about being guilty of the breach of section 4(2)(e) of the Act must fail. In our opinion, that will be the correct position in law.

116. However, we must at the same time take stock of the finding of CCI, since we do not agree with their finding of the relevant market being the market for services of stock exchange only in the CD segment. The CCI in its discussion on the subject first discussed the language of the section and goes on to consider the concept of associational link. The CCI also seems to have taken into consideration the language of Explanation to section 4(2) of the Act. It has also taken into consideration the fact that both the markets envisaged in section 4(2)(e) of the Act have to be relevant market. We must add that what is required to be guilty of the breach of section 4(2)(e) is under :- (1) That the enterprise has a dominant position in one market; (2) That the enterprise is dealing in not only the market in which it is dominant, but some other market also; and (3) It wants to enter into an entirely new market or protect the same.

117. In paragraph 10.83 of the impugned order an incorrect proposition of law appears as under:- While its conduct in the second market has to be separately examined for abuse if and after it acquires a dominant position there, the fact that it has used the strengths from the first market to wrongfully enter into or to protect the second market is independently considered harmful to competition under the Act.

118. We must point out that it is not necessary for the breach of section 4(2)(e) of the Act to be dominant in the second relevant market. It is enough, even if the enterprise wishes to use its strength in the market in which it is dominant to enter into or to protect the other (second market). In paragraph 10.84 of the impugned order, the CCI considered the differences in the language of sections 4(2)(a) to (d) of the Act on one hand and section 4(2)(e) of the Act on the other. Then a very interesting differentiation is made in the two markets. The first being the CD segment as X market, and the second being the non-CD segment as the Y market. The CCI then referred to the so called complexity on the basis of the fact that NSE has been considered dominant in the CD segment due to its strengths in all the non-CD segments. The CCI asked itself a question, as to how, once determined as dominant in the CD segment, could the charge of leveraging the position in the that market to enter or to protect the same CD segment itself be made. Then it goes on to say that this question assumes that once the CD segment has been taken as the relevant market, then wherever the word relevant market occurs in clauses 4(2)(a) to (e) of the Act, it should automatically refer to CD segment. In our opinion, this discussion is completely illogical.

119. In paragraph 10.85 or the impugned order, the CCI goes on to explain that the relevant market for clause 4(2)(e) of the Act can be different from the relevant market for clauses 4(2)(a) to (d) of the Act, but the aspects of dominance given in explanation (a) would apply equally to both. We also do not approve of this statement of law. We also do not approve of the further inference in fact the scheme of the section, particularly when read with section 19(4) is such that it is possible to take one market as the relevant market for sub-sections (a) to (d) of section 4(2) and the same market as the other relevant market for section 4(2)(e). Reliance is then placed on section 19(4) in paragraph 10.88 of the impugned order, which is as under :- In the Indian Competition Act, under section 19(4), the ability to leverage, in itself, is taken as one of the factors of dominance. This revalidates our observation above that both "position of strength" as well as the concept of leveraging has slightly different nuances in the Indian Act. Phrases like "size and importance of competitors", "vertical integration", "relative advantage" etc. are concepts that indicate the strength to leverage based on strengths in other markets. It is this strength that would render an enterprise dominant in the relevant market itself and would expose its conduct therein to evaluation of any other abuse of dominance separately. At the same time, the wrongful exercise of that strength by itself is also held as abusive conduct in its own right, under section4(2)(e).

120. The CCI has completely misdirected itself on the basis of incorrect and faulty logic. In paragraph 10.89 of the impugned order, the CCI proceeded on the ground that in both X as well Y market the enterprise was dominant and held that same was the situation in this case, meaning NSE was dominant in both CD market as well as in the so called non-CD segment. After thus discussing the law, the CCI posed four questions, which were:- (a) Whether NSE held a position of strength in the CD segment market comparable to its position in the CD and non CD segment markets as a whole? (b) Whether NSE enjoyed advantages in the CD segment market by virtue of its dominance in the non CD segment market? (c) Whether NSE customers in one market were potential customers in the other? (d) Whether NSE and its competitors could become competitors in both markets?

121. It then proceeded to hold that the two relevant markets had associational links and therefore, NSE used its position of strength in non- CD segment to protect its position in the CD segment. The CCI then answered all these questions in positive, on the basis of non-existent fact that MCX-SX was likely to be a competitor of NSE, since it had applied for the other segments. We do not think that the order of CCI was for valid reasons and was based on the correct interpretation of section 4(2)(e) of the Act. In our opinion, if the relevant market as held by the CCI was only the CD segment, then there could not be any other market like non-CD segment. There was no necessity of putting all the other segments in one group and indeed it could not have been done, much less to hold it as another relevant market. The logic of the CCI on this question is flawed and we reject the same. We, therefore, hold that the NSE could not have been guilty of the breach of section 4(2)(e) of the Act, basically on the logic that there was only one market and that market was the services of the stock exchange. Merely because at the relevant time period, the services of the stock exchange of MCX-SX were limited to CD segment, it does not mean that the relevant market had to be held as a CD segment market. We set aside the finding of the CCI on this issue. This brings us to the question of penalty.

122. Detailed submissions were addressed to us on the question of penalty. The CCI while considering the penalty considered the following contentions raised by NSE in respect of the imposition of penalty. They are:- (1) Novelty : Meaning thereby that the alleged violations being based on novel concepts and principles, they were incapable of having been anticipated for the purpose of compliance; (2) Uncertainty on application of law : It was suggested that since in the absence of guidance papers or a case law from the Commission, there was large element of uncertainty in the law; (3) Lack of cogent or convincing evidence; (4) Lack of intention or negligence : In this it was suggested that there was no ill-intention, nor was there any negligence on the part of NSE, in commencing the zero transaction policy; (5) No foreclosure : In the sense that since there was no foreclosure in the CD segment, there could not be an abuse of dominance. In this, it was pointed out to the CCI the losses incurred by MCX-SX as a result of zero pricing policy of the NSE were small in comparison with MCX-SX excess capital and MCX- SX not being harmed; (6) Benefit to ultimate consumers; (7) Expansion of the market; (8) Contribution to economic development; (9) Meeting the competition; (10) Full support and cooperation; (11) Principle of Proportionality; (12) Order contrary to foreign precedents; (13) Miscellaneous arguments made before them on the basis of case laws of European Commission.

123. It was also argued that the OECD document mentioned that lighter measures should be used when such a subject was never dealt with by the Courts in past. Even section 53N of the Act was referred to press the contention that no penalty should be imposed. Arguments were also made before CCI on the question of behavioural remedy. It was also urged that the cost estimates provided by the D.G. were incorrect. Further, the concept of turnover was also attacked and it was suggested that since turnover was zero in the CD segment, there was no need of penalty. The CCI seems to have discussed all these aspects in details in paragraph 18 of its order and answered each such aspect, particularly on the question of novelty and uncertainty on application of law. We are generally satisfied with that discussion. The CCI has also commented on the other aspects like lack of cogent or convincing evidence and lack of intention or negligence. A thorough discussion was done by the CCI in its order and practically on all the aspects it has given its comments. We are completely satisfied with the views expressed in this behalf while considering the penalty under section 27 of the Act. Lastly it was mentioned that even the mitigating factor wherever justifiable were not taken into consideration. It has been held by the CCI that besides the abuse of dominant position in terms of section 4(2)(a)(ii), it has cross subsidized from other segments of business; that it also camouflaged its intentions by not maintaining separate accounts for the CD segment; that NSE created a faade of the nascency of market for not charging any fees on account of transactions in the CD segment; it expressed that the small pockets were bound to be thrown out of the market, if they had also followed the zero transaction fee policy, which was adopted by NSE by incurring huge losses. The CCI also took into consideration the past conduct of NSE and its conduct in the CD segment suggesting longing on its part for dominance in any segments and therefore, it proceeded to pass on the order prohibiting the NSE from the practice of unfair pricing, exclusionary conduct and unfairly using its dominant position in other markets to protect the CD segment market. The CCI also directed NSE to maintain separate accounts and further directed it to modify its zero price policy in the CD segment and ensure that appropriate transaction costs were levied. It then proceeded to hold that in view of this behavior on the part of NSE the penalty at the rate of 5% at the average turnover was liable to be levied. Accordingly, it took into account the turnover being Rs.3328.98 crores, the average turnover being Rs.1109.66 crores. Accordingly, the CCI ordered the penalty @ 5%. In our considered opinion, the CCI has in fact practiced restraint on itself in ordering only 5% of the average turnover, when it could have gone right upto 10% of the average turnover. Considering the total turnover, the average turnover, in our opinion the CCI was quite justified in ordering the penalty at the rate of 5%. It was also repeated before us that this was a new law and like European Union in two cases, no penalty should have been levied. It was also argued that in past the CCI had imposed either no penalties or symbolic penalties. The examples of Distribution of package tours during the 1990 World Cup, [(1992) OJ L 326/31] as also Clear Stream (Clearing and Settlement), 2009 [OJ C 165/7] case were given where no impositions of penalty was ordered. The CCI has thoroughly discussed all these arguments, which were placed before it. We are quite satisfied with the appreciation of the material put before the CCI as also before us. The uncertainty in the application of law was also pressed into service like it was done before the CCI. We agree with the CCI in its observations about these aspects. It was tried to be said that there were benefits of zero pricing and that there was no denial of market access. We agree with the CCI for the reasons given for rejecting these contentions and for the same reasons, we also reject these contentions. Much was made about penalty levied @ 5% of the average turnover. Our judgment in the case of M/s. Excel Crop vs. CCI was pressed into service to suggest that the relevant turnover should alone be considered for the sake of penalty. All these arguments of relevant turnover should fall to the ground in the wake of our finding that the relevant market in this case IS the services of stock exchange in all the segments. In M/s. Excel Crops judgment, there were well defined distinct markets and it was a multi- commodity company. That is not a case here. Then, it was pointed out that the relevant turnover on account of the zero transaction fees policy was also zero. So, we were therefore asked to adopt a notional turnover figure. We have already pointed out that the NSE was making tons of profits from the relevant market on account of its services in the other segments. Therefore, there can be no justification for taking any lenient view, nor is it necessary to consider the concept of notional turnover figure, when the turnover of the NSE is well available on the basis of Annual Reports. We, therefore, reject the contention that the turnover for the CD segment should be the relevant turnover. In our opinion, we have given enough reasons to hold that the whole turnover of NSE should be taken into consideration, in view of our finding on the relevant market. We, therefore, do not propose to modify the order of CCI in this behalf. As regards, the other remedies imposed, the NSE has submitted that it should not be required to maintain segment wise account as in a multi-product firm, it is difficult to apportion shared based fixed costs and further that AS

17 does not requires the maintaining of segment wise accounts. In view of our finding that the relevant market is the services of Stock Exchange, in all the segments it will not be necessary to maintain segment wise accounts. We, therefore, do not approve of this direction by the CCI and it is ordered to be deleted. Before closing, some submissions were filed on behalf of the NSE on 30th July, 2014. In fact, these submissions have been filed much after the time granted for the same and therefore, they should not have been accepted by the Registry. We refuse to consider them on this ground alone. We do not see any merits in the Appeal and dismiss the same. Accordingly, the Appeal is dismissed. Pronounced in open Court on 5th day of August, 2014. ( Justice V.S. Sirpurkar) Chairman (Rahul Sarin) Member