This is Part 1/3 of a series of blogs on the history of the Market Data Wars. Feedback and suggestions more than welcome at benjamin.connault@iextrading.com
December 17, 2019 marked the 20th anniversary of the Market Data Wars.December 17, 2019 marked the 20th anniversary of the Market Data Wars.
The “Market Data Wars” refers to a long series of commercial, regulatory and legal developments around U.S. equities market data products and their prices. The outcome of the Market Data Wars will determine the extent to which market data products and prices can be regulated by the SEC. On one side are the incumbent U.S. exchange groups, which want as much commercial discretion as possible in the type and prices of market data products that they can offer. On the other side is a coalition of market data consumers together with, to an extent that has varied over time, the SEC (the exchanges’ regulator). Members of this coalition think market data products have monopoly good features and/or public good features that justify regulatory controls.The “Market Data Wars” refers to a long series of commercial, regulatory and legal developments around U.S. equities market data products and their prices. The outcome of the Market Data Wars will determine the extent to which market data products and prices can be regulated by the SEC. On one side are the incumbent U.S. exchange groups, which want as much commercial discretion as possible in the type and prices of market data products that they can offer. On the other side is a coalition of market data consumers together with, to an extent that has varied over time, the SEC (the exchanges’ regulator). Members of this coalition think market data products have monopoly good features and/or public good features that justify regulatory controls.
The Market Data Wars have been raging continuously for 20 years now and show no sign of abating. If anything, a June 5, 2020, decision by the U.S. Court of Appeals for the D.C. Circuit (the “D.C. Circuit”) will likely intensify the conflict. The decision was made in an important case revolving around a market data fee originally filed in 2006. (The decision focused on procedural issues rather than market data economics — Part 2 of this blog series will include more details.)The Market Data Wars have been raging continuously for 20 years now and show no sign of abating. If anything, a June 5, 2020, decision by the U.S. Court of Appeals for the D.C. Circuit (the “D.C. Circuit”) will likely intensify the conflict. The decision was made in an important case revolving around a market data fee originally filed in 2006. (The decision focused on procedural issues rather than market data economics — Part 2 of this blog series will include more details.)
With billions of dollars at stake, considerable resources have been spent on both sides of the arguments. In these blogs, I selected ten expert reports that fueled the Market Data Wars. I based my selection on historical relevance and conceptual novelty. These ten reports take us through 20 years of Market Data Wars history, from December 1999 to January 2020. No blood was spilled in these wars, but make no mistake about their intensity or importance: they are shaping capital markets for decades to come.With billions of dollars at stake, considerable resources have been spent on both sides of the arguments. In these blogs, I selected ten expert reports that fueled the Market Data Wars. I based my selection on historical relevance and conceptual novelty. These ten reports take us through 20 years of Market Data Wars history, from December 1999 to January 2020. No blood was spilled in these wars, but make no mistake about their intensity or importance: they are shaping capital markets for decades to come.
This blog series has three parts. Today’s Part 1 takes us from December 1999 to January 2011, as the Market Data Wars were launched and foundational theories were established. Part 2 covers January 2011 to March 2015 and describes, among other things, the background for the recent D.C. Circuit Decision. Part 3 covers March 2015 to January 2020 and focuses on the rise of “platform theory” arguments in the Market Data Wars. Together, the three parts include ten sections revolving around ten expert reports. Details and links to each expert report are provided at the beginning of each section.This blog series has three parts. Today’s Part 1 takes us from December 1999 to January 2011, as the Market Data Wars were launched and foundational theories were established. Part 2 covers January 2011 to March 2015 and describes, among other things, the background for the recent D.C. Circuit Decision. Part 3 covers March 2015 to January 2020 and focuses on the rise of “platform theory” arguments in the Market Data Wars. Together, the three parts include ten sections revolving around ten expert reports. Details and links to each expert report are provided at the beginning of each section.
1) December 1999: SEC Concept Release launches the Market Data Wars
Grawe, McConihe and Kolnik (with PHB Hagler Bailly) for NYSE, December 1999. Available here.Grawe, McConihe and Kolnik (with PHB Hagler Bailly) for NYSE, December 1999. Available here.
Published a few days before the turn of the 21st Century, the SEC’s “Regulation of Market Information Fees and Revenues” Concept Release proposed a new cost-based approach to regulating market data prices. This Concept Release marks the beginning of the Market Data Wars. Concerns about the possibility of for-profit exchanges and the cost of market data for retail investors prompted the proposal. Generally, the Concept Release correctly anticipated many trends of the coming two decades, including demutualization, computerization, and the growing importance of market data. But it failed to identify some crucial ones, most notably the rise of high-frequency trading.Published a few days before the turn of the 21st Century, the SEC’s “Regulation of Market Information Fees and Revenues” Concept Release proposed a new cost-based approach to regulating market data prices. This Concept Release marks the beginning of the Market Data Wars. Concerns about the possibility of for-profit exchanges and the cost of market data for retail investors prompted the proposal. Generally, the Concept Release correctly anticipated many trends of the coming two decades, including demutualization, computerization, and the growing importance of market data. But it failed to identify some crucial ones, most notably the rise of high-frequency trading.
NYSE, wanting to protect its ability to freely brand, market, and price its market data products, wrote a comment letter to that effect. In support of its arguments, it commissioned two expert reports from the consulting firm PHB Hagler Bailly. Among other things, the reports argued that cost-based price regulation is flawed, and that market data prices cannot be regulated separately from transaction fees or listing fees because they are “joint products”.NYSE, wanting to protect its ability to freely brand, market, and price its market data products, wrote a comment letter to that effect. In support of its arguments, it commissioned two expert reports from the consulting firm PHB Hagler Bailly. Among other things, the reports argued that cost-based price regulation is flawed, and that market data prices cannot be regulated separately from transaction fees or listing fees because they are “joint products”.
Shortly after the Concept Release, the SEC created an Advisory Committee On Market Information. In its influential final report, published in September 2001, the Committee recommended against the SEC’s cost-based proposal and in favor of a “competing consolidators” approach to the public tape. (The SEC adopted neither at the time, but the “competing consolidators” model is the cornerstone of the SEC’s February 2020 “Market Data Infrastructure” Proposed Rule.)Shortly after the Concept Release, the SEC created an Advisory Committee On Market Information. In its influential final report, published in September 2001, the Committee recommended against the SEC’s cost-based proposal and in favor of a “competing consolidators” approach to the public tape. (The SEC adopted neither at the time, but the “competing consolidators” model is the cornerstone of the SEC’s February 2020 “Market Data Infrastructure” Proposed Rule.)
2) 2006–2008: NYSE files ArcaBook I, NetCoalition petitions for review, and a “market-based” standard for fee filings is spelled out for the first time
Evans (with LECG Global Competition Policy Practice) for NetCoalition, July 2008. Available here.Evans (with LECG Global Competition Policy Practice) for NetCoalition, July 2008. Available here.
On June 7, 2006, NYSE Arca filed proposed rule change SR-NYSEArca-2006–21 to establish fees on ArcaBook, a depth-of-book market data product which had been free until then. Initially approved, the ArcaBook filing was suspended following a petition for review by NetCoalition, an Internet industry group including Bloomberg, Google, Yahoo, and other market data consumers. This became the first in a long a series of court cases around ArcaBook, some of which are still being litigated in 2020.On June 7, 2006, NYSE Arca filed proposed rule change SR-NYSEArca-2006–21 to establish fees on ArcaBook, a depth-of-book market data product which had been free until then. Initially approved, the ArcaBook filing was suspended following a petition for review by NetCoalition, an Internet industry group including Bloomberg, Google, Yahoo, and other market data consumers. This became the first in a long a series of court cases around ArcaBook, some of which are still being litigated in 2020.
On June 4, 2008, the SEC published and opened for comments a proposed order regarding the ArcaBook filing. The proposed order did two things.On June 4, 2008, the SEC published and opened for comments a proposed order regarding the ArcaBook filing. The proposed order did two things.
First, it described a new market-based framework for analyzing exchange market data fee filings. An exchange could argue that a proposed market data fee was “equitable, fair, reasonable and not unreasonably discriminatory” by showing that the fee was constrained by competitive forces. Two specific forces could constrain a fee: the availability of alternative data products, and competition for order flow to the exchange.First, it described a new market-based framework for analyzing exchange market data fee filings. An exchange could argue that a proposed market data fee was “equitable, fair, reasonable and not unreasonably discriminatory” by showing that the fee was constrained by competitive forces. Two specific forces could constrain a fee: the availability of alternative data products, and competition for order flow to the exchange.
Second, the proposed order applied the market-based framework to the ArcaBook filing and found that competitive forces did constrain the market data fee. Based on that analysis, the order proposed to approve the filing.Second, the proposed order applied the market-based framework to the ArcaBook filing and found that competitive forces did constrain the market data fee. Based on that analysis, the order proposed to approve the filing.
On July 10, 2008, NetCoalition submitted the Evans report as a comment to the SEC’s proposed order. In it, Evans assumed that the proposed order’s market-based framework was correct but, in applying it to the ArcaBook filing, found results opposite to those of the SEC. According to Evans, “the alternative sources of depth-of-book data identified by the SEC are likely not substitutes for an exchange’s depth-of-book data” and “the relationship between [order flow competition and depth-of-book data] is neither strong nor direct.”On July 10, 2008, NetCoalition submitted the Evans report as a comment to the SEC’s proposed order. In it, Evans assumed that the proposed order’s market-based framework was correct but, in applying it to the ArcaBook filing, found results opposite to those of the SEC. According to Evans, “the alternative sources of depth-of-book data identified by the SEC are likely not substitutes for an exchange’s depth-of-book data” and “the relationship between [order flow competition and depth-of-book data] is neither strong nor direct.”
3) August 2008: Nasdaq argues a “joint products with joint costs” theory
Ordover and Bamberger (with Compass Lexecon) for Nasdaq, August 2008. Available here.Ordover and Bamberger (with Compass Lexecon) for Nasdaq, August 2008. Available here.
Like the Evans report, the Ordover-Bamberger report was submitted as a comment to the SEC’s ArcaBook proposed order, in this case by Nasdaq. In contrast to the Evans report, the Ordover-Bamberger report supported both the SEC’s market-based framework and its application to the ArcaBook filing. The Ordover-Bamberger report argued further that the SEC’s findings apply “to other potential sellers of non-core market data, including NASDAQ, and to other products in addition to the specific NYSE Arca data products analyzed by the Commission.”Like the Evans report, the Ordover-Bamberger report was submitted as a comment to the SEC’s ArcaBook proposed order, in this case by Nasdaq. In contrast to the Evans report, the Ordover-Bamberger report supported both the SEC’s market-based framework and its application to the ArcaBook filing. The Ordover-Bamberger report argued further that the SEC’s findings apply “to other potential sellers of non-core market data, including NASDAQ, and to other products in addition to the specific NYSE Arca data products analyzed by the Commission.”
The Ordover-Bamberger report is also historically significant for two (not necessarily intentional) legacies of dubious value.The Ordover-Bamberger report is also historically significant for two (not necessarily intentional) legacies of dubious value.
First, it introduced the “platform” term to describe the economic activity of exchanges. This usage stuck and has created a lot of confusion, especially after the Supreme Court’s Ohio v. Amex decision propelled “multi-sided platforms” — a different concept — in the spotlight (see Part 3 of this blog).First, it introduced the “platform” term to describe the economic activity of exchanges. This usage stuck and has created a lot of confusion, especially after the Supreme Court’s Ohio v. Amex decision propelled “multi-sided platforms” — a different concept — in the spotlight (see Part 3 of this blog).
Second, it coined the expression “joint products with joint costs” to describe the economic activity of exchanges, where the joint products are market data and trading services. That expression became the boilerplate expression that many exchanges have used to justify market data fee increases in countless fee filings since.Second, it coined the expression “joint products with joint costs” to describe the economic activity of exchanges, where the joint products are market data and trading services. That expression became the boilerplate expression that many exchanges have used to justify market data fee increases in countless fee filings since.
The SEC eventually approved the ArcaBook filing on December 2, 2008. The approval was appealed by SIFMA and NetCoalition in front of the D.C. Circuit on January 30, 2009. What happened after will be covered in Part 2.The SEC eventually approved the ArcaBook filing on December 2, 2008. The approval was appealed by SIFMA and NetCoalition in front of the D.C. Circuit on January 30, 2009. What happened after will be covered in Part 2.
4) 2011: Attempted Linking of Market Data Fees to Trade Executions
Ordover and Bamberger (with Compass Lexecon) for Nasdaq, January 2011. Available here.Ordover and Bamberger (with Compass Lexecon) for Nasdaq, January 2011. Available here.
On January 20, 2011, Nasdaq filed SR-NASDAQ-2011–010, a rule change that proposed offering discounts on market data products for Nasdaq members which, among other things, would trade more than a given volume threshold on behalf of retail investors. This proposal broke new ground in linking market data fees to trade executions — “platform pricing” in Nasdaq terminology.On January 20, 2011, Nasdaq filed SR-NASDAQ-2011–010, a rule change that proposed offering discounts on market data products for Nasdaq members which, among other things, would trade more than a given volume threshold on behalf of retail investors. This proposal broke new ground in linking market data fees to trade executions — “platform pricing” in Nasdaq terminology.
Likely anticipating that the platform pricing proposal would be controversial, Nasdaq commissioned another report from Ordover and Bamberger and attached it as an exhibit to its fee filing. The report reiterated that availability of alternative data products and competition for order flow constrain the prices of Nasdaq’s market data products. It adds that price differentiation, volume discounts and bundling — the practice of pricing structures that involve conditions across several products — are common, efficient and not anticompetitive. The report generally argued that “each platform should be free to determine how best to recover the costs […] of its joint products.”Likely anticipating that the platform pricing proposal would be controversial, Nasdaq commissioned another report from Ordover and Bamberger and attached it as an exhibit to its fee filing. The report reiterated that availability of alternative data products and competition for order flow constrain the prices of Nasdaq’s market data products. It adds that price differentiation, volume discounts and bundling — the practice of pricing structures that involve conditions across several products — are common, efficient and not anticompetitive. The report generally argued that “each platform should be free to determine how best to recover the costs […] of its joint products.”
The report also includes an interesting discussion of “fairness” in the context of market data products. The Securities Exchange Act imposes a “fairness” statutory standard to exchange fee filings. The report acknowledges that “‘fairness’ is not a core concept of microeconomics or of industrial organization” and that “[a]fter all is said and done, the metric of what is fair or unfair has to be imported from elsewhere from outside of the model.”The report also includes an interesting discussion of “fairness” in the context of market data products. The Securities Exchange Act imposes a “fairness” statutory standard to exchange fee filings. The report acknowledges that “‘fairness’ is not a core concept of microeconomics or of industrial organization” and that “[a]fter all is said and done, the metric of what is fair or unfair has to be imported from elsewhere from outside of the model.”
The SEC eventually disapproved the Nasdaq filing, noting, “In fact, the Commission believes that preventing the linking of market data fees to trade executions will help bolster competitive forces in the area of market data, because exchange market data fees must appeal simultaneously to market participants that trade directly on an exchange and those that do not trade directly on an exchange.”The SEC eventually disapproved the Nasdaq filing, noting, “In fact, the Commission believes that preventing the linking of market data fees to trade executions will help bolster competitive forces in the area of market data, because exchange market data fees must appeal simultaneously to market participants that trade directly on an exchange and those that do not trade directly on an exchange.”