Daily Rules, Proposed Rules, and Notices of the Federal Government
NASDAQ is proposing to extend for three months the fee pilot pursuant to which NASDAQ distributes the NASDAQ Last Sale ("NLS") market data products. NLS allows data distributors to have access to real-time market data for a capped fee, enabling those distributors to provide free access to the data to millions of individual investors via the internet and television. Specifically, NASDAQ offers the "NASDAQ Last Sale for NASDAQ" and "NASDAQ Last Sale for NYSE/Amex" data feeds containing last sale activity in U.S. equities within the NASDAQ Market Center and reported to the FINRA/NASDAQ Trade Reporting Facility ("FINRA/NASDAQ TRF"), which is jointly operated by NASDAQ and the Financial Industry Regulatory Authority ("FINRA"). The purpose of this proposal is to extend the existing pilot program for three months, from October 1, 2012 to December 31, 2012.
This pilot program supports the aspiration of Regulation NMS to increase the availability of proprietary data by allowing market forces to determine the amount of proprietary market data information that is made available to the public and at what price. During the pilot period, the program has vastly increased the availability of NASDAQ proprietary market data to individual investors. Based upon data from NLS distributors, NASDAQ believes that since its launch in July 2008, the NLS data has been viewed by over 50,000,000 investors on Web sites operated by Google, Interactive Data, and Dow Jones, among others.
The text of the proposed rule change is below. Proposed new language is italicized; proposed deletions are in brackets.
In its filing with the Commission, the self-regulatory organization included statements concerning the purpose of, and basis for, the proposed rule change and discussed any comments it received on the proposed rule change. The text of those statements may be examined at the places specified in Item IV below. The Exchange has prepared summaries, set forth in sections A, B, and C below, of the most significant parts of such statements.
Prior to the launch of NLS, public investors that wished to view market data to monitor their portfolios generally had two choices: (1) Pay for real-time market data or (2) use free data that is 15 to 20 minutes delayed. To increase consumer choice, NASDAQ proposed a pilot to offer access to real-time market data to data distributors for a capped fee, enabling those distributors to disseminate the data at no cost to millions of internet users and television viewers. NASDAQ now proposes a three-month extension of that pilot program, subject to the same fee structure as is applicable today.
NLS consists of two separate "Level 1" products containing last sale activity within the NASDAQ market and reported to the jointly-operated FINRA/NASDAQ TRF. First, the "NASDAQ Last Sale for NASDAQ" data product is a real-time data feed that provides real-time last sale information including execution price, volume, and time for executions occurring within the NASDAQ system as well as those reported to the FINRA/NASDAQ TRF. Second, the "NASDAQ Last Sale for NYSE/Amex" data product provides real-time last sale information including execution price, volume, and time for NYSE- and NYSE Amex-securities executions occurring within the NASDAQ system as well as those reported to the FINRA/NASDAQ TRF. By contrast, the securities information processors ("SIPs") that provide "core" data consolidate last sale information from all exchanges and trade reporting facilities ("TRFs"). Thus, NLS replicates a subset of the information provided by the SIPs.
NASDAQ established two different pricing models, one for clients that are able to maintain username/password
NASDAQ also established a cap on the monthly fee, currently set at $50,000 per month for all NASDAQ Last Sale products. The fee cap enables NASDAQ to compete effectively against other exchanges that also offer last sale data for purchase or at no charge.
As with the distribution of other NASDAQ proprietary products, all distributors of the NASDAQ Last Sale for NASDAQ and/or NASDAQ Last Sale for NYSE/Amex products pay a single $1,500/month NASDAQ Last Sale Distributor Fee in addition to any applicable usage fees. The $1,500 monthly fee applies to all distributors and does not vary based on whether the distributor distributes the data internally or externally or distributes the data via both the internet and television.
NASDAQ believes that the proposed rule change is consistent with the provisions of Section 6 of the Act,
NASDAQ believes that its NASDAQ Last Sale market data products are precisely the sort of market data product that the Commission envisioned when it adopted Regulation NMS. The Commission concluded that Regulation NMS--by lessening regulation of the market in proprietary data--would itself further the Act's goals of facilitating efficiency and competition:
The recent decision of the United States Court of Appeals for the District of Columbia Circuit in
The Court in
NASDAQ does not believe that the proposed rule change will result in any burden on competition that is not necessary or appropriate in furtherance of the purposes of the Act, as amended. NASDAQ's ability to price its Last Sale Data Products is constrained by (1) Competition between exchanges and other trading platforms that compete with each other in a variety of dimensions; (2) the existence of inexpensive real-time consolidated data and market-specific data and free delayed consolidated data; and (3) the inherent contestability of the market for proprietary last sale data.
The market for proprietary last sale data products is currently competitive
Transaction execution and proprietary data products are complementary in that market data is both an input and a byproduct of the execution service. In fact, market data and trade execution are a paradigmatic example of joint products with joint costs. The decision whether and on which platform to post an order will depend on the attributes of the platform where the order can be posted, including the execution fees, data quality and price, and distribution of its data products. Without trade executions, exchange data products cannot exist. Moreover, data products are valuable to many end users only insofar as they provide information that end users expect will assist them or their customers in making trading decisions.
The costs of producing market data include not only the costs of the data distribution infrastructure, but also the costs of designing, maintaining, and operating the exchange's transaction execution platform and the cost of regulating the exchange to ensure its fair operation and maintain investor confidence. The total return that a trading platform earns reflects the revenues it receives from both products and the joint costs it incurs. Moreover, the operation of the exchange is characterized by high fixed costs and low marginal costs. This cost structure is common in content and content distribution industries such as software, where developing new software typically requires a large initial investment (and continuing large investments to upgrade the software), but once the software is developed, the incremental cost of providing that software to an additional user is typically small, or even zero (
An exchange's BD customers view the costs of transaction executions and of data as a unified cost of doing business with the exchange. A BD will direct orders to a particular exchange only if the expected revenues from executing trades on the exchange exceed net transaction execution costs and the cost of data that the BD chooses to buy to support its trading decisions (or those of its customers). The choice of data products is, in turn, a product of the value of the products in making profitable trading decisions. If the cost of the product exceeds its expected value, the BD will choose not to buy it. Moreover, as a BD chooses to direct fewer orders to a particular exchange, the value of the product to that BD decreases, for two reasons. First, the product will contain less information, because executions of the BD's trading activity will not be reflected in it. Second, and perhaps more important, the product will be less valuable to that BD because it does not provide information about the venue to which it is directing its orders. Data from the competing venue to which the BD is directing orders will become correspondingly more valuable.
Similarly, in the case of products such as NLS that are distributed through market data vendors, the vendors provide price discipline for proprietary data products because they control the primary means of access to end users. Vendors impose price restraints based upon their business models. For example, vendors such as Bloomberg and Reuters that assess a surcharge on data they sell may refuse to offer proprietary products that end users will not purchase in sufficient numbers. Internet portals, such as Google, impose a discipline by providing only data that will enable them to attract "eyeballs" that contribute to their advertising revenue. Retail BDs, such as Schwab and Fidelity, offer their customers proprietary data only if it promotes trading and generates sufficient commission revenue. Although the business models may differ, these vendors' pricing discipline is the same: they can simply refuse to purchase any proprietary data product that fails to provide sufficient value. NASDAQ and other producers of proprietary data products must understand and respond to these varying business models and pricing disciplines in order to market proprietary data products successfully. Moreover, NASDAQ believes that products such as NLS can enhance order flow to NASDAQ by providing more widespread distribution of information about transactions in real time, thereby encouraging wider participation in the market by investors with access to the internet or television. Conversely, the value of such products to distributors and investors decreases if order flow falls, because the products contain less content.
Analyzing the cost of market data distribution in isolation from the cost of all of the inputs supporting the creation of market data will inevitably underestimate the cost of the data. Thus, because it is impossible to create data without a fast, technologically robust, and well-regulated execution system, system costs and regulatory costs affect the price of market data. It would be equally misleading, however, to attribute all of the exchange's costs to the market data portion of an exchange's joint product. Rather, all of the exchange's costs are incurred for the unified purposes of attracting order flow, executing and/or routing orders, and generating and selling data about market activity. The total return that an exchange earns reflects the revenues it receives from the joint products and the total costs of the joint products.
Competition among trading platforms can be expected to constrain the aggregate return each platform earns from the sale of its joint products, but different platforms may choose from a range of possible, and equally reasonable, pricing strategies as the means of recovering total costs. NASDAQ pays rebates to attract orders, charges relatively low prices for market information and charges relatively high prices for accessing posted liquidity. Other platforms may choose a strategy of paying lower liquidity rebates to attract orders, setting relatively low prices for accessing posted liquidity, and setting relatively high prices for market information. Still others may provide most data free of charge and rely exclusively on transaction fees to recover their costs. Finally, some platforms may incentivize use by providing opportunities for equity ownership, which may allow them to charge lower direct fees for executions and data.
In this environment, there is no economic basis for regulating maximum prices for one of the joint products in an industry in which suppliers face competitive constraints with regard to the joint offering. Such regulation is unnecessary because an "excessive" price for one of the joint products will ultimately have to be reflected in lower prices for other products sold by the firm, or otherwise the firm will experience a loss in the volume of its sales that will be adverse to its overall profitability. In other words, an increase in the price of data will ultimately have to be accompanied by a decrease in the cost of executions, or the volume of both data and executions will fall.
The level of competition and contestability in the market is evident in the numerous alternative venues that compete for order flow, including thirteen SRO markets, as well as internalizing BDs and various forms of alternative trading systems ("ATSs"), including dark pools and electronic communication networks ("ECNs"). Each SRO market competes to produce transaction reports via trade executions, and two FINRA-regulated TRFs compete to attract internalized transaction reports. It is common for BDs to further and exploit this competition by sending their order flow and transaction reports to multiple markets, rather than providing them all to a single market. Competitive markets for order flow, executions, and transaction reports provide pricing discipline for the inputs of proprietary data products.
The large number of SROs, TRFs, BDs, and ATSs that currently produce proprietary data or are currently capable of producing it provides further pricing discipline for proprietary data products. Each SRO, TRF, ATS, and BD is currently permitted to produce proprietary data products, and many currently do or have announced plans to do so, including NASDAQ, NYSE, NYSEAmex, NYSEArca, BATS, and Direct Edge.
Any ATS or BD can combine with any other ATS, BD, or multiple ATSs or BDs to produce joint proprietary data products. Additionally, order routers and market data vendors can facilitate single or multiple BDs' production of proprietary data products. The potential sources of proprietary products are virtually limitless.
The fact that proprietary data from ATSs, BDs, and vendors can by-pass SROs is significant in two respects. First, non-SROs can compete directly with SROs for the production and sale of proprietary data products, as BATS and Arca did before registering as exchanges by publishing proprietary book data on the Internet. Second, because a single order or transaction report can appear in a core data product, an SRO proprietary product, and/or a non-SRO proprietary product, the data available in proprietary products is exponentially greater than the actual number of orders and transaction reports that exist in the marketplace. Indeed, in the case of NLS, the data provided through that product appears both in (i) real-time core data products offered by the SIPs for a fee, and (ii) free SIP data products with a 15-minute time delay, and finds a close substitute in last-sale products of competing venues.
In addition to the competition and price discipline described above, the market for proprietary data products is also highly contestable because market entry is rapid, inexpensive, and profitable. The history of electronic trading is replete with examples of entrants that swiftly grew into some of the largest electronic trading platforms and proprietary data producers: Archipelago, Bloomberg Tradebook, Island, RediBook, Attain, TracECN, BATS Trading and Direct Edge. Today, BATS and Direct Edge provide data at no charge in order to attract order flow, and use market data revenue rebates from the resulting executions to maintain low execution charges for their users. A proliferation of dark pools and other ATSs operate profitably with fragmentary shares of consolidated market volume.
Regulation NMS, by deregulating the market for proprietary data, has increased the contestability of that market. While BDs have previously published their proprietary data individually, Regulation NMS encourages market data vendors and BDs to produce proprietary products cooperatively in a manner never before possible. Multiple market data vendors already have the capability to aggregate data and disseminate it on a profitable scale, including Bloomberg and Thomson Reuters.
Moreover, consolidated data provides two additional measures of pricing discipline for proprietary data products that are a subset of the consolidated data stream. First, the consolidated data is widely available in real-time at $1 per month for non-professional users. Second, consolidated data is also available
The competitive nature of the market for products such as NLS is borne out by the performance of the market. In May 2008, the internet portal Yahoo! began offering its Web site viewers real-time last sale data (as well as best quote data) provided by BATS Trading. In response, in June 2008, NASDAQ launched NLS, which was initially subject to an "enterprise cap" of $100,000 for customers receiving only one of the NLS products, and $150,000 for customers receiving both products. The majority of NASDAQ's sales were at the capped level. In early 2009, BATS expanded its offering of free data to include depth-of-book data. Also in early 2009, NYSEArca announced the launch of a competitive last sale product with an enterprise price of $30,000 per month. In response, NASDAQ combined the enterprise cap for the NLS products and reduced the cap to $50,000 (
In this environment, a super-competitive increase in the fees charged for either transactions or data has the potential to impair revenues from both products. "No one disputes that competition for order flow is `fierce'."
In establishing the price for the NASDAQ Last Sale Products, NASDAQ considered the competitiveness of the market for last sale data and all of the implications of that competition. NASDAQ believes that it has considered all relevant factors and has not considered irrelevant factors in order to establish fair, reasonable, and not unreasonably discriminatory fees and an equitable allocation of fees among all users. The existence of numerous alternatives to NLS, including real-time consolidated data, free delayed consolidated data, and proprietary data from other sources ensures that NASDAQ cannot set unreasonable fees, or fees that are unreasonably discriminatory, without losing business to these alternatives. Accordingly, NASDAQ believes that the acceptance of the NLS product in the marketplace demonstrates the consistency of these fees with applicable statutory standards.
Three comment letters were filed regarding the proposed rule change as originally published for comment NASDAQ responded to these comments in a letter dated December 13, 2007. Both the comment letters and NASDAQ's response are available on the SEC Web site at
While containing a few superficial modifications from prior letters, SIFMA and NetCoalition's most recently submitted letter continues to mischaracterize the import of the original
SIFMA and NetCoalition further contend the prior filing lacked evidence supporting a conclusion that the market for NLS is competitive, asserting that arguments about competition for order flow and substitutability were rejected in
The foregoing rule change has become effective pursuant to Section 19(b)(3)(A)(ii) of the Act.
Interested persons are invited to submit written data, views, and arguments concerning the foregoing, including whether the proposed rule change is consistent with the Act. Comments may be submitted by any of the following methods:
* Use the Commission's Internet comment form (
* Send an email to
* Send paper comments in triplicate to Elizabeth M. Murphy, Secretary, Securities and Exchange Commission, 100 F Street NE., Washington, DC 20549-1090.