How Exchanges Compete: An Economic Analysis Of Platform Competition

The following is an excerpt from Phil Mackintosh and Michael Normyle's paper titled: How Exchanges Compete: An Economic Analysis of Platform Competition. The full paper can be read here.

Exchanges compete as platforms against each other and off-exchange venues.  Platforms are markets that facilitate interactions among multiple parties. The value of a platform is derived from inter-side externalities, or network effects, with the presence of participants on one side attracting participation on the other.[1] They often feature joint products, which are multiple products or services that are generated by the same production technology and therefore share the same costs.

Exchanges are quintessential platforms.

Exchanges exhibit the network effects characteristic of platforms: trade executions increase the value of market data; market data functions as advertisement for on-exchange trading; listings increase the value of trade executions and market data; and greater liquidity on the exchange enhances the value of ports and colocation services.

Why does this matter?

Competitive markets improve product quality, broaden consumer options, foster innovation and maintain lower prices. These benefits are available to equity market participants because the data shows that the market for exchange services is competitive. Because the market is competitive, the best regulatory framework for the delivery of exchange services—which collectively channel investors into productive opportunities that fuel capital formation and economic growth—is one that fosters competition.

As such, this paper is not just about the Exchange Act, or even economic theory. It is about creating a market structure that is best at supporting capital formation and economic growth through lower cost, higher quality, and more innovative on-exchange trading.

Platforms deliver exchange services efficiently and effectively—and thereby fuel capital formation and economic growth—because, when it comes to platforms, 1+1 is greater than 2.

Platforms in general, and exchange platforms in particular, exhibit positive network effects.  Larger trading platforms offer lower trading costs. As trading platforms attract more liquidity, bid-ask spreads tighten, search costs fall (by limiting the number of venues that a customer needs to check to assess the market), and connection costs decrease, as customers have no need to connect to all venues.[2]  The whole is therefore greater (in the sense that it is more efficient) than the sum of the parts.

This is not to say that smaller trading platforms do not have a role to play.  They provide specialized services that cater to individual customer needs. These specialized services help the smaller exchanges grow by driving liquidity to their platforms, and, if they are successful, achieve the economies of scale that benefit the larger enterprises.  Because the total costs of interacting with an exchange are roughly equal, smaller exchanges offset higher trading costs with lower connectivity, market data, or other fees.  While the mix of fees will change as exchanges grow, the all-in cost of interacting with the exchange remain roughly the same.   

The network benefits of exchange platforms go beyond what is expected from platform theory alone.  Exchanges have a unique role to play in market transparency because they publish an array of pre- and post-trade data that non-exchange venues, for the most part, do not. Non-exchange venues benefit from transparency by using published market data to set their own prices and assess their own trading strategies.

As the SEC recently noted in its market infrastructure proposal,[3] the number of transactions completed in non-exchange venues has been growing. Allowing exchanges to compete as platforms means that they will be better able to compete against non-exchange venues, and, to the degree order flow is shifted from non-exchange to exchange venues, overall market transparency is improved.[4]  Moving liquidity onto lit venues helps non-exchange venues by enabling them to provide more accurate pricing to their customers, and play their own role in capital formation more efficiently and effectively.  

Competition works as an effective system of fee regulation because, as we show in this paper, significant competitive forces constrain combined fees across all exchange products and services, and no exchange is able to provide services that cost more to use than any other in the aggregate.[5]  Although different exchanges have very different fee models for their products and services—and fees for specific services may differ markedly across venues—the “all-in” cost of interacting with an exchange is largely equalized across trading venues. 

This is because not all customers need to purchase all exchange services. In the case of market data, for example, while broker-dealers engaging in on-exchange trading have more of a need of multiple data feeds than other users, they still have the opportunity to route order flow to non-exchange venues that offer price improvement.  More importantly, many market data customers are not broker-dealers and therefore are not governed by the same regulatory obligations.  For example, non-exchange venues purchase market data to set internal bids and offers; investors purchase data to understand the state of the market generally; media companies purchase data to report financial news. Not all of these customers need to purchase market data from all venues; purchasing products from one venue or a subset of venues is enough.  As many proprietary market data products are substitutes, such products compete directly, adding to platform competition and therefore enhancing the overall competitive environment among exchange venues.[6]  As long as a sufficient number of customers can choose a different venue, exchanges are constrained from charging excessive fees.[7]  

Tailoring fees to the customer’s use case, needs and willingness to pay is fairer than cost-based fees or those based on some other criteria. In general, exchanges offer specialized fees in the form of discounts from a standard fee.  Examples of such discounts include non-professional fees, media enterprise licenses, and broker-dealer enterprise licenses. Offering discounts to specific classes of customers allow for broader dissemination of information, and provide customer discounts commensurate with the customer’s ability to pay. 

Reliance on competitive solutions is fundamental to the Exchange Act. Where significant competitive forces constrain fees, fee levels meet the Exchange Act standard for the “equitable allocation of reasonable dues, fees, and other charges among members and issuers and other persons using its facilities,”[8] unless there is a substantial countervailing basis to find that a fee does not meet some other requirement of the Act.[9]  As we demonstrate herein, competition exists at a platform level.  Evidence of platform competition demonstrates a competitive environment for each product sold as part of the array of trading platform services, provided that nothing about the product or its fee structure impairs competition.[10] 

Moreover, the “cost-based” fee analysis suggested in the SEC Staff Guidelines as a possible alternative to a competitive analysis may not be feasible. As discussed below, exchange platforms produce joint products. Economic theory suggests, however, that there are no objective criteria to allocate costs across joint products, and any cost allocation would therefore be arbitrary.[11] 

Further, fee caps eliminate incentives to invest and innovate, inevitably harming consumers, even when those fee caps are aimed at only a single product or subset of products. Moreover, regulating fees for a single product (or subset of products) would most likely not change the “all-in” cost of interacting with an exchange, and therefore any benefit to the customer is unclear.

For all of these reasons, competitive forces create a better and more efficient market structure that incentivizes innovation and efficiency better than any cost-based system of regulation.   

What is the evidence that platform competition works?

The combination of explicit all-in costs to trade and other implicit costs has largely equalized the cost to trade across venues.[12] This is a function of the fact that, if the all-in cost to the user of interacting with an exchange—taking into account the amount of liquidity of the exchange—exceeds market price, customers cease to buy the services of that exchange, and therefore the exchange must adjust one or more of its fees to attract customers. As such, platform competition has resulted in a competitive equilibrium in the market for exchange services, in which trading platforms are essentially price takers, taking into account the all-in cost of interacting with the platform. This competitive equilibrium is a natural consequence of competition, and demonstrates that no exchange platform can charge excessive fees and expect to remain competitive, thereby constraining fees on all products sold as part of the platform. The existence of platform-level competition also explains why some consumers route orders to the exchange with the highest explicit trading costs even though other exchanges offer free or a net-rebate for trading.[13]

The economic concept of platform competition has been recognized by both the SEC and the courts. SEC Staff Guidance published in 2019 acknowledged that platform competition can constrain aggregate returns, regardless of the pricing of individual products, and that platforms often have joint products.[14] The Supreme Court in Ohio v. American Express Co.[15] recognized that, as platforms facilitate transactions between two or more sides of a market, their value is dependent on attracting users to both sides of the platform (i.e., network effects).  Fees cannot be analyzed from only one side, but rather must be considered within the larger context of the platform to test for anti-competitive behavior.

The application of well-established economic theory shows that exchanges are platforms. Exchanges facilitate interactions between multiple sides of the market—buyers and sellers, companies and investors, and traders and market watchers—and their value relies upon their ability to draw customers to multiple sides of the platform, with the presence of participants on one side attracting participation on the other. The competition among exchanges as trading platforms, as well as the competition between exchanges and alternative trading venues, constrain exchanges from charging excessive fees for any exchange products, including trading, listings, ports and market data.  Indeed, the fees that arise from the competition among trading platforms may be too low because they fail to reflect the benefits to the market as a whole of exchange products and services, allowing other venues to free-ride on these investments by the exchange platforms, increasing fragmentation and search costs.

Given that the exchange market is competitive and that exchanges compete as platforms, platform competition is the most accurate model of the exchange landscape and should therefore be central to the Commission’s economic analysis of exchange fee filings.