As the Financial Information eXchange electronic communications protocol (FIX Protocol) gains further traction in a greater number of asset classes, it brings with it significant synergies for certain types of less-latency-sensitive market participants. In this chapter, W. Brennan Carley of Thomson Reuters explains how global standards continue to drive synergies. The widespread adoption of FIX enabled not only the rapid expansion of electronic markets but also played an important role in leveling the playing field among brokerage firms, exchanges, and electronic communication networks (ECNs), making it possible for firms to compete on price, execution quality, and other value-add services. Although the FIX protocol dates from the 1990s, its impact is not a thing of the past.
Going forward, FIX will continue to democratize new markets. In addition to its expansion into different asset classes like the OTC markets, FIX has the potential to streamline post-trade processes. As regulatory oversight imposes additional compliance demands, FIX will assume additional importance there as well. Finally, perhaps the crowning achievement of FIX is that it proves that the industry can work together cooperatively to deliver synergies and in doing so, provides a model for future joint development work.
The story of the widely accepted FIX protocol and how it came into being is unique. Standards are rare in financial services technology, especially in this highly proprietary industry. In general, developing and implementing standards entails a great deal of effort, work that does not positively impact a company's profitability. Plus, with such a large and varied group of participants, it is a difficult process to develop universally accepted standards in any area, let alone in the complex and highly customized are of information technology. Given this background, the achievements of FIX are even more impressive.
Brennan Carley is global head of the Elektron Transactions and Enterprise Platform businesses at Thomson Reuters. As head of Elektron Transactions, he leads the firms business in electronic trading in equities and other listed securities, including FIX order routing and indications of interest. As head of Enterprise Platform, Brennan leads the Thomson Reuters' core data platform business that allows financial professionals to control information flow by connecting all of the different applications needed for trading, investment, risk management, and compliance.
Brennan joins Thomson Reuters from Proton Advisors and brings almost 30 years of experience in trading technology and trading communications. As founder and managing principle of Proton Advisors, Brennan provided advisory services to private equity funds, venture investors, and their portfolio companies in trading technology and telecommunications. As a consultant through Proton Advisors, Brennan served as SVP of product marketing at Spread Networks, a provider of ultra-low-latency connectivity to the financial markets.
He previously was COO of NYFIX (now NYSE Technologies), was CTO and founder of financial extranet BT Radianz, and was head of technology for Instinet. He has been a director on the boards of companies including Marketcetera and Yipes, and is an adjunct professor of business and technology management at New York University.
Brennan earned a bachelor of arts degree in economics from New York University and has done graduate studies at Columbia University, Harvard Business School, and Massachusetts Institute of Technology.
The FIX protocol has done nothing less than revolutionize trading and reshape the entire trading ecosystem. While electronic trading would have happened without FIX (and some electronic trading preceded FIX), the rate at which the equities markets have converted to electronic trading would never have been achieved without FIX. This has lowered the cost of trading, enabled entirely new trading strategies, and has accelerated the shift in power in the industry to the buy-side.
When FIX was first developed (by Salomon Brothers and Fidelity investments), it did not initially appear to be a foregone conclusion that FIX would gain widespread adoption. There were a number of proprietary protocols in use at the time, some developed by brokers and some developed by stock exchanges. Many of those exchanges and brokers (including Instinet, where I worked at the time) would have preferred to keep their proprietary protocols; they were up and running, and once a customer had programmed to them, they were less likely to switch. So brokers and exchanges with existing protocols had a lot of reasons to resist FIX. But the industry quickly embraced FIX for three reasons.
Electronic trading was still in its infancy when FIX was developed. While screen-based trading had been around for a while, the idea of programmatic trading (i.e., machine to machine communications) was still very new, and the use of an order management system (OMS) or execution management system (EMS) was also quite new. These things needed message-based communications protocols, and FIX came along before any other protocol was well established. Even though several exchanges and brokers had proprietary protocols, the use of these protocols was very limited and the market was fragmented. No one protocol was dominant. So FIX was not trying to displace an entrenched proprietary standard, it was trying to displace a babble of protocols. Probably the most important was that customers, the buy-side, wanted it. Sophisticated buy-side firms like Twentieth Century (now American Century) and Fidelity told their brokers that they wanted to use FIX. While exchanges and brokers might have preferred to continue using their own homegrown protocols, it really was inevitable that an open, customer-supported protocol would displace a bunch of proprietary protocols, none of which was very well entrenched or dominant. Ironically, the initial adoption of FIX was the asset management and the brokerage community, both profit-driven entities, and the last adopters were the exchanges and clearing houses, which at that time were mutually owned non-profit industry utilities.
By establishing an easily implemented open standard, FIX has not completely leveled the playing field; brokers, ECNs, alternative trading systems (ATSs), exchanges, and multilateral trading facilities (MTFs) in Europe have many opportunities to differentiate their execution services. But FIX has eliminated differentiation based on communications technology, and indirectly based on trading front ends (EMSs and OMSs). Asset managers and hedge funds can use virtually any EMS or OMS to trade with just about any broker, ECN, or exchange. The technical barrier has been lowered by FIX, and it has become much easier for the buy side to switch from one broker to another. The FIX protocol has standardized messaging formats, and FIX networks have made the physical connectivity to many destinations easier and more cost effective. This has allowed electronic trading to become much more friction free, reducing cost, and allowing brokers and ECNs to compete on the basis of execution quality, market impact, and other services.
FIX has created whole new businesses and enhanced profitability for some participants, while driving down profitability for others. For the buy side, FIX has reduced the cost of trading and created greater competition for their order flow. It has enabled enhanced transparency of execution costs, which has contributed to declining costs of execution.
FIX has enabled the creation of FIX-based businesses, ranging from OMS and EMS platforms to FIX routing networks. Other businesses have sprung up to provide tools for trading, based on FIX. And of course FIX has made it possible for alternate trading venues (ECNs, ATSs, MTFs) to be established and attract liquidity.
While FIX has reduced cost and improved profitability for the buy side and created a whole industry, like any disruptive technology it has contributed to decreasing revenues for brokers and exchanges in the face of intensified competition. Some of those have thrived by driving down their costs; others have suffered with high cost bases and decreased profitability.
On the one hand, standardization reduces the opportunity for innovation of the thing being standardized. By standardizing a single trade communications format, it is much harder for any firm to offer new innovations in trade messaging. It also slows industry innovation in trade messaging, because it requires some level of consensus and cooperation. Imagine if Apple had to go to a committee every time it wanted to offer a new feature.
On the other hand, standardization accelerates innovation in two ways. It allows innovation to occur within the context of FIX (i.e., FIX Protocol Limited [FPL]). Once an innovative development is agreed upon, the adoption of that innovation can occur much more quickly. So the development of innovative technology may be slower, but the dispersion of that innovation into the market can be much quicker. Then, by standardizing messaging, the industry has been able to shift its resources from developing “yet another messaging protocol” onto more useful forms of innovation that build on top of FIX.
While there are battling standards for things like symbology and XML-based standards for electronic documents (e.g., XBRL), in the area of messaging standards the main ones besides FIX are ISO 20022, SWIFT, and FpML. FpML is optimized for over-the-counter (OTC) derivatives, where deal structures and trading workflows are much more complex than for instruments traded on-exchange.
SWIFT has developed several messaging standards over the years, and focuses primarily in inter-bank payments messaging, and more recently on post-trade messaging in securities.
ISO 20022 is really an umbrella standard that allows messages to be described in a standard way. Current versions of FIX, FpML, and SWIFT messages are all described using ISO 20022, and all of these organizations work together.
While FIX started as a U.S.–centric cash equities protocol, it quickly evolved to embrace global markets and other asset classes including FX (foreign exchange) and fixed income. Customers can connect to global FIX networks that enable them to establish their presences around the world. This makes it possible for a U.S. asset manager to connect its OMS to a broker in Brazil, or a hedge fund in Hong Kong to connect to a broker in London, just as easily as it is to connect to a domestic broker. In practice, of course, while FIX solves the connectivity problem, brokers and buy sides must still establish credit arrangements, clearing and settlement, [and so forth]. And while a broker in Tokyo can use FIX to connect to an ECN in New York, all of the commercial arrangements must still be in place.
Until recently, this has meant that global trading took place by using global banks. The hedge fund in Hong Kong would trade with the Hong Kong office of a global bank that had a brokerage license and exchange membership in London. Following the financial crisis, however, many banks with global ambition have consolidated their operations and focused on their home markets. French banks with formerly global ambition have slimmed down their Americas and Asian operations and focus now on the Eurozone. Japanese banks have done the same with their European and Americas operations. A model that is re-emerging is one in which brokers in one country have relationships with friendly brokers in another country. So the hedge fund in Hong Kong might have a (FIX-enabled) local relationship with a broker in Hong Kong that routes the order (via FIX) to a UK bank that, in turn, routes the order (via FIX) to market in London. The standardization on a single protocol makes it easier for cross-border activity to take place, even as market structure continues to evolve.
Within the cash equities market, FIX adoption is pervasive across all market participants. It began with participants who were either large enough to have sophisticated in-house order management and trading systems, or who had systematic trading strategies (e.g., pairs trading, statistical arbitrage) and the brokers who facilitated those customers. It spread from there, and is now widely adopted by all but the smallest firms (e.g., those who outsource their trading.)
As FIX is being adopted in FX and in fixed income, it is following the same trajectory as it did in equities, but with an accelerator: Firms that have already implemented FIX for their equity trading, especially those who have cross-asset trading strategies, have been among the first to adopt FIX for FX and fixed income by leveraging their existing systems and skills.
As FIX is being adopted in FX and in fixed income, it is following the same trajectory as it did in equities, but with an accelerator.
Over the next few years, I expect that FIX will be broadly used across all exchange-based (or “exchange-like”; e.g., swap execution facilities (SEFs) and organized trading facilities) asset classes, a category that will expand as regulatory forces push more trading onto an exchange-like model.
FIX started in the [United States], but is now truly global, with leadership coming from all areas of the globe. FPL, the governing body for FIX, has active development taking place from participants in Europe, Asia, Latin America, and the United States.
FIX is the protocol of choice for all but the most latency-sensitive trading. For firms that are pursuing purely latency-based trading strategies, or where low latency is a substantial element of their alpha capture, exchange proprietary and binary protocols are generally used. While this kind of trading has had high visibility in the press and accounts for a substantial portion of daily volume, the number of participants who engage in these strategies is fairly small (around 40–50 globally), and they tend not to use standard software or protocols of any kind. They are like the Formula 1 racecar drivers, happy to trade off cost-effective and standard systems in return for a few microseconds of performance advantage.
FIX has had the most impact on the pre-trade through trade stages of the trade cycle, which makes sense when one considers that the post-trade world was (and remains) very centralized and standardized around a small number of industry utilities (e.g., Central Securities Depositories).
FIX has made some inroads into the post-trade market, in particular in the area of trade allocations, and I expect that this will continue to grow as it is a natural extension of the existing FIX-based workflow.
FIX has had the most impact in order routing, initially in cash and block trading, and more recently in algorithmic and systematic trading—delivery of orders from buy-side firms to brokers. With our current fragmented market structure, institutions have increased demand to get trades done in block size, and I expect that FIX-based messaging will have a growing impact on the block trading world, primarily in the form of indications of interest.
Where proprietary interfaces are not dominant or where there are multiple proprietary interfaces to choose from, FIX has an opportunity to grow and replace such interfaces. But where there are well-established proprietary interfaces, there is little opportunity (or advantage) for the industry to switch from those to FIX, at least not in the near term.
The choice of FIX versus proprietary standards is really driven by two things. In the case of the ultra-low-latency community, proprietary protocols are chosen for pure performance reasons. In other areas, the primary drivers are market factors; [that is] in segments of the market where there a small number of dominant players (such as central securities depositories (CSDs) or some of the derivatives exchanges) with established proprietary protocols, the market will use those protocols. In areas of the market where there are many competitors, competitive intensity is high, and barriers to entry are low, FIX is a much more attractive standard.
The primary area where FIX is not used extensively today, and which could benefit from greater FIX adoption is post-trade, which would allow greater reuse of FIX-based systems/technologies, and end-to-end processing of messages. Given the number of well-established protocols in this area by dominant CSDs and other industry utilities, it seems unlikely that we will see substantial post-trade adoption of FIX in the near term.
FIX (and FPL) helps with risk management in three ways. First, the adoption of FIX as a standard trade messaging protocol has made it easier for firms to build risk management systems. Second, FPL has defined a set of industry best practices for pre-trade risk checks, which of course can be implemented by systems that intercept FIX-based trading messages. Of course, these same best practices can also be used on proprietary messaging interfaces as well. Finally, FIX is also being used as a messaging protocol to provide reporting to regulators.
Regulators are struggling to catch up with the explosion in trading volumes and order volumes, and need visibility of trade messages to better understand and surveil (both in real time and afterward) financial markets. Adoption of FIX messaging has helped regulators to perform this function.
As instruments that have traditionally been traded OTC become more electronic and shift to an exchange-like model, FIX will be used as a standard messaging protocol. FIX already has functionality for trading OTC instruments, and FPL is working with ISDA (International Swaps and Derivatives Association) to develop best practices for trading interest rate swaps and credit default swaps using a combination of FIX and FpML. On the clearing side, for example, FIX 5.0 supports messages for standardized credit default swaps for clearing. A number of derivative clearing houses are using FIXML today for post-trade messaging for clearing listed derivatives, and many are adopting these systems to provide OTC derivative clearing as well.
While the cooperative nature of development leads to slower development of new technologies, it has some significant benefits. It leads to more robust standards, because more participants have had the opportunity to validate the technology. Plus, the industry has greater buy-in to the technology because it has been developed in a cooperative manner. Once defined, the rate of adoption can be greater.
The single biggest technological achievement in the financial services industry has been the automation of the trade execution function. While electronic trading is not without its flaws, and the markets have seen vulnerabilities such as the Flash Crash, it is easy to forget what the markets were like before the introduction of automation. Trading costs were higher, as measured in commissions, in spread, and in market impact. Markets were less liquid, and it took more time and money to get trades done in thinly traded securities. And the markets were more opaque and more vulnerable to conflicts of interest and rigging of prices. While not perfect, electronic trading has reduced costs for investors; made it easier, more reliable, and more transparent to trade; and has enabled new kinds of investments altogether (such as exchange traded funds [ETFs], which have reduced the cost for investors to get exposure to broad segments of the market).
I can't predict the next “quantum leap” in trading technology for three reasons. First, we are operating in an environment that is characterized primarily by greater regulatory focus and increased pressure on costs. Most of the focus of the industry is therefore concentrated on cost reduction, standardization, and consolidation of functions (e.g., into shared utilities, whether for-profit or non-profit.) In the [United States], the financial sector's share of GDP has grown over the last 60 years from 2 percent to over 8 percent, a growth rate that is clearly not sustainable, which is one reason we are seeing consolidation of the financial sector today. Technologies like FIX are valuable tools to create efficiencies, so I am optimistic about the continued growth of FIX. But it is hard to see something that will revolutionize the industry.
As tempting as it is to look for revolutionary technologies, most technologies are incremental improvements on what was there. FIX was an evolution of earlier trade messaging protocols (at least in concept, if not in the precise syntax). Over the next few years, I expect most technology development in trading will be evolutionary, and a lot of it will be adopting and extending existing technology (including FIX) into areas that are today less automated.
Finally, as the physicist Nils Bohr (who received the Nobel Prize for his work on quantum physics, so who knows something about quantum leaps) said, “Prediction is very difficult, especially about the future.” So everything I just said could be totally wrong.
While it is important to get the technology right, technology rarely succeeds or fails based solely on the technology. Understanding the market that you operate in is critical to the success of any technology. Betamax was arguably a “better” technology than VHS, but VHS ultimately won. It won because it had a longer recording time and was more readily licensed, which led to greater availability of content. And content drove the success of VHS players. Ultimately any technology needs to succeed in the marketplace, and good technology is only one variable. FIX succeeded not because it was better than alternate technologies, but because it made it easier for buy-side institutions to build one system to connect to multiple brokers, which lowered their costs. And in that market, the bargaining power of the buy side was and remains greater than the bargaining power of the sell side. To really understand these dynamics, I would encourage any technologist to read Michael Porter (Harvard Business School), and in particular study Porter's “Five Forces” model.
Key Learning Points
By establishing an easily implemented open standard, FIX has not completely leveled the playing field; brokers, ECNs, ATSs, exchanges, and MTFs in Europe have many opportunities to differentiate their execution services. But FIX has eliminated differentiation based on communications technology, and indirectly based on trading front ends (EMSs and OMSs).
The adoption of FIX as a standard trade messaging protocol has made it easier for firms to build risk management systems. FPL has defined a set of industry best practices for pre-trade risk checks, which can be implemented by systems that intercept FIX-based trading messages.
As instruments that have traditionally been traded OTC become more electronic and shift to an exchange-like model, FIX will be used as a standard messaging protocol. FIX already has functionality for trading OTC instruments, and FPL is working with ISDA to develop best practices for trading interest rate swaps and credit default swaps using a combination of FIX and FpML.