The electronic execution landscape has undergone a slow but steady transformation over the past decade.
Increasingly seen as a commodity, execution infrastructure provides essential functionality without which it would be impossible to operate, but few brokers see it as a means of gaining any real advantage over their rivals. Today’s differentiators are found higher up the chain; sophisticated algos, execution consulting capabilities and other innovations that sit on top of the execution infrastructure.
More crucial than ever to the bottom line, execution is now something that brokers need to get right. But getting it right has become a lot more difficult.
The costs associated with execution have ballooned during a period when brokers’ margins are already under tremendous pressure. As markets continue to grow and diversify, the execution landscape has become far more complex too. Electronic trading has spread into new regions and new asset classes, each with their own regulatory dimensions. The increasing need to be everywhere and support everything, and yet still do this in a controlled and consistent manner, will define the battle for supremacy between global and super-regional brokers.
The last decade has seen a staggering proliferation of new venues around the world, and the trend is set to continue. In Europe, for example, there were 175 markets registered at the end of 2007 and this has risen to around 280 today. Including the 60 or so markets that have come and gone in between, that’s more than 160 new venues – nearly double in just seven years. The increase globally has been gradual, but stark, and all of these venues come replete with their own subtleties and idiosyncrasies. Additionally, each venue typically makes one or two mandatory upgrades a year – that alone means brokers are running pretty hard just to stand still. But the world isn’t standing still.
Regulation has compounded the issue. A new regulatory consensus has emerged in the years following the crisis, focused on transparency. The regulators’ belief that lit markets are good and OTC is bad means they aim to push trading away from OTC and dark pools and onto lit markets – a trend that shows no sign of slowing down. This will put further pressure on margins, especially for the scale players, and will accelerate the spread of electronic trading to new asset classes.
In the US the move to push OTC derivatives trading onto exchanges has resulted in the creation of over 20 SEFs (Swap Execution Facilities) since 2013. No doubt this number will fall, but at this stage brokers have no way of knowing for sure which will make the cut in the long run. In the absence of a crystal ball, they need the flexibility to switch in and out of new markets with speed and ease, and the ability to test the waters of new markets with minimal risk.
A number of European reforms underway, such as Mifid II, serve to further highlight this global regulatory trend and place execution quality firmly in the spotlight. Even though decisions about when and where to trade are made above the execution layer, brokers will need to make major changes to their execution infrastructure in order to support new monitoring, control and transparency requirements.
This shift in landscape has been dramatic, but because it’s been gradual many brokers’ electronic execution capabilities have evolved in a patchwork manner. Extending an existing system beyond its original intended scope carries its own risks, and often the cost alone is prohibitive. Instead, market access infrastructure – whether developed in-house, outsourced to multiple vendors, or some hybrid of the two – has been bolted together piecemeal as firms have expanded their capabilities and moved into new regions, new markets and new asset classes. The upshot is a multitude of systems underpinned by different technologies, duplicating routes to market in some areas, and failing to provide any route in others – with all the resulting inefficiency this entails.
A realisation is dawning that this approach no longer suits the vastly different trading environment that brokers now find themselves in.
Time for a rethink
To be successful, brokers need to provide their buy-side clients with globally consistent trading behaviour. At the same time they need to be able to deliver their individual USPs through this client experience. Performance, reliability and cost are all part of the mix too. In the face of all these challenges, many are asking themselves some fundamental questions about their existing approach to execution infrastructure and the sustainability of their in-house and/or multi-vendor systems.
The key to success lies in the pursuit of consistency. Above all, this means creating a single, unified execution layer across the entire business that insulates the trading function from everything underneath it. No matter the geography or jurisdiction, the broker should be able to offer its clients a uniform trading experience. Whether a traditional trader or algo developer, the true value comes from deciding what to trade, where, when and how, not in understanding the subtleties of each market interface. Users should be able to send whatever orders they need to whichever markets they want, without having to worry about what order types markets support, which interface to use, or how long the order will take to get there.
This requires much more than simply changing the format of an order message; it requires a system that understands the regional nuances as well as the specifics of the market. For example, it needs the ability to create synthetic order types on exchanges where an order type is not supported.
Achieving consistency is a challenge in itself, achieving it on a global scale with reliable performance is harder still. Historically, firms have tended to treat performance in one dimension – almost like a 100m sprint – seeking to achieve faster and faster top speeds, sometimes even at the expense of other factors. Interestingly, as execution has steadily become commoditised, these other factors have come to the fore.
Firms are realising that execution is in fact a multi-discipline event, more akin to a decathlon. Yes, speed is very important: you’re not going to win unless you can run fast, but a successful decathlete needs to be much more rounded, able to outperform in a number of quite different disciplines. He needs brute speed, but also copious amounts of stamina, technique and strength, not to mention accuracy, reach and flexibility.
Brokers are starting to see the importance of treating execution performance in a similarly multi-dimensional way. The ability to be on time every time is far more important as a measure of latency than the ability to be fastest on the field under time-trial conditions. If some orders take tens of microseconds and others take thousands, then it is near impossible to define a reliable execution strategy. Latency needs to be consistent, regardless of throughput.
Latency measures must also take into account the entire process. Apart from the obvious trading tools, algorithms and SORs, a decision can’t be made without market data and an order can’t go to the market without undergoing risk checks. All of this takes time and adds to the effective latency of the system. A market gateway that takes a couple of microseconds is of little use if the risk checks in front of it amount to several milliseconds.
Reliability is another essential piece of the puzzle. Performance, reach and intelligence will count for nothing if the broker cannot rely on the infrastructure day-in, day-out. As market complexity and the regulatory backdrop continue to change, reliability is becoming an ever-more challenging goal. Many firms are finding that achieving this goal through a patchwork of electronic trading infrastructures is all but impossible.
The final dimension to consistency is cost. It becomes very hard to understand cost, let alone control it, when it’s spread across multiple vendors. Whether or not these silos remain at the organisational level, there’s still huge scope for reducing fixed costs – connectivity, datacentres, developers – by collapsing them at the technology level.
Control is crucial. Historically, a number of brokers have wished to deploy their own competitive differentiation around their electronic execution desks. For example, firms may have developed their own more efficient risk management systems or more effective algorithms. Unfortunately the proprietary nature of these services has led firms into either having to build out their own execution layers or build multiple connection points into third-party systems. Ideally what is required, then, is the ability to plug these services into a single control surface so that their competitive differential is maintained without the exorbitant cost of managing the market access complexity.
A new foundation
Scale players and global or super-regional brokers will soon find themselves struggling if they fail to act. In time, those that don’t adapt will be smothered by a perfect storm of cost, regulation and inefficiency.
Key, though, is for brokers to have the freedom and flexibility to connect the execution layer to whatever differentiators they have sitting higher up the chain. Ultimately, the point of an execution layer is to give brokers the best possible market access through which to innovate and effectively deploy their own competitive services.
Make no mistake: while the commoditisation of the execution layer means that it is the innovation on top that affords the real competitive advantage, failing to get execution right will render this impotent. Commodity or not, powerful, comprehensive, reliable, smart and consistent execution will remain a crucial ingredient for those firms that truly want to compete on a regional or global scale.
Will Winzor-Salle is an Electronic Executino product specialist at Fidessa