Buy-side dealing desks, tasked with achieving the goal of best execution, have very real challenges in the face of increased trade volumes and ever-rising complexity. Buy-side cost structure is under pressure around the globe, so adding more headcount is often not the answer. The need to manage large volumes of increasingly complex orders and execute them with quality presents an ideal opportunity for technology to fill the gap, delivering the kind of intelligent trade automation which allows trading operations to scale.
Below a certain notional value, for instance, a human trader is unlikely to add much value to a trade. They are not making the best use of their precious time, and likely incurring an opportunity cost in diverting their attention away from a more difficult trade where their skill and expertise can drive more incremental value.
The challenge, then, is to filter out the noise and provide traders with an alternative way of examining their orders. The aim here is to help identify where the high-value opportunities for intervention lie, and let automation do the rest.
In certain ways, trading is already moving towards exception-based management. Technology can be applied to create a framework of intelligent rules that analyses orders when they come in and, if they fit certain criteria, automatically allocates the orders to different buckets for intelligently automated execution. In this scenario, the trader is only notified if something goes wrong, or if a material event occurs that could impact price in an unexpected or unpredictable way, giving the trader a chance to get ahead of it.
The adoption of this framework has proven efficient, and our research indicates that traders now automate as much as 70% of their trades. The residual 30% of their order flow that is managed manually is, not surprisingly, the really hard stuff where human expertise has the greatest impact. The traders get to focus their energies squarely on the orders where they’re going to make the most difference, which is both measurable and valuable—and typically, these orders are more satisfying for the trader to work.
Other techniques coming into play are centered around synthesizing a wide range of internal and external data points to help trading desks determine both how and where to execute their orders. These take into account factors such as broker commissions payments, or ranking brokers based on past performance for a specific name, sector or capitalization band, and then using those factors to devise a weighting system that determines where a given order should be routed.
Artificial Intelligence and Predictive Analytics to Optimize Algorithmic Trade Execution
Algorithmic trade execution is another important area where technology is driving efficiencies and performance. A typical institutional buy-side trader could easily have 12 or more broker algorithm suites on their desk. However, it is almost impossible for a trader to know at any given point in time, with statistical certainty, the optimal strategy to be used to execute a particular order—a challenge that is compounded by the volume of orders they are working on over the course of a normal day.
Furthermore, it is not realistic to expect a trader to master the myriad nuances of all of the strategies offered by their brokers. Understanding these strategies to the degree necessary to deploy them optimally over time for a single order—let alone for an entire day’s worth of trades, many of which will have different characteristics, and often quite different factors influencing price movements and thus the trader’s ability to maximize alpha capture (or minimize alpha degradation, as the case may be)—is a tall order, and technology can help ease the burden.
In fact, helping to alleviate that complexity, while at the same time allowing a trader to get the very best from their algorithm provider of choice—who no doubt has poured lots of R&D dollars into the continued development of their execution strategies—is at the heart of some of the most exciting work going on in trading platforms today.
The cutting edge of execution technology today offers the ability to work within a particular broker’s algorithm suite, applying principles of artificial intelligence and predictive analytics to understand not only the key factors likely to impact a given trade, but also take into account the intentions and the style of the portfolio manager that identified the alpha opportunity in the first place. This all feeds models which tailor the execution profile for each order that a trader feeds into the system.
Once the schedule for an order is determined, the trade is implemented across the different strategies available to it over the life of the order. Whereas a human trader might only have the bandwidth to look at a given order a handful of times throughout the day—perhaps changing strategies a couple of times at discrete intervals—a system with intelligent capabilities will have the ability to monitor all orders continuously, running a tight feedback loop, tracking execution outcomes in real-time against a set of criteria and dynamically adjusting the execution strategy and its associated parameters automatically, in order to maximize alpha capture.
Creating a Single, Unified Trading Workspace
While it’s perfectly acceptable for the middle office to keep the order management system (OMS) on its desk, and fine if the portfolio managers have it to monitor their positions, the trading desk should not need direct access to both an OMS and execution management system (EMS) in order to trade.
Instead, the message we’re hearing is that for any activity a trader might need to conduct in the OMS, whether that’s splitting out an order from a particular account, or aggregating a group of orders for a more streamlined execution, or managing allocations back to the OMS for post-trade processing, the preference – and sometimes even the mandate – from traders is to be able to access these functions directly via the execution platform where they do their actual trading.
Not only is this “single workspace” concept becoming more evident in the interplay between the OMS and EMS, but it also applies to external systems that provide traders valuable color and insights crucial to efficiently executing their trades. Whether that’s incorporating pre-trade analytics from one or more providers, integrating data from market microstructure experts which can assist traders in seeking liquidity, or acting quickly when material news breaks on a stock they’re tracking, traders have come to expect decision support throughout the trading day.
Ideally, all this market information, color, and insight is synthesized from multiple sources and made available to the trader in the EMS—their trading cockpit—in real time, allowing them to harness the power of that information, making it alertable and, ultimately, actionable.