Trends in Electronic Trading and the opportunities they bring.

Background

In today’s environment, scalability, profitability and performance continue to be the fundamental drivers for traders. A deciding factor to enable this is the ability to access liquidity and execute against it in a cost-effective way. Technology has proven itself to be a key element in a trader’s ability to access, execute and manage trades. How a trader decides on which technology to use for order execution and order management is increasingly complex.

One of the great debates revolves around the issue of integrating and migrating trades from front-end execution to back-end clearing and settlement — to manage trades throughout the trading process. Most software providers have historically taken a compartmentalised approach to the process, focusing on one aspect of trading, for example, trade execution or order management.

This is an approach based more in technological capability than on trader need. In a way, the traditional OMS/EMS debate is irrelevant because it focuses more on the technology than on the underlying forces that drive trading behaviour. The real issue underlying the OMS/EMS discussion is not which system is the best, but which one addresses workflow seamlessly without commingling different applications. If asked, sell side clients want one screen and preferably one platform where they can perform the majority of the tasks associated with the order/trade lifecycle. The fact that it is still being debated has more to do with the technical challenges of integrating systems that were originally designed for a specific part of the workflow. If an OMS vendor acquirers an EMS vendor it does not automatically create an OEMS. The products are still the same, albeit repackaged and marketed as one solution. New technology must be leveraged where the workflow is seen as a process of walking the order from A-Z already in the design phase.

Current Trends driving the OMS space

System Consolidation

In our client conversations we notice a few clear trends, the most prominent one being the desire to consolidate the sales and trading activities on one platform, regardless of flow type or asset class. Clients typically want to be able to have a scalable platform that can handle more and more exception types. From primarily handling market rejections or limit breaches, the exceptions can be generated by underperformance, liquidity, instrument type, client or whatever arbitrary reason. Clients are also looking to outsource the management and maintenance of the platform by signing SLA:s with details on hardware, service hours, uptime, etc., effectively “renting” the technology stack to get a clearer view of the total cost of ownership. Onboarding of new clients is key and the time from landing a new account to have it sending orders have to be minimized. This is an area sell side have been reluctant to give up control over, but with better onboarding tools and automated testing, this will most likely change in favor of a hosted managed model.

Regulatory Compliance

In attracting order flow, the ability to evidence compliance with regulatory requirements, such as best execution, have taken center stage. The protection for the buy side have been expanded by an update of “Best Execution” as defined by MiFID II, the most recent update in what is now a Global movement of regulatory tightening. The regulatory frameworks are also expanded into new asset classes, that to some extent have been unregulated, at least from a Best Execution point of view. Derivatives and Fixed Income are areas that will experience considerable change taking into account the full effect of new regulation. Trade reporting and transparent audit trails must be an integrated part of the workflow. Particularly challenging is to consolidate all order related activities into a readable and easy to access format.

Stability

Another important factor to consider is stability or reliability which, in today’s electronic markets, hinges on the performance of the technology that constitutes a broker’s trading infrastructure. Typically, this stack will be made up of various, interrelated components such as FIX engines, order management systems, market data platforms, smart order routers, algo engines, etc. Each of these represent an additional point of latency and/or failure – scenarios that could impact execution and, in turn, cause clients to defect to another broker.

Automation

Humans interacting with systems through a keyboard is very error prone. Just the fact that there is a need for a “backspace” or “delete” key, pretty much says it all. One way of avoiding this is to look at automating some of the more routine parts of order processing. To a large extent, this has been done already. Orders are sent electronically directly from the client’s OMS/EMS and routed automatically to the relevant market. The trades are received and allocated to the order in real time and once the order is fully filled it is automatically booked. This is an automatic chain of events supported by a number of different systems and is also only possible when the clients have taken the trading decisions themselves or transferred that decision to an algo.

When an order is to be handled by “care”, the decision is instead transferred to a sales trader, albeit based on some sort of order instruction. In the past, clients relied on their Sales Trader’s knowledge, experience and network of other clients to achieve best execution. With electronification of markets and workflows orders leave electronic footprints that is stored and analysed. This vast collection of data, optimally used, opens up for robo-trading on a whole new scale, partly replacing the knowledge the Sales Trader once monopolised.  

Convergence of Flow types

In the infancy of electronification and low touch, markets weren't really fragmented and Low Touch was basically a way for a Buy side client to place an order “directly” on the market using access sponsored by their broker. This was mainly used for non-complex orders such as market- or limit orders in blue chip stocks. Everything else was basically routed to the High Touch desk as care orders. As the market complexity increased so did the use of algos. First, as a High Touch tool for the Sales Traders, but soon access to algos was offered directly to the buy side, essentially giving them the same tools as the traders on their broker’s desk.

Low Touch evolved into something a little different with clients using algos. Order had to be monitored more carefully and parameters could be entered wrong to trigger a rejection or other error. The flow started to look more like High Touch. On the care desk, trading was mainly about algo selection if the client had not provided one with the order and this selection process was more and more subject to automatic “rules”, introducing low touch features in high touch.

The manual order management is becoming more about handling exceptions due to market rejections, illiquidity or deviation from the relevant KPI’s. The underlying workflows are becoming more or less the same and with block discovery venues and buy side liquidity pools for large in scale orders and the separation of HT and LT is even more blurry. With flow converging, it is time to revisit the separation of flow in the infrastructure, since the order may start as “No Touch”, triggering an exception requiring some minor trader intervention and ultimately becoming an order in need of a fully manual workflow. Orders need to be able to passed around in the system without affecting latency, stability or performance. Clients are expecting some level of service regardless of order channel so the labeling of Low Touch and High Touch will soon be a distant memory. Humans are needed to provide context and define the rules and that is not changing in the short- to medium term.

Leveraging the front office

”Post crises” staff reductions are levelling off, but the industry is roughly at 70% of pre-crises levels and seem to hold there for the foreseeable future. The front office is following their orders further into the life-cycle, often performing middle office services like allocation and book outs. With flow converging, coupled with lower average trade sizes and fragmented markets, this translates into more trades per trader and further processing by the front office. The industry is turning to technology to solve that equation and want to see a system that is flow agnostic and can handle multiple asset classes, with a large degree of automation for true STP.

Status today

The industry has seen numerous technology solutions, beginning with the automation of the back office, and the development of order management systems in the 1990s. As traders’ needs have come to be more precise, technology firms have responded, resulting in trade execution tools (e.g., algorithms, transaction cost analyses, etc.) and order management systems (e.g. trade allocations). This sudden influx of technology has left many users with multiple systems and legacy issues which continuously need to be either integrated or updated.

MiFID (I) was a great catalyst for technical development and brokers were rushing to offer advanced execution tools to both buy side clients and other sell side participants. Mainly, the offer consisted of EMS capabilities exclusively tied to the brokers infrastructure. It was not uncommon for buy side clients to have 2-3 or even more brokers through which they had Direct Electronic Access to markets. At first, the service was offered by a stand-alone PC on the desktop, but with FIX integration this was soon integrated in the client’s OMS to some extent. This may be a viable solution for buy side where there is no need for example trade allocation.

With increasingly complex flows, transparency issues and other factors, the need for a neutral and independent solution started to become more apparent. Instead of employing multiple systems integrated via FIX, the integrations will have to be deeper and seamless. Partly because of latency, but mainly because of usability and workflow.

Opportunity

However, with major parts of the industry’s offering stuck in legacy platforms, there is a window of opportunity to offer something else. Something that is flow and asset class agnostic, truly global and based on a modern platform that scales well and offer the flexibility to sidetrack from the main architecture without it becoming an issue. Technology choice is crucial and the underlying architecture must be able to cater to workflows that look completely different around the globe or across asset classes. It does not necessarily have to do everything a classic High Touch system can do today, mainly because the workflows are changing. We need to be able to predict or even suggest how our clients would/should be set up in the future, but most importantly, offer the flexibility to adapt to different internal and external factors. The power of an app based, modular platform becomes evidently clear.

To really capture and benefit from the trends the industry needs to analyse in more detail what each trend mentioned above suggests in terms of functionality or architecture. Let's walk them through one by one:

1.    System Consolidation

What does it really mean? Partly it is a cost of ownership issue, but also a response to the trends of flow convergence and asset class electronification. It is also a staffing and stability issue. Ideally a system should be bespoke to each purpose or situation. That is quickly a recipe for disaster as the solution must be implemented on site and customized to fit the need of the client. That inevitably becomes expensive and hard to manage.

A solution to this problem is to find commonalities between asset classes and develop functionality and workflows that work in a high throughput environment, while still being able to manually insert and manipulate orders and trades. The platform should be a flexible, yet resilient infrastructure or framework with monitoring capabilities on top of which it is possible to add or remove more advanced functionality.

1.    Regulatory Compliance

Regulatory tightening is a global phenomenon. There is a definitive push against transparency, which from a vendor perspective is good news. To be able to stay compliant, orders must be tracked throughout their lifecycle and every event needs to be recorded. The best way to do this is by keeping electronic audit trails and vendors will have to be able to store and access large amounts of order and trade related data, without adding latency to the front end users. Transparency for the user of the system is equally important and keeping the order in the same platform makes it easier to achieve this, especially if the algo box and SOR are part of the build. There is also an increased demand for trade reporting to authorities, which also have to be handled correctly and without delay.

2.    Stability

All the bells and whistles in the world won't make a difference if the platform is unreliable. When listing the most important factors for platform selection, stability will come as number 1, 2 and 3. This has partly to do with the outsourcing trend. With in-house or bespoke systems, clients had to have IT people on site. It is expensive, but the response time is short and emergencies are tended to without delay and without having to compete with other clients for attention. This, together with customisability is one of the best arguments for a deployed solution. However, clients are now willing to trade some of that flexibility away for a lower and more predictable cost, but it means they will value a vendor with a reputation of being stable.

Stability is achieved by standardisation across the client base and a sensible release cycle with minor upgrades. It is maintained by streamlining functionality and having a tight control what is put in the system, but more importantly, what is not put there.

3.    Automation

Automation have been one of the main drivers from the very beginning. Achieving Straight Through Processing have been crucial to reduce the overall transaction cost in the industry and can be said to be the first phase of automating the workflow. Phase one enables a buy side client to select broker, select strategy, pre-allocate and send the order for execution through sponsored access. At the broker, the order is received electronically and automatically routed. The allocation instructions are applied and confirmed and the order is sent to back office for clearing and settlement. This is a complete order cycle and typically how a low touch platform operates.

However, phase two of automation will include what is considered “high touch” today and require analytical capabilities within the system, or using an external tool integrated with the workflow. Specifically we want to be able to automate the investment decision itself, as well as the strategy selection. There are a number of tools being developed at the buy side that involves the investment decision, meaning the decision to buy or sell a certain instrument. That part of the business is beyond the scope of a sell side OMS and instead we want to focus what happens when the investment decision have been made. To be able to suggest a trading strategy we need access to order and trade history, trade patterns, access to third party ownership data, performance indicators and ranking for all accessible algos, in house as well as brokers. In theory we can also store data on how a specific trader is performing when trading certain instruments and suggest “manual intervention” if the trader is consistently performing better than algos.

Automation is key to free up time for traders to deal with the an increasingly complex liquidity situation where humans really can add value. Freeing up time by automating the more straight forward orders to pursue the hard to find blocks and leveraging their relationships will be key for a successful Sales Trader.

4.    Convergence of flow types

Low Touch has evolved very quickly. From being the minority part of the flow by offering clients access to the market via the existing infrastructure and adaption of High Touch systems, Low touch in especially equities is the major flow type. The flows have become more complex and direct access to strategies is now the norm. High touch is now to some extent limited to very large blocks or illiquid instruments that depend on the broker′s network of clients. As described above, even that part of the business relies more and more on data processing, by cross referencing orders against ownership data or trading/order history. Low touch being injected with more and more complex strategies, means that performance monitoring is requiring more attention. The fragmentation of markets and lack of a golden source of static data increases the likelihood of rejections, which in turn creates a need for exception handling tools similar to those used in the High Touch part of the business.

The Convergence of flows requires a consolidated system to be able to scale and seamlessly move orders from no interaction to manual processing by the trader and back again. The platform needs to be flexible enough to allow for order and trade manipulation and for actions taken on a single order or multiple orders.

5.    Front Office leveraging

It is not uncommon that equities and equity related instruments are handled by the same Sales Trader. This means that a system needs to support the workflows of equities and derivatives, as well as allow for ETFs and funds. If we move further into the life cycle of the order, allocations are often performed by the front office. For middle officers, the need to allocate and confirm instruments cross asset classes is becoming more frequent. This require multiple systems and ideally this could be solved by consolidating into a single platform, with similar user experience and monitoring abilities.

Finally, it is all about Finding the Buy and build sweet spot.

Tier 1’s and large Tier 2’s have historically almost always opted for in-house development. The reason for this have mainly been tailoring a solution with a near perfect fit to the business requirements. Other reasons may be that the Tier 1’s have hard to find a partner that can meet the rather tough requirements of uptime and reaction time. As the technology evolves, some elements once considered complex are commoditised. Connectivity, order monitoring, SOR, TCA etc., are mature areas and firms need to consider spending their intellectual capital where it makes most sense. Outsourcing the commoditised components of workflow and infrastructure to free up capacity for in-house development where the firm has an edge will make them more efficient and ultimately more profitable. The vendors need to offer open API:s and sometimes even provide the source code allowing firms to find that perfect blend between buy and build.

Hope you enjoyed the read! Please comment or reach out for further discussions on the topic.

Lars Wiberg, VP Product Management,

Itiviti