Whether a broker is executing a client’s orders, offering direct market access (DMA) or providing sponsored market access (SMA), it is constantly taking on risk. The priority is to ensure that those risks are always visible and never build up.
But given the sheer speed and volume of order and trade messages that brokers are in one way or another involved in, managing the risk in real time is a formidable technological challenge, not to mention a costly one.
To illustrate the scale of the problem, consider this: The number of trade-related messages a typical broker is processing on any given day can easily reach into the multiple millions. Worse yet, that massive volume of messages, involving bids, offers, amendments/cancellations and executed trades, can be at dozens, if not hundreds, of venues all over the world, each of which may require different APIs for their gateways.
Cost, scalability, and performance are all key considerations for any solution. It stands to reason that the higher the performance, the better the risk management will be because it will capture all of the key trade detail with minimal lag. This is particularly the case if some proportion of the flows relate to high-frequency trading activities. And since financial flows are always shifting, scalability is paramount.
But all of that comes at a high price. Brokers need to write and maintain APIs for all of the venues they trade on, capture the receipts for the trades they have executed and listen in on those that are done via DMA or SMA.
For firms offering client flow, and potentially multiple different flavors of it as a full-service bureau, being able to effectively monitor and manage that flow is key. This requires having a holistic, real-time view of the flow to manage risk and also to service client requests. Having a slow-performing system is not only visible to end clients but also impacts them at the time when they rely on the service the most.
Many solutions have involved consolidating drop copies of trades in a centralized system, based on FIX protocols. However, firms can have multiple disparate flows and technology platforms and this adds to the challenge of bringing all of the flow onto a single system. Another challenge is that traditionally, much of the focus, in terms of technological investment, has been on the front end rather than the back end. These types of periphery systems often don’t get the same investment and focus as those on the critical path for order entry (although that is starting to change given the regulatory requirements that have been coming through recently).
Then, once these drop copies are processed, there is still the question of normalizing the data and arriving at a consolidated view of risk. That, in itself, is no mean feat when you consider the volumes and variety of orders and trades that brokers are engaged in.
So why not adopt the same performance standards for post-trade as for pre-trade and execution? For instance, at Vela, using low-latency gateways that have been built for high-speed trading, we normalize trade data at the point of execution, instantly sending it back in a format that allows it to be consolidated into a holistic risk assessment in real time, all within a hosted solution. Key to this is that the solution is using the exchange-level drop copy, not the individual client’s copy.
This has several immediate benefits. For a start, it simplifies the workflow. A broker can thus focus on the part of risk management that needs it, namely the risk reporting and analysis, rather than the collection of trade receipts. Meanwhile, the cost benefits in terms of saved time and effort are clear.
One additional benefit that may be easily overlooked is that outsourcing the trade normalization and risk management can also affect top-line revenues. Any broker that is lacking in performance in terms of message capture and normalization could run into trouble in trying to onboard new clients that want to trade at high frequency, for example.
Such systems need to be able to provide kill switches, either at a firm or account level, which are compatible with a venue’s protocols. Third-party offerings in this space, such as the one Vela offers, should use industry-standard FIX protocols and to plug into existing infrastructure. We’ve also found that clients want the option of cloud delivery so that data can be accessed anywhere.
Market trends favor pre- and post-trade solutions such as these. While different markets have their ebbs and flows, financial market activity in general is almost always growing. At the same time, SMA is rapidly becoming a preferred route for many clients, so brokers need systems that can cope with the particular challenges that come with that.
On the surface, addressing performance, achieving scalability and minimizing cost may sound like incompatible objectives. But in the area of real-time risk management, a third-party approach based on high-performance gateways and point-of-trade normalization could allow a firm to tackle all three in one fell swoop.