Redefining The Trading Stack
By Ben Jefferys, Head of Trading Solutions, IRESS
Market participants need to take a clear, objective look at their own individual requirements before selecting an appropriate trading system.
Market complexity and escalating regulation are forcing both brokerages and fund managers to install new dealing technologies. The orthodoxy, in part promoted by influential service providers, is that an integrated multi-asset trading system is the panacea.
Typically, we are told that it is the role of the order management system (OMS) to collate order flows from all asset classes within one channel. However, although this is suitable for some firms with sufficient size to bear the high costs of investment and organizational structure to warrant a combined process, it is often less appropriate for others.
Unfortunately, many firms feel they have limited choice, compelled to use compromised or over-purposed solutions which are inflexible and hard to change by entrenched supplier incumbents or in-house solutions. Contracting commission pools for institutional brokers, means there is even further consolidation. Moreover, financial markets are highly segmented and tiered which can mean that all-encompassing multi-asset OMSs can complicate rather than rationalise processes.
Instead, it makes sense to re-define the trading stack. We should firstly recognise that not all sell-side firms are the same, nor are their clients. So it should be accepted that in order to best serve clients now and in the future the solution may in fact be to use separate systems that openly communicate across the stack.
To understand why, it is necessary to clarify the purpose of the OMS. In the first instance, its role is to manage order workflow, routing and execution to and from counterparties, exchanges, algo engines and dark pools. In addition, it should control risk and its design must be robust to capture uptime and resilient to ensure minimal disruption.
But, the selection of the best OMS depends on other factors too. It needs to be open and flexible, yet easily controlled so that clients’ individual risk can be managed. It must also be scalable and adaptable to changing future requirements, which means it should work for the benefit of the user in all circumstances. Too often a firm is stuck with an OMS that is no longer efficient, unable to swap it for a better system because of prohibitive cost or overwhelmed by the complexity of transferring.
Critical to the ultimate decision is whether to opt for a single- or multi-asset model. If multi-asset then what does that really mean? It is especially important now, because the industry is in a state of flux, having to reassess operational practices and structures concurrent with regulatory changes, notably Markets in Financial Instruments Directive (MiFID) II, while making sure it is well-placed to adapt or exploit unforeseen later developments.
Adapting to change
In these fluid times, when all industry participants have to make often costly adjustments, it is right to re-evaluate the trading stack and incorporate an ability to react to future structural changes with minimal disruption.
There are several key issues to address if considering a multi-asset platform, but they all lie within one fundamental question: does that option really suit my requirements? In fact, am I totally clear on my requirements?
For instance, if you need flexibility and specialisation, but lack the economies of scale or the ability to change the whole trading stack at the same time, then an all-encompassing multi-asset OMS might not only be a pipe-dream, but also expensive, defeating the objective of lowering costs. A sell-side firm – even a tier 1 investment bank – whose clients trade different asset classes need to clearly understand what multi-asset capability is required. It’s not a one-fits-all scenario.
Indeed, many clients look for specialisation in certain asset classes while maintaining an ability to handle others in a simpler form. Matching your trading stack to your firm’s core capabilities and structure is key. It is not uncommon to see different OMSs used across the asset classes traded by the sell-side.
This model has ramifications for the traditional middle office. But it need not be directly coupled to the OMS whereby each system used has its own middle office function. Instead it is shared by each vertical system. The sell-side can then bring it all together post-trade when the middle office becomes agnostic to the trading process. At all times this middle office layer remains informed; it is not a disconnected process.
This model creates an independent middle office function, integrated with several discrete OMS and execution management systems (EMS). This normalises the middle office workflow and interaction with third party trade processing agents and with of course the back office giving the sell-side full control and flexibility of their post trade process.
Another way of redefining the trading stack that we see is to incorporate middle office processing into the back office. If the latter has already been rationalised or restructured, it perhaps makes sense to extend its functionality to cover all aspects of the post-trade lifecycle.
When these designs are coupled with an open EMS strategy sell-sides can implement a powerful yet perfectly flexible and specialised multi-asset solution that isn’t limiting in the future.
However, there is a fine line between delivering specialisation and flexibility, and introducing cumbersome complexity. Problems often occur when in-house systems are introduced in a piecemeal fashion. To avoid them, it is essential to build your structure like a Lego edifice, making sure the parts fit together according to a plan in order to create the whole.
We’d love to hear your feedback on this article. Please click here