What’s the best way to achieve seamless, end-to-end digitalization?

How much time, money and soul-searching does it take to craft a workable modernization strategy? Especially one that integrates with an existing data strategy and architecture.

The prudent answer is, “That depends; it varies dramatically from enterprise to enterprise.”

However, although the resources might differ, organizations face common challenges when implementing the changes. Over the years, most will have relied on manual workarounds and time-worn operational decisions, which are now incapable of supporting their digital ambitions. Modern applications will most likely have been reconfigured to compensate for inflexible legacy technology.

Impromptu modifications limit the effectiveness of core trading and risk systems, especially when they’re obliged to handle messaging, data management and archiving without a unified suite of solutions.

Why keep it lean?

The pragmatic approach is to use systems and applications for their intended purpose. Keep your core trading and risk management systems lean, so they perform as expected.

And you must decouple integration. Point-to-point integration does the job, but locks you in. Consequently, when you try to update, your systems are so closely coupled that shifting application B directly affects application C. Testing becomes complicated as well. Often, you need to make changes at both ends, and that can make system upgrades a nightmare.

Could you occupy the middle ground?

One of the most significant improvements is to implement some form of middleware, the exact size and shape of which will depend on the organization’s scale and complexity. The solution might be an enterprise service bus (ESB) or an integration platform as a service (IPaaS) if you’re working with on-premise and cloud-hosted apps. A more modern approach to integration enables significantly better management of transformations, filtering and routing compared to the traditional point-to-point approach.

What if business execs could DIT (do it themselves)?

Suppose you need to track the timing of cross-asset activity for specific clients or monitor the frequency of unusual funding requirements over time. Traditionally, this would involve engaging technology teams, prioritizing, defining requirements, then building and testing solutions — a process so lengthy that any business advantage is often lost by the time results are delivered. With real-time capabilities, business stakeholders can access and analyse this information themselves, instantly. The data exists—but how do you unlock it?

Consider implementing a data warehouse for structured data (e.g., reg reporting, inter-system files, end-user reports) and a data lake for unstructured data and underpinning advanced data pipelines. This offers business insights and hands workable capabilities back to your business users.

Then, when business users say, “Right, what’s been happening in this space over the last 6 months across the organization?”, the information is right there. And with the correct tools, it can be self-served.

That way, you allow technology to do things only technology can do and users to resolve issues, avoiding misinterpretations of the “Is this what you want? No, not quite” kind, for themselves.

How do you choose the right tool for the job?

Another worthwhile approach is to create a separate, purpose-built repository for static and reference data. Often, we see multiple systems using the same reference data but with different versions built into various applications. Sometimes, all the data is being mastered and distributed from the trading and risk system, which might not have been designed for this purpose and often ends up holding, distributing and archiving data it’s not even using.

The ideal situation is to have a centralized enterprise data management system (EDM) where your data is captured once and catalogued correctly. With the benefits of data provenance, lineage and a single, unique data version, you can reduce the need for reconciliation and many other challenges caused by running multiple versions of the same data.

Although these moves won’t solve all your headaches, they’ll provide the tools for switching out clunky workarounds and segregating applications so they can do what they’re supposed to do. This will release your flagship trading and risk applications to perform at their peak, while also opening the door to smoother, more predictable change.

What does technology actually do?

Above all, technology is an enabler. In partnership with the business, IT leaders must be empowered to advise, explore new technologies and maintain/upgrade the existing estate, so they’re ready to move at the required pace.

And that’s important. It’s not uncommon for an organization to embark on a well-intentioned, technology-led project with limited business value. Five years later, they’re still toiling away. Everyone’s waiting on “the project” while other requests pile up. For some, riding out the never-ending project is a dream come true. Meanwhile, millions have been spent, but little has been delivered to the business.

In contrast, having short-term objectives and deliverables that can plug into a flexible technology ecosystem adds real value to the business.