Matt Barrett, CEO of Adaptive, shares best practices for deploying next-generation trading technology, reflecting key insights from Markets Media’s latest industry research.

Research conducted by Markets Media has revealed a highly varied level of technology deployment in place today, varying by company type and function.

Matt Barrett, CEO of Adaptive, reviews the results and outlines the current state of the capital markets.

How much pressure are capital markets firms under to transform tech provision models?

Every capital markets participant clearly recognises the need of a technology refresh, from very large banks to smaller venues and exchanges. There is now broad acceptance that this transformation is inevitable and will occur during their tenure, driven by volatile regulatory and geopolitical factors.

In the equities space, increased retail participation has changed investment flows and boosted volumes. The sector faces a constant discussion about 24/7 trading. Separately, the industry also faces vocal concerns about the risks associated with AI technology and investments. The question coming from the US West Coast is, how are we actually going to make money out of AI? It is an industry primarily fuelled by debt, and so risk sits in equity and fixed income private markets.

Are we seeing technology-led leaders break away from the pack?

Yes. One major US bank, for example, has notably re-invested its profits into technology, successfully building a huge internal technology cost base and wielding it with great success. Moving forward, a key challenge will be to manage the flexibility of that cost base in the future.

Within the industry, CEOs traditionally come through from sales or risk desks, they’re not technologists. In a few places, notably smaller firms, we are seeing CTOs being promoted to CEOs. For example, many exchanges are really technology-driven businesses, but the sales team often sees technology as complex and prefers not to manage it, so future CEOs need to have tech backgrounds to optimise their business.

Who’s calling the shots in the decision making process around technology deployment today?

Decision-making related to technology often faces significant inertia driven by resistance. Tier one sell-side institutions are facing CxO-level demands to standardise trading technology because they cannot afford the ongoing operational expenditure (Opex) to run disparate systems across multiple business lines. Within those firms, leaders view multi-asset, modern trading technology infrastructure as deployable across multiple divisions and want to assess it. Friction often occurs where technology organisations hold a different philosophical approach and consequently dismiss alternative solutions as unsuitable or not a good fit. Until those organisations resolve this resistance, meaningful change will be difficult.

Organisations see this inertia most recently with cloud. Before I started Adaptive, I worked at a tier-one bank and it talked proudly about its ability to build and run proprietary data centres better than its competitors. Today, no one would think that that is something to be proud of, given the efficiency and ubiquity of the cloud. It’s the same thing with banks specialising in non-differentiating technology, a historical source of pride that now represents a strategic lag.

How can that change?

You need to appoint people who are very comfortable with innovation and change sitting in senior roles. People who are prepared to take very difficult decisions and say, ‘We need to cut £100 million of spend, and I assure you in two years’ time, the whole thing isn’t going to fall apart.’ They need to be that bold.

The locus of technology control is tightest around a trading team and relaxes the further you move away from the desk. Why?

Technology that is controlled by the trading desk tends to be systems that become part of the trading workflow, and therefore directly affect the cost of sale and front-office performance, rather than simply being categorised as Opex.

How do you see deployment models impacting performance and cost management?

The industry survey found – as we also see – that buy-side firms typically prefer to buy or take a hybrid approach rather than build technology in-house, whereas the sell-side develops more technology internally.

Yet the opportunity not discussed, which would really enhance performance, sits in the gap that exists in understanding data centre infrastructure as a place to meet your counterparties. There is a persistent front-office attitude that technology used for counterparty connection must be hosted on-premise. I believe that cloud providers or market participants have not demonstrated a good enough, or better, replica of a co-location model in the cloud, which would allow market participants to connect and trade with each other in a cost-effective and latency-sensitive way. Cloud providers are working to achieve this, and we are collaborating with them to deliver it.

Today, when someone wants to build a new venue, for example, their default reaction is to build and deploy it into one of the existing data centres, because all their counterparts are there. The reality is that it’s far quicker to connect to a venue in the cloud, but people still have ‘muscle memory’ for deploying on-premise.

How does the wide acceptance of componentisation change capital markets firms’ capabilities?

It is a great risk and a great opportunity. If you get componentisation right, you are going to maintain pricing power over your vendors. In a healthy vendor ecosystem, a specific component provides a specific replaceable capability, allowing the firm to control technology costs during vendor renewal negotiations.

If you get it wrong and become dependent on capabilities that are not easily replaceable, then you’ve created a very complex system and lost pricing power by not being able to renegotiate at contract renewal time.

The work being done by FINOS, the open source foundation with its Common Cloud Controls initiative, is trying to create pricing power with the cloud providers. Currently, cloud providers offer services using APIs that are often not like-for-like, creating user dependency. This dependency prevents firms from effectively renegotiating. They need to create a ‘shim’ across the top of services, a thin wrapper over an external API, component or capability that abstracts it, and allows the user to migrate from it if they want.

That is a good snapshot of lagging best practice.

Does legacy technology always underperform?

If you are forced to deal with legacy technology and have a successful business, you’ve likely been around long enough for something to become a legacy system. Legacy is a recognition of earlier success. New firms entering the market are more aggressive in their adoption of modern technologies and are further along the adoption curve, as they don’t have the same technical debt. The survey is a very good snapshot of where we see many of the established firms.

What would you recommend for firms to optimise their vendor engagement?

I was pleased to see a deeper understanding of the increased return on investment that technology changes delivered over the last five to 10 years. That tells us we’re pushing against a slightly open door. For capital market firms to truly benefit from new deployment models, it is necessary to take a fresh approach, examine technology that can provide a layer of abstraction, and utilize componentization language, which breaks down your system architecture into components.

Someone with a high-level view across the organisation, down into its organisational structure, can map that onto their technology structure to understand the right places to componentise their organisation. This allows them to switch from an internal, high-cost base to an external, far lower-cost base, thereby gaining organisational efficiency and differentiation.