THETRADETECH DAILY THE OFFICIAL NEWSPAPER OF TRADETECH 2026
For years, innovation on the buy-side has been tied to large-scale, high-cost projects, but are smaller, targeted improvements perhaps more viable and impactful today? Traditional providers of OMS and EMS applications deliver a value that has been risky to build yourself and expensive to change. As newer technology becomes available, these providers provide increasingly better ways to connect themselves to the modern trading desks. This includes end-to-end automation( no touch trading) and workflow automation( create and place a basket of orders based on a set of factors).
When you combine this with more efficient developers using AI-assisted coding, we see a multiplier effect in terms of bridging the gap between manual trading workflow and code-behind workflow.
The moat on data and technology is shrinking while the moat on talent is increasing, fewer people can do more. It is easier and cheaper to build out better infrastructure now than before.
When reimagining the trading desk’ s core design, which legacy assumptions or workflows are most in need of being challenged or discarded? The biggest assumption to discard is that trading performance is individual. We ' ve moved fully to group-based outcomes. This includes group-based performance, crossteam collaboration on larger trade programs and challenging each other on strategy or tactics.
We found it useful to group the desk into natural focus areas. A classic equity trading desk and a quantitative trading desk. This split was not done to separate tools, flow or people. It enabled us to concentrate skill sets and extract maximum value from our technology partners, algorithmic providers, high-touch trading desks, liquidity sourcing and overall global relationship strategy.
Trading is moving from classic workflow of moving an order from system A to system B and routing a list to algorithm X at broker Y. This is now being systematised into workflow tools with as few clicks as possible. Your traders will never look back.
We also follow our chief executive’ s core strategy on not only adopting AI, but innovating with AI. This includes actively prototyping agents that assist traders in real-time with information and decisionsupport. These agents will act on events, perform scheduled tasks and can be interacted with by our traders. We use AIassisted coding to prototype tools and our developers use it to build out functionality in our asset management suite more effectively.
How can firms strike the right balance between reducing reliance on expensive vendor ecosystems and still accessing the
The changing tide of AI
The TRADE catches up with Peder Viervoll, head of quantitative trading at Norges Bank Investment Management, to explore how AI, modern infrastructure, and workflow automation are enabling buyside trading desks to shift from costly, large-scale transformations to faster, talent-driven, and highly collaborative incremental improvements that enhance efficiency and decision-making.
liquidity, data, and tools needed to stay competitive? For a smaller to medium sized asset manager, a classic vendor might solve most of the firm’ s problems. The larger the firm, the more likely it is that they can allocate money and resources on moving towards composing something using both vendor technology, albeit legacy, and modern APIs to create more seamless workflows. When it comes to data, I would recommend moving to modern and scalable database technologies in the cloud. These are not necessarily expensive and can be scaled nicely with transparent costs. Working with data vendors, we try to connect them as much as possible through our strategic database technology, reducing footprint and need for technical maintenance and support internally. This allows for quick access to new data sets and faster time-to-market for the end-user.
For tools, prototyping with AI-assisted development is the lowest hanging fruit and we believe we are among the most active and forward leaning users in our industry. My view here is that we experiment and don’ t lock in on specific technology. It changes monthly. Build fast. Fail fast. Iterate relentlessly.
Looking ahead, what does a truly adaptable, data-driven buy-side trading desk look like, and what practical steps can firms take now to start moving in that direction? No one has a crystal ball that show us where we are in the next five to 10 years. First advice; get the data to the users as fast as possible. Make data actionable. Build better workflows. Live with legacy technological debt- but put the systems on a diet. Convert and move towards modern APIs. Second advice; right now, experimentation is cheap. Use it to lift the level of skill across the desk. In one way or another, AI will be here. The real question isn ' t whether AI replaces traders, it ' s how much you can augment them. That gap is widening quickly.
20 THETRADETECH DAILY