THETRADETECH DAILY THE OFFICIAL NEWSPAPER OF TRADETECH 2026
When you look three years ahead, which investments will genuinely move the needle for your trading desk, and which long-standing practices are you prepared to leave behind? Decision grade analytics, not just better or more TCA reporting. Our current roadmap includes the evolution of AI-enhanced TCA, more financial management and dealer behavioural insights alongside real-time views which hopefully will see investment now, bringing it into dealers’ day-to-day activities versus TCA analyst dependencies. The trading workflow needs to be a repeatable decision science, not just a post-trade defence of execution performance and quality for our clients.
We have numerous tools and vendors around our trading workflows, but already we need to further integrate our data and analysis. Our data lineage needs to be easily explainable, trade execution review forums need to evolve from- in some cases-storytelling, to more signal reviews and ensuring that the exception monitoring performed can be further automated with triage that scales with our volumes growth, so we can create capacity without additional headcount.
Review forums should exist to catch signals and deltas, not to explain static charts that buy-side dealers already understand. I’ d like to see less analyst assembled slide packs and retrospective explanations without decision consequences and have more“ what changed and what action we take” framing and exception first agendas over the next three years.
A new era for buy-side trading teams
The TRADE sits down with Brian Mitchell, global head of dealing and implementation at M & G Investments, to discuss how trading desks are shifting from fragmented, retrospective analysis toward integrated, AI-supported, real-time decision-making across the industry.
are emerging. The best traders will be comfortable with data, change while also being curious about new tools and able to evolve as products and regional regulations shift.
On our own operating model, we are looking for additional opportunities to scale, automate elements of our flow further, optimise dealers’ judgement and invest in data-literate, commercially aware, cross product and / or asset class traders.
AI is rapidly entering the execution workflow. Where do you see it delivering meaningful edge for the buy-side, and where is it still falling short of the hype? One of AI’ s more credible contributions to date would be on pattern recognition at scale, less so on autonomous trading. Where some“ edge” potentially comes from detecting outliers across millions of data points that buy-side dealers would not see. Flagging execution drift, venue degradation or broker under performance earlier and supporting the likes of FCA consumer duty aspects, through faster, more contextual oversight rather than simple retrospective reviews.
Other aspects include the automation of judgement light, rules-based workflows like the auto-population and classification of TCA exceptions using in house or third-party data feeds and even dealers ' standard report outputs on markets and moves, for investor consumption. While behavioural analysis of execution decisions versus price prediction can improve consistency rather than chasing the odd bps, we should be strengthening audit trails while AI also starts to support buy-side connectivity cost attribution, broker value assessments and even wallet size logic.
For me, AI increasingly supports our monitoring and reviews- less so the hype of continuous autonomous optimisation of trading abilities, as there are still challenges around data completeness, latency, fragmented OMS /
EMS architecture, complex derivatives and experienced trader judgement still seems to dominate liquidity access, particularly across OTC markets and products.
As desks look to scale trading activity without scaling headcount, how is the role of the execution trader evolving and what skills will define the trader of the future? The buy-side trader’ s role is shifting away from manual order slicing and reactive venue switching towards overseeing increasingly automated and algorithmic flows, interjecting when exceptions or signals merit trading across portfolios and less individual tickets. Multi-asset skills and increased data usage are essential to frame decisions, with the value-add shifting more to experience in complex, illiquid or even event-driven scenario management.
Increasingly buy-side traders must understand how expected-cost models are constructed, be fluent in TCA, able to interpret outputs- not just receive them- and have a broader commercial and broker or venue strategy awareness. To usher this along we need to develop our talent pipelines, barbell experience across the desks and live the evolving market structures and new instruments that
Many firms are reassessing their broker lists and leaning more heavily on bilateral partnerships. What is driving that shift, and what does it mean for the traditional broker model? Many buy-side firms are prioritising brokers that can commit balance sheet and facilitate risk transfer with portfolio trades and blocks consistently across the regions. Broker performance is more observable, comparable and we require it to be persistent over time, so that as the data matures, the more‘ optional’ broker become visible. Long broker lists are not helpful when trying to simplify controls and surveillance and create an oversight burden.
Basic coverage only types of brokers remain under pressure, and the historic model of broad coverage, occasional flow, limited differentiation is being squeezed. Those that are providing little balance sheet, rely on more generic sales coverage and cannot easily demonstrate execution value in TCA data are being increasingly marginalised.
We need brokers and venues to have high quality electronic execution and analytics, risk warehousing and principal liquidity provision at times, alongside strong block facilitation and portfolio trading. The old relationship management model has been replaced by more outcome-driven engagement, lists are shrinking not to reduce competition but to increase accountability, there’ s been a re-pricing of value and you must earn flow continuously!
18 THETRADETECH DAILY