Mind the gap: Bringing market context and execution closer together

• Separation in trading workflows: Traders often analyze market data and then switch to different systems to execute trades, a gap rooted in the historical development of distinct technologies for market data and execution. 
• Different system purposes: Market data desktops focus on information and analysis, while execution systems handle order management and connectivity, making integration complex due to differing business areas and development histories. 
• AI’s impact on trading: AI enhances trading by compressing and filtering vast amounts of market information quickly, shifting the trader’s role towards judgment and increasing the cost of switching between systems. 
• LSEG’s integrated solution: LSEG uniquely combines a broad data foundation, sophisticated desktop workflows, execution capabilities, broker connectivity, and AI into a connected system aimed at uniting decision-making and order execution.

There are moments on many trading desks that rarely appear in a workflow diagram.

They happen after the analysis is done. A trader has read the news, checked the charts, looked at where the volume is forming, seen which brokers are advertising interest. The decision is effectively made. And then the trader moves to a different system to act on it.

Of course, plenty of desks have closed this gap. But for those that haven’t, the separation is so routine that it’s stopped registering as a choice at all, and it persists for reasons that have more to do with how the technology developed than with how trading is done.

Market data desktops and execution systems evolved along different tracks, serving different buyers, solving different problems. Desktops were built for information: coverage, depth, history, the ability to understand what’s happening and why. Execution systems were built for action: order handling, routing, broker connectivity, control over how a trade reaches the market.

Both matured into sophisticated products, but few firms have made them work together as one thing. Owning both ends of the workflow requires a data foundation broad enough for genuine research, an execution capability with real operational credibility, and the connectivity to reach the market. Those are three separate areas of the business and assembling them takes either a very long time or a series of decisions that only make sense in hindsight.

For a long time, this was tolerable because the quantity of context bearing on a trade was manageable. Filings, estimates, broker research, volume patterns, block activity, options positioning, macro releases and liquidity signals from a hundred brokers at once – none of that is new individually. What is new is the expectation that a trader will have synthesised all of it before the opportunity moves.

This is precisely where AI is beginning to make a practical difference. Not by making trading decisions, but by compressing the research layer: filtering a large volume of material down to what is relevant, summarising it and surfacing the thing that matters. The funnel narrows faster than it used to.

This sharpens the problem rather than solving it. If the analytical stage compresses from an hour to a few minutes, the manual handover to a separate execution system stops being a minor administrative step and becomes a visible share of the time between insight and order.

There is a second effect, less obvious than the first. As AI takes on more of the filtering, the trader’s remaining contribution concentrates into judgement: deciding what the synthesis is worth and what to do about it. That is the part of the job carrying the value, and it sits immediately either side of the handover. Breaking off to move between systems is a more expensive interruption than it used to be, because there is less routine work left around it to absorb the disruption.

Very few organisations hold all of the pieces required to address this, and fewer still have spent decades building each of them. LSEG is in the unusual position of already owning every one, built separately for their own markets over a long period. The work of the last few years has been connecting them.

A data foundation covering roughly 80 million instruments, deep enough to make a desktop a genuine analytical environment rather than a quote screen. A desktop workflow built out over many years of client-driven development. Execution capability, in the form of LSEG TORA engineering and operational know-how, now folded into REDI on Workspace. A connectivity layer reaching around 1,000 brokers through LSEG Autex Trade Route. And an AI layer beginning to sit across all of it.

What they all add up to is convergence on a single point: the moment a decision becomes an order.

A trader tracking a set of securities through Monitor or Pulse is watching live movement as it happens. Conventionally, spotting something actionable there meant turning to another system to do anything about it. In desktop environments such as REDI on Workspace, that observation moves directly into order entry and execution monitoring.

The blended order book shows available depth; intraday volume analytics show how much of the day’s activity has already happened. That context is not just informational; it shapes urgency, timing and routing. Where the two sit apart, the judgement it produces is carried across in the trader’s head. Where they sit together, it is applied directly to the order.

Autex IOIs show where broker interest sits, and block trade monitoring reveals where large-in-scale activity is occurring. Both are tools for finding liquidity and, in the same environment as the execution layer, finding and reaching that liquidity become part of the same action.

Options context shows the cost of separation more sharply than most. A trader assessing options activity, pricing and the volatility surface is usually working towards a hedge or a structure with several legs, where the relationship between them is the whole point. Rebuilding that view in another system is where errors enter, and where the market moves while the trader is retyping.

Macro context is the mirror image, arriving from the top down. A trader reviewing economic indicators or a scheduled release is forming a view about rates, a currency or a sector rather than a single name. That view has to be translated into specific positions, and the translation is where intent degrades.

None of these describes a dramatic failure, which is why no single one justifies rebuilding a workflow. Each is a small transfer of intent from one screen to another, performed correctly, dozens of times a day, by people who have stopped noticing they are doing it. The cost is invisible because it is distributed, surfacing only in the aggregate: in the trade that moved while the order was being retyped, in the hedge that went on late, in the position that reflects the decision approximately rather than exactly.