For institutional buy-side desks, the gap between low-touch and high-touch FX execution is both an operational inconvenience and a measurable cost. Fragmented workflows mean manual reconciliation, delayed or incomplete transaction cost analysis, and execution decisions made with less information than the situation demands. This article examines why that gap exists, what it costs in practice, and what a more integrated approach to FX execution looks like.
Two Workflows, One Desk
Most institutional FX desks today operate a split reality. Electronic, low-touch orders flow through an execution management system cleanly and efficiently. High-touch trades such as larger orders, complex structures, and relationship-driven executions are handled separately, often through chat, voice, or standalone messaging platforms. Fills are typically tracked manually and reconstructed later in spreadsheets.
That division made sense when it emerged. High-touch trades require human judgment, bilateral negotiation, and counterparty relationships that no algorithm can fully replicate. The problem is everything that happens around it.
When a trader works a large resting order throughout the day, fills arrive incrementally across separate communication channels. Each update requires a manual entry. Each manual entry introduces a lag and a margin for error. By the time post-trade TCA is run, the data picture is already compromised because of how it was recorded.
The Cost of Disconnection
The direct costs include time spent on manual reconciliation, errors in fill tracking, and TCA that reflects incomplete data. The indirect costs are harder to see but often more significant.
When execution data is fragmented, the feedback loop that should inform future trading decisions breaks down. Pre-trade analytics lose value if they cannot be compared against accurate post-trade outcomes. Traders make judgments on execution style, liquidity provider selection, and algo deployment without the full picture. Over time, this persistently chips away at execution quality.
There is also a compliance and oversight dimension. As regulatory scrutiny on best execution continues to develop, the ability to demonstrate a consistent, documented decision-making process matters. A workflow that lives partly in a spreadsheet and partly in a chat window does not support that audit trail.
The Case for Convergence (With Caveats)
A logical response is to bring high-touch and low-touch workflows onto a common platform, but it requires clarity about what integration does and does not mean.
Integration does not mean automating away trader judgment. The negotiation, relationship management, and discretion in high-touch execution remain human activities. Integration addresses the data layer around those activities: capturing pricing, fills, execution progress, and order status digitally within the same system that handles electronic orders, rather than across disconnected tools.
That gives traders a single view of order and execution progress, removes the need for manual fill tracking, and ensures TCA covers the full scope of desk activity. It also makes pre-trade and post-trade analytics more meaningful, because the data feeding them is complete.
However, poorly implemented integration creates its own complexity: a platform that technically connects both workflow types but does not reflect how traders actually work, leading to workarounds and shadow processes that replicate the original problem.
The standard-setting question to ask for the integrated approach is: Does it reduce friction for the trader without removing the judgment the role requires?
Best Execution in a Fragmented Market
Workflow integration is one part of a larger execution-quality conversation. The FX market has become more fragmented, with a growing number of alternative liquidity venues offering midpoint execution alongside traditional bank liquidity pools.
From a spot FX perspective, execution quality is the primary source of risk. Venues that consistently (key word) provide midpoint execution can materially reduce transaction costs and are often considered to represent best execution. Access to midpoint liquidity is valuable. Assuming that access equals best execution but without rigorous measurement is a misstep.
Firms need to verify whether fills are achieved at the midpoint and assess the consistency of that performance over time. Midpoint execution that holds in normal conditions but degrades in volatile periods tells a different story than one that is stable across market environments. Pre-trade TCA—combining historical data, liquidity provider analysis, and execution style modelling—is what makes that verification possible. Without it, best execution is a claim rather than a conclusion.
Where This Points
The direction of travel for institutional FX execution is toward greater integration of execution and analytics. Not as a single product category, but as a connected practice with traders and portfolio managers working from the same data, applying pre-trade analysis to execution decisions, and capturing post-trade outcomes to inform the next trade.
That requires:
- Platforms capable of handling both workflow types without forcing a choice between them.
- Data discipline around high-touch execution that matches what electronic trading already delivers.
- A measurement culture that treats best execution as something to be demonstrated, not assumed.
Buy-side desks are being asked to manage more complexity with either static or shrinking resources. Desks that navigate it well will be the ones that treat their execution infrastructure as a source of analytical advantage, not just a mechanism for getting orders done.
