Deep capital markets expertise, the proven xMetrics® platform, and the responsiveness of a specialist engineering team — focused on solving your specific challenges, not selling you workarounds.
Global bank operating an e-FX Distribution desk serving over 850 institutional clients – asset managers, hedge funds, corporates, and regional banks. The firm handles $85B in daily notional volume across 2.4M quotes and 180K trades, distributing pricing via direct API, its single-dealer platform, multi-dealer platforms (360T, FXall, Refinitiv) and ECNs.
Three of the firm’s largest clients – collectively representing $47M in annual revenue – raised serious concerns about execution quality. Each had independently conducted TCA analysis showing the LP underperforming arrival price benchmarks with elevated reject rates. One had already shifted 60% of flow to competitors; another threatened complete termination.
The LP’s internal analysis told a different story: its pricing engine was consistently top-3 on multi-dealer platforms and fill rates were strong. However, neither side had complete data. The LP had no visibility into client-side latency or aggregator processing; the clients’ TCA measured outcomes without understanding causes– particularly around latency attribution, reject root cause, and true execution quality.
Beyond the immediate revenue at risk, the LP faced reputational damage, regulatory scrutiny under the FX Global Code, and rivals positioning themselves as the more transparent alternative.
xMetrics was deployed to provide microsecond-accurate visibility across the entire quote-to-fill lifecycle. Hardware-timestamped capture points at every boundary – network ingress, pricing engine, quote distribution, order receipt, fill decision, execution report, and market data feeds with full latency decomposition from client to LP and back.
Custom UDFs (user-defined functions) built using the xMetrics Mutations Framework added analytics not available from raw capture: quote freshness (measuring quote age at order receipt to identify stale-price hits), reject attribution, toxicity scoring and independent TCA.
Within four weeks, xMetrics revealed a picture fundamentally different from the client’s TCA reports:
The LP’s median processing time was sub-1ms, with over 87% of round-trip latency attributable to the client’s own order path.
77% were caused by clients hitting stale quotes, with genuine last-look protection accounting for just 4%.
73% of trades flagged as poor execution in client TCA had been filled at or better than true market price with aggregator latency misattributed to the LP.
Sub-1ms median processing, 96.2% fill rate on valid orders, and a 0.5% last-look reject rate.
| Client | Outcome |
|---|---|
| Asset manager | 100% flow restored; client-side latency reduced ~76%; became a reference client |
| Hedge fund | Termination averted; reject rate dropped ~68%; flow increased 40% |
| Corporate treasury | Relationship preserved; began routing time-sensitive trades direct to LP |
| Client | Outcome |
|---|---|
| Asset manager | 100% flow restored; client-side latency reduced ~76%; became a reference client |
| Hedge fund | Termination averted; reject rate dropped ~68%; flow increased 40% |
| Corporate treasury | Relationship preserved; began routing time-sensitive trades direct to LP |
The LP faced the loss of $47M in annual revenue when three major clients cited poor execution quality. xMetrics provided microsecond-accurate capture across the full trade lifecycle, revealing that performance issues lay in client-side latency and aggregator routing – not LP execution. With verified data, the LP defended all three relationships, recovered a departing client, and generated $12M in new revenue by turning transparency into a competitive differentiator.
To view a high level online demo of xMetrics please complete this form. We will then provide an access code to view the content.
Address: 20-22 Wenlock Rd, London N1 7GU
Telephone: 020 7030 3791
Email: info@instrumentix.co.uk