"Firms should consider the points we make in this Market Watch and consider whether changes to their surveillance arrangements are required." - FCA, Market Watch 86.
That challenge goes beyond whether a firm has surveillance technology in place. The more useful question is whether the firm can explain how its controls work, test what they detect and miss, and demonstrate why it believes those controls remain effective.
It is a theme we explored recently on an industry panel discussing Deloitte’s paper on surveillance vendor management. Most firms now rely on third-party technology for at least part of their surveillance, but outsourcing the technology does not outsource the accountability. The firm remains responsible for whether its controls work. That responsibility only means something if the firm can see how the controls it relies on behave: how alerts are generated, which parameters decide what is and isn’t flagged (including defaults nobody has revisited), and how changes to data or software affect what the surveillance detects.
Getting this right is a shared effort. Firms need to understand, challenge and evidence their controls. Vendors should make that easier through clear documentation, transparent configuration and tools that let firms test their surveillance for themselves. Market Watch 86 shows what happens when that understanding is missing.
Our view is that effective surveillance is not defined by alert volume or by the presence of a system. It depends on whether firms can investigate with the right context, calibrate controls to their own risks, test them against real behaviour and give management enough evidence to challenge performance.
Market Watch 86 follows supervisory visits to seven UK retail brokers, where the FCA reviewed the systems, controls and governance arrangements used to identify and report potential market abuse. While the findings relate to retail brokers, the FCA notes that they may also be relevant to other types of firms.
Market Watch 86 is useful not because it gives firms a checklist to follow, but because it exposes six practical tests of surveillance effectiveness.
Start by asking these six questions of your trade surveillance:
1. Are your investigations thorough enough?
Effective investigations bring the relevant context together and apply the correct threshold for reasonable suspicion.
Under UK MAR, firms do not need proof that market abuse has taken place before submitting a Suspicious Transaction and Order Report (STOR). The FCA found that several firms were applying too high a threshold for reasonable suspicion.
In some cases, procedures steered analysts towards reasons to close an alert rather than considering the available information and weighing aggravating and mitigating factors. The FCA observed more consistent decision-making where firms had well-documented investigation and escalation procedures supported by formal analyst training.
Our view: an alert should be the start of a decision process, not the start of a data hunt. Bringing relevant trade, communications and market data into one investigation case helps reviewers assess activity consistently, escalate where needed and retain the evidence that supports STOR preparation.
2. Are your alerts calibrated to your business?
Calibration should be risk-based, explainable and revisited as the business changes.
The FCA identified examples where alert parameters were not sufficiently tailored. This included firms applying flat price-movement thresholds across stocks with different levels of volatility and minimum trade-value parameters that excluded significant amounts of client activity from monitoring.
One-size-fits-all thresholds are a warning sign. Configurable volatility and minimum trade-value thresholds can support calibration, but the control is only defensible if the firm can explain why the setting fits its instruments, trading activity and risk profile. The practical test is simple: why is this threshold set here, and what evidence tells you it is still appropriate?
3. Can you explain and challenge your alert logic?
Third-party technology can supply the control, but the firm still needs to understand and challenge the logic behind it.
There were instances where individuals responsible for surveillance could not adequately explain alert logic, calibration or the circumstances in which alerts should trigger. Some firms appeared to rely on out-of-the-box vendor settings.
Defaults should be a starting point, not a substitute for judgement. Firms need documentation and guidance that make scenario logic and settings explainable, so they can challenge whether the configuration still reflects their business and risks.
4. Are you testing whether your surveillance is effective?
Testing should show not only that alerts fire, but how sensitive the control is to different parameter choices.
The FCA found that testing was often limited. It points to above and below the line testing as good practice, alongside approaches such as using synthetic trade data to assess whether alerts trigger as expected in particular market abuse scenarios. The regulator also emphasises the importance of periodically reviewing alert parameters and logic.
The useful question is not simply how many alerts a system produces. It is what the surveillance would detect, what it might miss and how that changes when parameters move. Native backtesting can help by showing how different settings would have affected historical alerting. That is different from synthetic scenario testing, which the FCA has previously cited as a good-practice example.
5. Can management demonstrate effective oversight?
Management information should help leaders challenge surveillance effectiveness, not just count alerts and cases.
The regulator states examples of senior managers with limited understanding of alert triage, investigations, escalation processes and calibration. It also highlighted the role of quality assurance in identifying inconsistent decisions, errors and areas where further training or guidance may be needed.
Useful management information connects volumes to decisions: which scenarios are driving alerts, where calibration has changed, where quality assurance is finding inconsistency and what remediation is underway. Reporting should make those questions easier to answer. SteelEye can support this by surfacing surveillance information for oversight and review.
6. Is your MARA driving surveillance improvements?
A market abuse risk assessment should be an input to surveillance design and review, not a document that sits alongside it.
In its investigations, the FCA found that firms gained the greatest value from their MARAs when they used them to identify gaps in surveillance coverage, assess whether controls were effective and prioritise improvements. It specifically cautioned against assuming that having an alert in place is, by itself, evidence of an effective control.
Used properly, the MARA becomes a practical bridge between risk assessment and controls. It should inform discussions about coverage, calibration and priorities, and expose where a firm's surveillance capabilities or guidance need to improve. That is why MARA conversations should be part of implementation and ongoing review.
The real test: can you explain and evidence effectiveness?
The takeaway is simple: surveillance should not just be running. Firms should be able to demonstrate that it is working.
That means being able to explain why controls are configured as they are, test what they identify and what they might miss, investigate with the right context and give management the evidence to challenge whether arrangements remain effective.
SteelEye can support those processes by bringing relevant trade, order, communications and market data into investigations, providing configurable alert parameters, native backtesting, case management and insights reporting. Those capabilities can support investigation, calibration, testing and oversight, while responsibility for the firm's risk assessment, governance and final decisions remains with the firm.
If some of these questions are difficult to answer, we can help you strengthen the surveillance processes behind them.
FAQS
What should firms review following FCA Market Watch 86?
Firms should consider whether investigations have sufficient context, alert thresholds reflect their risk profile, surveillance logic is understood and challenged, controls are tested for effectiveness, and management has enough information to oversee performance. The MARA should also inform coverage, calibration and priorities.
How can firms demonstrate that trade surveillance is effective?
Firms need evidence that controls are designed for their risks and work as intended. That includes explainable calibration, testing of what controls detect and miss, documented investigation decisions, quality assurance and management information, and a clear link between the MARA and surveillance priorities.
SteelEye can support this by bringing relevant trade, order, communications and market data into investigations, providing configurable alert parameters, native backtesting, case management and Insights reporting. These capabilities can help firms investigate with greater context, test calibration and provide management with clearer evidence of how surveillance controls are performing.
Is automated trade surveillance enough?
Automated surveillance can support a firm's controls, but technology alone does not demonstrate effectiveness. Firms still need to understand, calibrate, test, investigate and oversee their surveillance arrangements, with final decisions and accountability remaining with the firm.
SOURCES
.png)
.png)
.jpg)
.png)
