By Gustaf Gräns, Chief Product Officer and Mikael Ohlsson, Head of Product, Scila AB
The scaling challenge in modern compliance
In modern financial markets, volatility and liquidity can shift in milliseconds. Yet, the surveillance rules designed to monitor these markets often rely on rigid, static thresholds. This disconnect creates a heavy burden on the analysts: as markets move dynamically, compliance teams are forced to manually retune parameters across thousands of instruments, leaving highly skilled analysts managing parameters instead of hunting actual market abuse.
Scila Surveillance addresses this challenge by adding tools to complement rules-based surveillance. By introducing intelligent, self-adjusting parameters, Scila Surveillance allows compliance teams to deploy alert rules that adapt to market conditions. The goal of this technology is not to replace the analyst’s judgment with an opaque automated system, but to provide a scalable, proactive framework that reduces the manual maintenance burden.
The limits of static scaling
At times, static limits can struggle to cope with the market’s natural ebb and flow. A primary driver of this challenge is shifting volatility regimes, which creates a significant operational liability where rules perfectly tuned for a quiet period become sources of excessive noise during a volatile period.
The volatility trap
During periods of high volatility, fixed thresholds can generate a flood of false positives, forcing analysts to sift through noise rather than investigating genuine risks. Conversely, during periods of low volatility, these same loose thresholds fail to detect subtle manipulation, leaving the firm exposed to regulatory risk.
The maintenance burden
The operational cost of manually retuning rules can be high. Compliance teams often manage thousands of instruments, ranging from highly liquid stocks to illiquid bonds. Treating a major index future the same as an illiquid commodity contract inevitably leads to poor surveillance quality. Attempting to manually calibrate parameters for each instrument type consumes valuable resources that should be dedicated to analysis.
Modular and expert-driven adaptability
To overcome the limitations of a ”one-size-fits-all” approach, Scila Surveillance solves the static threshold problem by anchoring surveillance in parameters calibrated by statistics rather than nominal values. This is an inherently transparent, modular architecture. The compliance officer still defines the risk profile while the platform provides the dynamic building blocks that allow those rules to scale mathematically alongside market conditions.
Z-score-based dynamic thresholds
To automatically distinguish between normal market noise and statistically significant outliers, Scila Surveillance incorporates Z-Score statistical benchmarking into its core price movement alert rules.
Statistical significance over nominal value
The system shifts the focus from absolute values (e.g., Price change > €5) to statistical deviations (e.g., Price change > 3 standard deviations). By utilizing Z-scores and log returns, the system normalizes behavior across different asset classes. A 3-sigma move is statistically significant whether the asset is priced at €10 or €1000, ensuring consistent surveillance coverage.
Hybrid configuration for tailored safety
For maximum flexibility, these rules now support a combination of limit types. Compliance teams can configure rules to require Ticks and Percentage thresholds alongside Z-Scores, creating a multi-layered safety net that respects both nominal market structure and statistical reality.
Scenario: Ramping in different market conditions
Quiet market: A trader attempts to aggressively ramp a price. Because volatility is low, the dynamic thresholds have tightened, and the system detects the anomaly immediately.
Turbulent market: The market reacts to a macroeconomic event with sharp price swings. The thresholds widen automatically, ignoring the normal volatility and preventing a flood of false positives.
Dynamic order book grouping
While surveillance granularity is essential, configuration must remain manageable. It is not enough for the rules to adapt, the instruments themselves must be monitored according to their current behavior.
Instead of manually assigning rules to broad static instrument groupings, Scila Surveillance utilizes automated segmentation to group them based on statistical behavior. To further highlight the modularity of the platform, these groups can be configured broadly or highly specifically. For instance, users can define logic such as ”Top 20% by Volatility” across all asset classes, or narrow it down to ”Top 5% Equity Options by Volatility”. This ensures that an instrument is always monitored according to its current behavior, not how it traded six months ago. For sophisticated risk grouping, the platform also allows for weighted scoring (e.g., Volume + Volatility + Trade Count).
Scenario: Automated liquidity buckets
A compliance team needs to apply strict spoofing rules to illiquid energy contracts. Instead of manually defining parameters per asset class or instrument type, they utilize automated classification. The system identifies instruments with low liquidity behavior and automatically places them into the ”Illiquid” bucket, applying the stricter rules without manual intervention.
Monitor. Identify. Act.
By automating the classification of instru- ments and the adjustment of thresholds, Scila Surveillance drastically reduces the time required for parameter maintenance.
This augmented approach empowers the human expert. A lower false-positive rate means analysts spend their time investigating genuine market abuse risks rather than dismissing noise. Ultimately, utilizing dynamic parameters demonstrates to regulators that the firm possesses a surveillance function that actively evolves with the market, rather than one that simply reacts to it
Techie insights 2026: Dynamic alert
rule parameters in adaptive surveillance (PDF)
About the authors
Gustaf Gräns is the Chief Product Officer at Scila, where he leads the firm’s overarching product strategy and vision. With a career at Scila spanning nearly two decades, Gustaf has a proven track record of bridging deep technical architecture with commercial success. He brings extensive experience in driving complex product development and executing high-throughput system integrations for tier-1 clients such as Deutsche Börse, Erste Group Bank, and Glencore. Gustaf holds a Master’s degree in Electrical Engineering, alongside studies in Economics, from Uppsala University.
Mikael Ohlsson is the Head of Product at Scila and holds a Master’s degree in Computer and Information Engineering from Uppsala University. With over a decade of experience at the firm, Mikael combines strategic product leadership with a hands-on background in software engineering. He focuses on developing robust, multi-asset solutions that address the technical and regulatory complexities of modern market infrastructure, including practical AI implementation. Specializing in advanced algorithms to prevent complex market abuse, Mikael operates at the intersection of technical execution and client strategy, successfully leading complex integrations for varied market structures such as MTFs and interdealer brokers.
If you would like to discuss Scila’s roadmap or explore how these technologies can be integrated into your surveillance operations, please get in touch via LinkedIn or contact our team at sales@scila.se.
