Neutralizing Infrastructure Performance Anomalies in Electronic Trading

Network performance anomalies, like spikes in latency, microbursts of a large number of messages, or out of sequence network packets, are an issue for every business. In electronic trading environments, however, they take on more significance because they can have a direct impact on the bottom line, so you need to get them fixed, and fixed quickly.

The challenge with high-performance trading is that rapid change is common when interacting with markets – it’s not unusual for unusual things to happen – so the difficulty is in separating meaning from noise, real anomalies from explainable patterns of behavior. When there are so many metrics to digest, you need smart and automated monitoring and analysis so that you can identify genuine performance issues, why they occurred, and how to remedy.

Corvil’s Intelligence Hub solution is purpose-built for precisely this kind of monitoring, with detection techniques and analytics algorithms that are able to alert you when something genuinely meaningful is occurring. Here are five factors beyond the scope of most monitoring tools that are ‘all in a day’s work’ for Intelligence Hub:

  1. Context: We look at orders rejected, canceled, or not filled; crucially we allow customers to see them side by side with infrastructure performance metrics. We provide context, so if there is a correlation between infrastructure performance and fill rate performance, our metrics will expose it.

  2. Impact: The only way to fully understand the impact of an anomaly and the scale of a problem is by being able to examine client, market, and even symbol data. Intelligence Hub excels here too and will dashboard, for example, which individual security trades are being impacted by infrastructure performance.

  3. Patterns: It’s important not to confuse anomalous activity with recurring patterns, such as spikes when markets open and close around the world. Intelligence Hub will recognize recurrent activity and can be configured to only report on deviations from the patterns.

  4. Trends: The ability to follow trends over time is also important. Behaviors exhibited over long-term trends (message-rate growth, for example) are relevant, but short-term intraday trends can be important too. Are you on track to keep order-to-trade ratios within critical limits at the end of the current trading day? Intelligence Hub includes predictive analytics to answer this type of question before limits are breached.

  5. Correlation: With a wide range of metrics analyzed continuously on a per-client, per-system and per-market basis, it’s critical to draw links between related anomalous events occurring at the same time in different data series. Sophisticated alert correlation and coalescence functions built into Intelligence Hub perform this task automatically, generating warnings that are more meaningful and more actionable.

By combining these five capabilities in a platform that also supports large-scale data collection and exploration, Intelligence Hub enables operations and analysis staff to isolate and explain problems faster, and to spot critical factors that impact trading results - sometimes even before they happen.

Schedule a demo to find out more about Corvil anomaly detection.

Learn more about Corvil Intelligence Hub

Fergal Toomey

Fergal Toomey, Chief Scientist
Pico is a leading provider of technology services for the financial markets community. Pico provides a best-in-class portfolio of innovative, transparent, low-latency markets solutions coupled with an agile and expert service delivery model. Instant access to financial markets is provided via PicoNet™, a globally comprehensive network platform instrumented natively with Corvil analytics and telemetry. Clients choose Pico when they want the freedom to move fast and create an operational edge in the fast-paced world of financial markets.