Latency management specialist Corvil has introduced its Trade Lens functionality, designed to map the technicalities of latency management onto trading business operations. IntelligentTradingTechnology.com spoke to Corvil CEO Donal Byrne to get the details.
Q: Can you describe what Trade Lens does, and why it’s of benefit to trading operations?
A: The Trade Lens transforms the business visibility of order-flow offered by the Corvil solution. It presents business users with an intuitive and concise view into key metrics such as fill rate, traded volume, cancel to order ratios, order types, latency profile, and session errors. These metrics can be interactively aggregated over sessions, customers, gateways, protocols, venues or any other user-defined grouping. Key features:
– Configurable dashboards provide a business relevant overview of trading performance.
– Interactive tables allow slicing and dicing to investigate issues in detail. Views can be sorted, filtered and saved for future reporting.
– Corvil’s Role Based Access Control allows multiple teams to use the Trade Lens while controlling access to sensitive data.
Q: What kinds of analysis does Trade Lens enable?
A: The Trade Lens presents information that directly relates the business requirements to the performance of the trading infrastructure. The user can quickly identify opportunities or understand problematic behaviours. Examples of the kind of analysis possible with the Trade Lens are:
– Display a dashboard showing DMA Clients ranked by cancel-to-order ratio, and overlay their executed volume.
– Rank all Trading Gateways by their latency, and then see all customers ranked by order-volume on the slowest gateway.
– Quickly see all sessions to a particular venue, and see how well balanced the order-flow is across sessions.
To take a more detailed example: a DMA dashboard might display the latency for the most valuable clients ranked by executed volume. To understand the causes of a large latency, the per-client summary would be broken down to see the latency of each client session and the gateways they connect to. Switching to a per-gateway analysis shows the overall performance of a problem gateway across all clients.
Q: How might a trading firm use a report to understand what infrastructure issues might be impacting trade performance?
A: Network and application metrics can be displayed side by side in the Trade Lens views to identify correlations between business outcomes and infrastructure issues. Notional value filled, latency and message rates can be displayed side by side for all of a firm’s exchange sessions; a poorly performing session might be on a slow gateway or might be getting throttled by the exchange for violating message rate quotas.
Q: Presumably to provide this analysis, Trade Lens needs to understand and decode the trade execution protocols being used. What such protocols are supported?
A: Corvil supports over 150 public protocols, covering trade execution, market-data and middleware across most major markets and geographies. The decoder library is updated and released monthly to our customers, to include exchange-driven changes, additional features and maintenance fixes. Many of our customers are using our Software SDK to create decoders for their proprietary order-flow and market-data protocols without having to disclose any intellectual property or come to us when they extend their applications.
Q: What is the new Advanced Deployment Manager. What does it do and how does it simplify the job of latency management?
A: Advanced Deployment Management is a multi-release strategy to streamline configuration and maintenance of CorvilNet. We understand that some aspects of latency management can be complex. Our approach is to listen to our customers, identifying their concerns and what is taking their time, and then work to automate and assist with those workflows.
In CorvilNet 8.0 we are introducing two specific features. First, “Application Auto-configuration” will allow customers to get automatic configuration updates from our monthly protocol-decoder releases. These updates will draw from our experience across hundreds of deployments and can include complex correlation rules, configuration of metrics such as fill-rate and cancel-to-order ratios, and changes to market-data network addresses. Customers can choose, per-protocol, whether to accept automatic updates; if they choose not to, they can examine the configuration available and apply manually.
Second, a new GUI “Correlation Assistant” will simplify the task of configuring and debugging latency correlation, which can be complex – especially across multi-protocol data transformation. Users can see the application messages at each observation point, examine the correlation rules and get live updates of packet/message/lookup counts. We can’t remove the complexity involved in application transformations, but we are committed to making the configuration task as straightforward as possible.
Q: You also have a new network monitoring appliance – the CNE 2100. Where does this ‘fit’ in your appliance lineup, and what kinds of trading operations is it being pitched at?
A: The new CNE-2100 appliance has been introduced to dramatically lower the cost of entry while offering the full functionality of CorvilNet including order flow and market data analysis. We have clients with 10s and even 100s of remote sites that want to ensure the integrity and robustness of the order flow and correlate this with the latency and loss across the remote site WAN link. Exchanges can also use this product to report latency and loss against service provider SLAs for their remote access points.
Q: What’s your outlook for the marketplace for latency management? It remains pretty specialised, with just a couple of strong vendors. Do you agree? Do you think that will be the case for the next year? Any game changes on the horizon?
A: The latency management market experienced strong growth last year despite the overall flat to declining spend on financial technology. We would agree that the market was reasonably specialised in 2008, 2009 and 2010 where the primary focus was on the reduction of latency for high frequency trading strategies. However we disagree with the statement for 2012. The market requirements started to change in 2011 where we witnessed a broader adoption of low-latency technologies across all electronic trading infrastructures, independent of trading strategy. In addition we witnessed a demand for features that could help people better manage the cost and risk of their trading infrastructure.
Today, Latency Management platforms like CorvilNet are being deployed as the consolidated IT monitoring platform of all trading infrastructure with a focus on speed, cost and risk management. As our customers start to leverage their investment in this low-latency, high performance network, we are now starting to see the need to support non-trading related applications across the enterprise. Therefore we envision latency management as becoming the standard IT applications and network monitoring system for next generation enterprise IT systems where latency and performance assurance will be table stakes.
And there are always game changes on the horizon.