ITRS Launches FLM Data Feed Latency Comparison System

 In a move that could establish an enabling technology for those interested in latency arbitrage, ITRS Group has released Feed Latency Monitor (FLM), a plug-in to its Geneos platform that lets users analyse relative latencies of real-time data feeds in real time.

According to ITRS CTO Misha Kipnis, early adopters of the system are using it to compare delivery latency in the data feeds they receive through their multiple vendors. Tantalisingly, FLM generates a set of relative latency metrics that can be analysed in Geneos and output to algorithmic trading engines.

This can ensure, for example, they are receiving data with the lowest latency throughout the trading day, when different vendor systems may undergo varying degrees of stress that can impact latency.

To date, trading firms wishing to do this kind of analysis have used Excel to check the data on a one-off basis, after the instance. FLM brings real-time capability to the application, meaning that firms can respond as latency performance fluctuates over time.

For now, FLM is available for five data delivery channels, among them: Thomson Reuters Integrated Data Network (IDN) and Reuters Data Feed Direct (RDFD), via the Reuters Foundation API (RFI), giving access to the Reuters Source/Sink Library (SSL); NYSE Technologies' (former Wombat) data feedhandlers and Middleware Agnostic Messaging API (MOMA), supporting connections to legacy Wombat TCP middleware, Informatica's LatencyBusters Messaging (LBM; originally 29West) and Tibco Software's Rendezvous; SunGard Global Trading (formerly GL Trade); and Trading Technologies. Kipnis says ITRS is in discussions to add more suppliers to its roster.

FLM works by allowing users to analyse and compare the latency performance of any given data feed across multiple suppliers’ delivery platforms. By directly subscribing to the monitored platform’s incoming market data feed, FLM can calculate the relative latency of this feed compared to a reference market data feed. It also identifies the relative latency between both direct and consolidated data feeds.

For example, a firm using Nasdaq OMX real-time market data may wish to compare latency across Thomson Reuters’ consolidated IDN network and RDFD, Thomson Reuters’ direct data feed delivery infrastructure. FLM ‘subscribes’ to a representative set of instruments through both platforms, and measures the delivery latency of each. These statistics are then fed into Geneos, where they can be analysed according to a variety of user criteria.

According to Kipnis, FLM “analyses market data distribution performance on a local and wide-area basis enabling operations teams to redirect market data sourcing away from under-performing feeds. The analysis is based on real-time performance as measured against a reference feed; FLM quickly identifies those vital external price feeds which are running slowly and allows data sourcing to be redirected before trading processes are impacted.”

The resulting analysis can be fed into applications, such as algorithmic trading engines, or used to drive graphical dashboards that can be made available to traders or to administrators and support personnel. In the case of the former, the metrics can help the application make a determination of which of the data sources of data available to it to use as their delivery performance fluctuates during the trading day. In the case of the latter, traders can use the dashboards to help decide which vendor platforms to use at any given time; administrators can be warned of potential latency issues as they arise.

Kipnis says FLM was developed as a result of client demand. ITRS’s Geneos is used as a network and trading equipment monitoring system, and FLM is an extension to its existing Market Data Managing and Monitoring Suite.

The system runs on the Geneos server at the client site, and makes use of the standard Geneos client application to run the visualisation dashboard. The system went into production in the fourth quarter of 2010. Kipnis reports significant interest from a full range of financial institutions, from smaller brokers and managers to major sell-side banks.