Skip navigation

Market Data Distribution

Market data is the lifeblood of all firms in the financial services industry. Historical and real-time information on asset classes such as equities, options, fixed income and commodities are critical for buy-side and sell-side firms in trading as well as modeling. Market data distribution has several unique operational challenges:

  • Volume—Market data volumes continue to grow across the financial services spectrum. In the options market alone, the Options Price Reporting Authority (OPRA) estimates volume to approach a million messages-per-second this year. Other sources of messages, such as NYSE Consolidated Quote System (CQS) and NASDAQ TotalView-ITCH, are exhibiting similar increases due to such factors as automation, regulatory requirements and volatility. This cross-spectrum rise in data will continue to tax existing at-capacity software-based messaging systems.
  • Volatility—Market data does not arrive at flat and consistent rates. Market openings and closings contribute to data bursts, as do external events such as Federal Reserve rate announcements. The bursts often overrun the message systems and the subscribing endpoints, forcing devastating retransmission and latency-inducing queuing challenges.
  • Latency—With electronic trading and processing of market data, latency is a critical operational variable. Millisecond latency numbers don’t cut it in the world of microsecond low-latency demands. Equally important to low-latency is the ability to control deviations and latency while eliminating the occurrence of outliers.
  • Scale—Market data must now be delivered to more endpoints than ever before. Because of the volume of market data, more servers are being deployed at the front-end to partition and shore up the publishing capability. Not only does this complicate and increase the challenges of the ingress data feeds, but it ends up impacting and overwhelming both the network and subscribing nodes.

The existing software-based messaging infrastructure used by many firms for market data distribution framework is breaking, yet the traditional mechanisms to compensate only serve to exacerbate the problem.

 

The Tervela Advantage

The Tervela Message Network™ is based on the deployment of Tervela TMX Message Switches™ and it overcomes the challenges of market data distribution by delivering a silicon-accelerated message transport platform that does not contain the inherent deficiencies of legacy software-based messaging systems and message-oriented middleware. Several capabilities of the Tervela Message Network combine to provide a level of intelligence that cannot be matched with software-based peer-to-peer messaging platforms:

  • Intelligent Messaging—The Tervela Message Network can perform byte-rate monitoring on a per-topic basis so that information distribution and consumption patterns can be both filtered and controlled, ensuring both messaging performance and network infrastructure integrity.
  • Intelligent Networking—The Tervela Message Network makes real-time transport decisions, moving topics in and out of multicast groups while dynamically optimizing which topics are sent out on specific multicast groups. When transport decisions are made, they are communicated to the Tervela Client API™ via the control path. As a result, the API automatically switches transport type without application involvement. The API and TMX will also coordinate the changeover to ensure that messages are being received on the multicast channel before they are taken off the unicast channel to safeguard against network problems.
  • Slow Consumer Management—The Tervela Message Network also features a feedback loop that allows for policy-based forwarding in response to real-time changes in traffic patterns. One example is the slow consumer issue, in which the TMX dynamically provides egress queuing and rate throttling for slow consumers without impacting the flow to other consumers. Another example is ingress bandwidth monitoring, in which fast producers are policed to assure that high-priority messages are guaranteed bandwidth.
  • High Fan-Out—Unlike some financial services environments, today’s market data distribution scenarios require the ability to absorb tremendous amounts of information and vast amounts of topics while intelligently filtering and disseminating information on a broad scale.

For more information, learn about the TMX and how it can also support high-performance trading infrastructure, market making and order routing applications.