Corvil today introduced a new class of data analytics to address the challenges concerning transparency, surveillance and compliance for trading businesses participating in high speed financial markets and algorithmic business in general.
Machine-time analytics are being introduced as part of Corvil's latest release of its streaming analytics platform.
New research by the Tabb Group called "Speed II - Have we reached the tipping point" examines the changing role of speed and the new challenges that have emerged.
"We are beginning a new chapter for speed in financial markets where information about speed will become as, or even more, important than speed itself. We call this Speed II. In this new world, the business with the best machine-time intelligence for execution, operations and compliance will gain competitive advantage," said Larry Tabb, CEO and founder of the Tabb Group.
Machine-time is defined as the time it takes a machine to act or respond to information. In the world of electronic trading, machines can execute trades autonomously in under one hundred microseconds and can make decisions to trade in less than ten microseconds. These machines are capable of taking actions that can be one million times faster than a human. We need a new way for humans to make sense of machine decisions and actions occurring at these timescales. Machine-time analytics is therefore the ability to analyze the actions and decision of machines at the timescale and granularity with which they act.
"It's time to re-think the role of speed in today's financial markets," said Donal Byrne, CEO of Corvil. "With markets operating at blindingly fast speeds, coupled with new ad-hoc mechanisms to slow them down and a growing array of complex order types, we need a reliable way to make sense of everything that is happening. We need to safeguard and provide transparency into all trading activity. This is the reason we introduced the new class of data analytics we refer to as machine-time analytics."
Machine-time analytics starts with the ability to continuously monitor, analyze, track and forensically verify the precise time and sequence of all machine actions and events involved in executing a trade. This provides us with the necessary granular and accurate data to get full transparency into machine level activity. We can now make sense of machine events and detect potential market abuse based on trusted data.
Corvil's machine-time analytics platform is based on streaming analysis of granular data with precision timestamps. Corvil uses network data as its primary data source because it is the richest and most granular source of machine data available and provides an independent, immutable record of what actually happened.
Corvil is also introducing App Agent which delivers machine-time visibility of internal application events with ultra-low overhead. The App Agent solves the problem of efficiently offloading the work of serializing, batching and publishing time-stamped application events. With the addition of App Agent, the Corvil solution now provides machine-time visibility from wire to application to wire.
Machine-time analytics requires accurate timestamping for all data to make sense of the order of events, and to correctly identify causality. For electronic trading, machine-time analytics platforms must maintain accurate time. Corvil's platform maintains time accuracy in the microsecond range synchronized to Coordinated Universal Time (UTC) time sources and uses nanosecond timestamps throughout. This makes Corvil fully compliant with the upcoming MiFID II regulations which require all investment firms to capture, manage, synchronize and maintain all machine messages involved in an algorithmic trade to within one hundred microseconds of UTC.
"Platforms that can provide firms with better fine-grained speed/time transparency and manage machine-time intelligence will help businesses successfully navigate the new challenges of the Speed II era," said Larry Tabb, CEO and founder of the Tabb Group.
The new Corvil Tera+ Release (9.1) which supports streaming machine-time analytics is now generally available.