Our new Intelligence Hub analytics platform helps ensure intraday trading always stays compliant
An annual event in the Corvil calendar is the Deutsche Börse IT Open Day, when the Frankfurt-based exchange provides it members with technical updates on the latest (and planned) enhancements to the Eurex and Xetra trading systems. We attended last week with a booth in the main atrium to hear the keynotes from one of the most technically progressive venues and to network with its members, many of whom are Corvil customers.
In January this year, the MiFID II regime introduced new scrutiny and sanctions around Order to Trade ratios (OTR) in an attempt to mitigate the risk of algorithmic trading systems creating disorderly trading conditions.
To stay compliant with the new regulations, trading members need to be able to track the ratio of unexecuted orders to trades intra-day, and in as close to real time as possible so that corrective action can be taken to ensure regulatory violations do not occur. Sanctions can be significant, with the threat of fines of up to €1m and a ban from trading for 30 days.
Matt Davey and I had the opportunity to deliver a half hour presentation and demo of our predictive Order-to-Trade Ratio (OTR) solution built upon our new Intelligence Hub platform. The talk certainly caught people’s attention driving visitors to the Corvil booth afterwards.
Deutsche Borse continuously tracks OTR behaviour per-product and per-firm during market hours, issuing a report at the end of each trading day, which includes evidence of any violations. But of course, this is too late for the trading member to take corrective action, as the horse has already bolted and the firm are left to handle the consequences.
This is where Intelligence Hub comes in. By providing continuous insights into trading activity including Order to Trade ratio throughout the day, Intelligence Hub generates alerts that allow time for the firm to intervene and take corrective action. We achieve this with predictive analytics powered by machine learning.uUsing regression modeling that combines intraday data with historical data. Behaviour is learned on a per-instrument and per-product basis, enabling continuous predictions to be made and compared against exchange-mandated thresholds.
It works in fours steps:
Feeding Intelligence Hub are Corvil appliances, which are capable of acquiring network packet data at scale. Working across highly distributed trading environments, spanning multiple desks and algorithms; they collect granular data that can now be surfaced in Intelligence Hub dashboards for visualization and analysis with unprecedented speed and accuracy.
Deutsche Börse is a long-standing customer – they use Corvil to monitor the performance of market data as it travels from the exchange to their regional POPs. Last week’s IT Open day it was a great opportunity to connect with many of their members, and introduce them to a solution, that will help them avoid fines and sanctions by staying within the guiderails of MiFID II RTS-9.