BMLL Technologies Limit Order Book Analytics
Analytics on a global, multi-asset class, deep-history, multi-petabyte limit order book dataset
BMLL Technologies Limit Order Book Analytics can be rapidly deployed on AWS
Investment firms are increasingly required to make and justify decisions based on empirical data science. In order to stay competitive, they need to ensure this is done using the most relevant datasets whilst exploiting the latest advances in statistical science.
BMLL Technologies provides a web-based platform that combines a global, multi-asset class, multi-petabyte limit order book dataset with a Jupyter and Python front end. Limit Order Book Analytics is driven by scalable Spark clusters with a range of advanced analytics based on machine learning.
BMLL Technologies is an APN Advanced Technology Partner and has achieved AWS Financial Services Competency. Competency Partners have industry expertise, solutions that align with AWS architectural best practices, and staff with AWS certifications.
Utilize Limit Order Book Analytics for a series of applications
Core platform: API access to a curated object store through a logical data model combined with scalable computing power and value-add analytical toolboxes
Best execution: Next generation TCA product driven by machine-learning providing actionable intelligence using machines to leverage human capability, not replace it. Examples include broker recommendations, counter-signaling (“what information would I leak to the market if traded in this way?”) and trade timing.
Market abuse and trade surveillance: An application delivered to the desktops of compliance officers and traders that allows them to detect manipulative patterns in their own order flows as well as the wider market, while ensuring that fills lie within user-defined tolerances
LOB simulation: A distributed simulation and back testing environment for aggressive and passive orders allowing for transaction costs and market impact. Advanced features include synthetic data generation for a limit order book based on a feature space from historical data, enabling trade execution to be designed using reinforcement learning.