DATA MANAGEMENT

BIG DATA ANALYSIS

  • The need to analyze Tick Data and Order Book Data led us to adopt dedicated technologies to store and process Big Data. We adopted RAPIDS for strategies developed with Python and the CUDA for trading systems built with C++/C. 
  • Even, more importantly, we employ specific algorithms built for BIG-DATA (Leskovec et al., 2016, Ait-Sahalia and Jacod, 2014, Aldridge and Avellaneda, 2021).
  • Our data are cleaned and organized to maximize their informative content: relative entropy maximization, structural breaks identification, Maximum Memory Preservation, …
  •  We are working on building an efficient infrastructure to store and analyze FIX data.
web-3706562_1280-min

DATABASES

To make our strategies operational, we employ an extensive number of databases:

  • OptionMetrics for data on options.
  • the CRSP/COMPUSTAT databases for factor analysis and equity data.
  • ICE for 5 minutes bars on futures.
  • IBES to extract measures of uncertainty and sentiment.
  • Interactive Brokers for streaming data
  • TickData for the tick data needed for execution algos.