Back-test & stress-testing: record, replay, back-test

Overview

Recordings of live market and pricing data enables our clients to back-test market data for algos, test algos in real-time conditions and define sophisticated trading strategies.

Running with the Horizon replayer, back-testing replays historical events as fast as possible, providing a reliable simulation of the trading infrastructure and market conditions. The recorder is a tick-by-tick recorder plugged to the Internal Market. It will record all tick events (i.e. bid, ask, last changes of price and quantity) into either a proprietary format file – called binary logs, optimised but not legible by humans – or into CSV, or into a dedicated database for further analysis.

The replayer takes a binary log as input and replays it as if it were a real-time stream. The replayed stream will be monitored in the GUI’s views used to monitor the market. The start time and end time of the replay can be chosen as well as the acceleration factor (2x, 5x, 10x, 60x, best effort).

Screenshot - Automated Trading

Key features

Top benefits

  • Recording data to test market-making and algos in the most accurate conditions
  • Test algos in real-time conditions
  • Replay a selected day of historical market data (summary, depth) and pricing data (price vectors)
  • Connect to historical data and back-test trading strategies
  • Output custom data for analysis in Horizon or for analysis with external tools (e.g. Python Libraries)
  • Ultra-low latency impact
  • Accurate market simulation
  • Extreme conditions simulations
  • Build custom reports for continuous validation or for regulators
  • Customizable order matching simulation, latency simulation and trading universe management