Skip to content
All library documents

Using a Historical Data Window for Pair-Strategy Simulation

Article BigQuant
Read the original

Summary

This platform-support exchange concerns a pair-trading strategy that failed to run in simulation. The suggested fix is to leave the data-extraction module’s start-date option unbound to live-trading parameters. The explanation is that the strategy’s regression step needs data covering a period of time; binding the extraction to trading parameters can provide a live-oriented data selection instead of the required historical window.

The user confirms that the change worked. The exchange therefore offers a concise troubleshooting lesson about matching input data scope to a model’s needs: a regression used in a pair strategy requires a suitable span of historical observations. It does not describe the regression specification, the selected period, or any trading results, so it should not be read as a complete strategy guide.

Key ideas

  • The pair-trading simulation failed because its regression needed a historical data window.
  • The suggested fix was to avoid binding the extraction module’s start date to live-trading parameters.
  • The user reported that the change resolved the simulation failure.
  • The exchange does not provide the regression details or evaluate the strategy’s trading performance.

Tags

Pairs tradingStatisticsBacktesting

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.