Avoiding Stale Cache Results in BigQuant Backtests

Summary

A BigQuant user reports that a stock AI strategy template fails with an index error when its prediction or backtest date range includes January 1, 2016. The issue is described as intermittent and associated with using certain factors. The suggested workaround is to enable a cache-related parameter in each module. For older code-based strategies without modules, the reply suggests adding a changing dependency value to a function so the cache is not reused.

The post provides a reproduction path through a built-in strategy template, but it does not establish the underlying cause or show that either workaround fixes the error reliably. The changing-value approach may force recomputation, at the cost of reducing cache reuse. Treat this as community troubleshooting advice rather than a diagnosis or a general recommendation for handling empty data or index errors.

Key ideas

• The reported failure occurs when a backtest date range includes January 1, 2016.
• The error is an index-out-of-bounds exception in backtest and trade modules.
• The reply recommends enabling a cache parameter in each module.
• For older code strategies, a changing function dependency is suggested to avoid reusing cached results.
• The post does not verify the root cause or demonstrate that the workarounds resolve it.

Tags

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