Diagnosing Missing Trades and Unexpected Exits in an Equity Backtest

Summary

This platform support discussion explains two apparent backtest anomalies in a single-stock strategy: why a position did not begin on the requested date at full allocation, and why the stock appeared to be sold later. The response attributes the delayed start to a basic stock-selection module that depended on other data tables; those data were unavailable at the beginning of the test, so the strategy did not establish a position until later. Removing that module reportedly shifted the first position to the following day.

The later event is attributed to a trading halt. Missing market data produced a NaN factor value, which prevented the strategy from buying that day. The discussion cites an external finance site as corroboration of the halt, but supplies no code, execution trace, or fuller account of the alleged sale. It illustrates how data availability and missing prices can affect backtest timing and decisions; the brief exchange does not establish a general platform defect or address how to model suspended securities robustly.

Key ideas

• A selection module may delay trades when it relies on data that are unavailable early in the backtest.
• The reported strategy began holding the stock after that module was removed.
• A trading halt can leave price inputs missing and cause factor calculations to return NaN.
• The discussion offers a case diagnosis rather than a complete investigation or general platform rule.

Tags

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