Optimizing Multi-Factor Weights by Tuning Expression Lists

Summary

The discussion addresses how to search for effective weights when combining two stock-ranking factors. It proposes using BigQuant’s parallel tuning module, m.tune, rather than treating each weight as a separately exposed parameter. The key implementation idea is to treat the complete ranking expression as the tunable parameter and supply a list of expressions representing different factor-weight combinations for evaluation.

The post gives a conceptual method but no complete example code, specified objective function, backtest results, or chosen weights. It points readers toward an example and notes that a template would be helpful, while one participant says the tuning process remains unclear to them. Consequently, the document illustrates how to expose composite factor expressions to a parameter search; it does not establish that any particular combination is robust or profitable. Any result would depend on the backtest design and data, which are not detailed here.

Key ideas

• The proposed task is to search for weights in a combined stock-ranking expression.
• The post suggests passing multiple weighted expressions to m.tune as the search candidates.
• The expression itself can be treated as the tunable parameter even if weights are not separate module parameters.
• The document does not provide completed code, a scoring criterion, or evidence that a selected weighting generalizes.

Tags

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