VeighNa 2025 Review: ML Research, Risk Controls, and Framework Updates

Summary

This annual review summarizes VeighNa’s 2025 development, including its tenth year as an open-source project and the release progression from version 4.0.0 to 4.3.0. It describes vnpy.alpha, a workflow for machine-learning research and trading that covers factor feature engineering, model training, cross-sectional and time-series strategies, data management, signal generation, and backtesting. The review also covers support for Python 3.13 and related engineering changes.

Trading-focused updates include a redesigned pre-trade RiskManager with configurable rules for order counts, order flow, duplicate submissions, order size, and order validity, along with persistent logging of key engine actions. The article reports project adoption and community activity figures, but these are not evidence of strategy performance. Its Python speed comparison is presented as a general benchmark claim, not a VeighNa-specific trading test. The review is a project account rather than an independent evaluation of the tools or their results.

Key ideas

• The vnpy.alpha module supports factor engineering, model training, strategy development, and research workflows.
• The review describes support for Python 3.13 and updates to the project’s build and code-quality tooling.
• VeighNa 4.2.0 introduced configurable pre-trade risk rules and enhanced operational logging.
• Adoption and community metrics describe project activity but do not establish trading performance.
• The reported Python performance comparison is not presented as a strategy-level latency test.

Tags

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