v0.11.0 - Portable refresh and ETF stock rankings

v0.11.0 - Portable refresh and ETF stock rankings

This release extends the monthly ETF and stock research framework with current-date refresh, research-validation experiments and interactive sector reports.

Added and improved

• One-click Windows setup and refresh: prepare a project-local Python environment, download prices, archive raw and calculated data, and generate reports using the current clock.
• Eleven independent ETF-sector stock-ranking pages. Click an ETF from the overview or either portfolio report to view its sector stock sample.
• All 100 indicator definitions, formulas and statuses, with selectable ascending/descending rankings and CSV exports. Missing values remain unranked and sort last.
• English default reports and documentation, with Chinese companion pages.
• All calculated scalar performance metrics and per-asset actual weight statistics in the portfolio reports.
• Nine predefined research experiments covering transaction costs, holdings count, ranking window and factor-group contribution removal. Save all results and failures.
• Completed-session and month-end checks, source/artifact hashes and preserved run history. Failed refreshes leave the previous successful report available.

Validation evidence

• 109 local unit and integration tests passed on Windows.
• The Windows launcher completed a live refresh for 55 stocks and 11 ETFs. All nine validation experiments completed.
• The final English-report run contained 22 sector language pages; 765 local links and 68 artifact hashes were checked.
• GitHub Actions is configured for tests and offline demonstrations. Consult the checks on the tagged commit for the actual remote result.

Scope and limitations

Decisions remain monthly. Each sector currently contains five configured stock samples; these are not complete or historical ETF constituents. Of 100 candidate indicators, 65 price/volume factors are implemented and 35 financial factors await point-in-time fundamentals. The workflow downloads prices and generates reports; it does not download company annual-report or SEC filing text.

A chronological holdout does not certify unseen data. Fixed-universe selection bias and vendor adjustment revisions remain. Internal data checks do not constitute independent price certification. macOS/Linux launcher scripts are supplied but have not been tested on those systems.

Getting started

Download and extract 'Quant-research-lab-v0.11.0-source.zip', then double click 'START_V011.cmd' on Windows. Internet access is required. Open 'START_HERE.md' for instructions. The source bundle excludes generated reports, market data, databases and runtime caches; the launcher builds these locally.

The product name v0.11 corresponds to package 0.11.0 and release tag v0.11.0. This release includes the earlier research-validation milestone discussed as v0.1.1. Daily, hourly and AI event-driven versions remain planned.

中文辅助：本版增加当前日期一键更新、11 个 ETF 板块独立个股排名页、全部指标升降序查看和研究验证报告。默认英文，可切换中文；每板块当前为 5 只样本股，35 项财务指标仍待数据。

v0.1.0 - Monthly research baseline

Preserve the existing monthly research framework as the baseline before upgrading to v0.11.0.
Features:

• Monthly sector ETF allocation.
• Multifactor stock selection and backtesting.
• Portfolio accounting with transaction costs.
• English and Chinese research reports.
  Baseline commit: 7bf8fd0
  This baseline is tagged retrospectively to preserve the version history.