Backtesting Your Ideas
A guide to using our backtesting tools to validate your strategies.
Backtesting helps you see how your investment ideas might have performed in the past. With ValQuantIDX, you can access historical index constituents, weights, and performance data through our APIs and test how these indices could fit into your existing research or portfolio models.
This is a high-level overview to help you get started.
What You Can Backtest
You can explore many types of research questions, including:
| Example Use Case | What You Might Learn |
|---|---|
| Compare portfolio performance with and without ValQuantIDX | See if adding ML-driven indices improves results |
| Test your stock screens on historical index constituents | Identify ideas supported by data-driven selection |
| Check performance across market regimes | Understand when the approach may work best |
| Add factor overlays (value, quality, momentum, etc.) | Combine your factors with ML-selected universes |
| Evaluate diversification benefits across regions | Explore multi-market integration opportunities |
These are only starting points—users are encouraged to explore creatively.
Data Available for Backtesting
Through the API, you can retrieve:
- Historical constituents + equal weights
- Historical returns / performance data
- Reference information (e.g., region, sector, metadata)
Data can be accessed in Python or downloaded into Excel/CSV for other tools.
Example High-Level Workflow
Option A — Python / Jupyter Research
- Request historical index data from the API
- Load your own pricing, factor, or research datasets
- Join/merge datasets on date + ticker
- Run your tests (performance, risk, factor comparison, etc.)
- Review findings and iterate on ideas
Option B — Excel / CSV Workflow
- Export index history via API download endpoint
- Open in Excel, Power BI, or another analysis tool
- Compare against your portfolio or watchlist
- Calculate returns, weights, or sector/region patterns
- Create charts and insights for internal review
Tips for Practical Backtesting
- Start simple, then go deeper as needed
- Test multiple time periods and regimes
- Consider liquidity, turnover, and practical constraints
- Avoid assuming outperformance continues unchanged
- Use backtesting as research support, not as a guarantee
Next Steps
If you want to go further, you can explore:
- Multi-region allocation testing
- Factor overlays on top of ValQuantIDX universes
- Portfolio optimization and risk analytics
- Combining internal research signals with ML-based indices
This section is only an introduction—you can build far more advanced workflows over time.
Disclaimer
Past performance does not guarantee future outcomes, and all investment decisions should be evaluated with proper risk management and professional judgement.