Argentum Capital
A Python-based investment research and timing framework focused on building long-term positions at attractive prices.
Argentum Fund is a personal investment research project built in Python and versioned on GitHub under the Argentum Capital label.
The core idea is simple:
Good investments are not only about what you buy. They are also about when you buy.
This project is designed to help identify securities that appear attractive on a longer-term basis, while emphasizing buy timing, disciplined cash deployment, and structured portfolio building.
Rather than chasing the fastest possible gains, the Argentum Fund is built around a more measured philosophy:
- identify strong longer-term opportunities
- monitor valuation and price dislocation
- deploy capital gradually and intelligently
- build positions at favorable entry points
The current project has two main research avenues:
A research workflow focused on value-oriented securities, especially within the S&P 500 Value universe. The objective is to identify securities that have declined in price, yet still display attractive underlying metrics such as:
- Price-to-Earnings Ratio (P/E)
- Price-to-Book Ratio (P/B)
- Price-to-Sales Ratio (P/S)
- EV/EBITDA (Enterprise value relative to operating earnings)
- Dividend Yield
- Free Cash Flow Yield
- other quality and valuation indicators over time
A supporting workflow that evaluates holdings derived from:
- 13F-based portfolios
- copied or mirrored investor portfolios
- Autopilot tracker holdings
The goal here is not to blindly copy public portfolios, but to evaluate the best-performing or most compelling securities inside those portfolios using the same valuation, quality, and timing logic.
Argentum Fund is not intended to be a high-frequency trading system or a prediction machine.
It is intended to be a decision support system for longer-term investing.
The emphasis is on:
- disciplined accumulation
- timing entries rather than reacting emotionally
- managing cash deployment during uncertainty
- comparing value and momentum in a practical way
- learning through transparent, repeatable analysis
In practical terms, this means the program aims to answer questions such as:
- Which securities currently look attractive?
- Are they cheap for a good reason or a bad reason?
- Is now a reasonable time to buy?
- Should new cash be deployed normally, accelerated, or held?
- Are there strong candidates inside a copied 13F or Autopilot portfolio?
- Is portfolio concentration becoming too high?
- Config-driven universe building
- Manual CSV universe support
- Standardized schema for securities and strategy sleeves
- Support for synthetic internal identifiers such as cash holdings
- Support for international CSV conventions including comma-decimal formats
- Standardized
current_universe.csvoutput for downstream analysis
- Historical price ingestion
- Value / quality / momentum factor generation
- Buy-timing signals
- Cash deployment logic
- Tracker-specific portfolio analysis
The universe builder creates a standardized investment universe from either:
- manual CSV inputs
- future SEC 13F ingestion workflows
This ensures that all later analysis operates on a consistent structure.
A major goal of the project is to identify when to buy, not only what to buy.
That means the platform is designed to consider:
- drawdowns from recent highs
- rolling returns
- valuation changes
- quality metrics
- potential recovery signals
- cash available for deployment
This project also serves as a software and research portfolio. It is being developed to improve skills in:
- Python programming
- modular software design
- GitHub version control
- financial data analysis
- quantitative research workflows
- eventual machine learning experimentation
The roadmap is intentionally modular. Future versions may include:
- price history builder
- factor engine for valuation, quality, momentum, and risk
- ranking engine for candidate securities
- weekly report generation
- market condition summaries
- portfolio overlap and concentration analysis
- paper portfolio support for testing alternative ideas
- SEC 13F ingestion module
- Autopilot holding analyzer
- configurable scoring engine
- backtesting framework
- buy-signal calibration
- performance comparison against benchmarks such as VTI, VOO, or value ETFs
- sell-discipline module
- rebalancing rules engine
- brokerage API integration
- alert system for buy and sell conditions
- dashboard or web app interface
- machine learning or ranking models
- market regime detection
- paper trading or simulation environment
A mature version of Argentum Fund could eventually support a workflow like this:
- Build or refresh the investment universe
- Pull price and fundamental data
- Score securities on value, quality, and timing
- Compare opportunities within the value universe and 13F-based portfolios
- Generate a ranked list of candidates
- Recommend whether to buy now, scale in gradually, or wait
- Track performance in both live and paper portfolios
- Eventually identify not only when to buy, but also when to trim or sell
A simplified project structure is expected to look like this:
argentum-fund/
│
├── README.md
├── requirements.txt
├── config/
│ └── universe_config.yaml
│
├── data/
│ ├── manual/
│ ├── processed/
│ └── outputs/
│
├── src/
│ ├── universe/
│ ├── data/
│ ├── features/
│ ├── scoring/
│ ├── portfolio/
│ ├── reports/
│ └── backtest/
│
└── tests/
Argentum Fund is a personal software and research project.
The author is not a licensed financial advisor, investment adviser, broker, or other regulated financial professional. Nothing in this repository constitutes:
- financial advice
- investment advice
- tax advice
- legal advice
- a solicitation to buy or sell any security
- an offer to manage money or operate an investment fund
All content in this repository is provided for educational, research, and software development purposes only.
Any investment decisions made using this code, analysis, or output are made entirely at the user’s own risk. Markets involve risk, including the possible loss of principal. Past performance does not guarantee future results.
No warranties or guarantees are made regarding:
- accuracy
- completeness
- reliability
- timeliness
- fitness for a particular investment purpose
Users should conduct their own due diligence and, where appropriate, consult a qualified financial professional before making investment decisions.
This project was developed with the assistance of generative AI tools.
Generative AI was used to assist with:
- brainstorming project structure
- drafting code scaffolding
- discussing system design
- refining documentation
- exploring analytical ideas
All code, documentation, and investment logic should be reviewed critically by the project owner and by any downstream user. AI assistance does not imply correctness, financial suitability, or regulatory compliance.
Argentum Fund is the name of this repository and research project.
Argentum Capital is the branding identity used in project materials and visuals.
This branding is informal and conceptual. It does not indicate the existence of a registered investment company, fund, adviser, or legal financial entity.
This project sits at the intersection of several interests:
- long-term investing
- value-oriented research
- disciplined portfolio construction
- software engineering
- GitHub-based project development
- quantitative thinking
- learning by building
At its best, Argentum Fund should become both:
- a practical investment research tool
- a serious public software project that demonstrates technical growth over time
This is currently a personal project, but feedback, discussion, and constructive ideas are welcome.
Planned releases may gradually expand the project from a universe-building and buy-timing framework into a broader research platform that includes:
- sell signal logic
- backtesting
- portfolio comparison
- paper trading
- broker integrations
- reporting dashboards
- machine learning experiments
The immediate focus, however, remains clear:
build a disciplined framework for identifying attractive securities and improving the timing of long-term buys.
License to be added.
Jaemin Eun
GitHub project under the Argentum Capital label.