How raymondycp/quantfinance_algotradingstrategy works
Project type and purpose
• This is a collection of Jupyter-notebook–based quantitative-finance tools rather than a traditional multi-service application. It focuses on strategy development (SMA crossover), backtesting, parameter optimization, and position-sizing simulation/visualization.
File structure
• Top level contains four notebooks and no separate “src” folder.
• [QuantFinance]Sma_FineTurningParameter_Backtesting_py.ipynb – SMA crossover strategy, data fetching, backtesting, Sharpe-ratio optimization
• [QuantFinance]Backtest_Backtrader.ipynb – uses Backtrader framework for backtesting
• [QuantFinance]PositionStrategy.ipynb – Kelly-criterion–based position sizing, Dash app for interactive visualization
• [QuantFinance]Scalping_Tradingview.ipynb – (presumed) scalp‐strategy implementation via TradingView data
• Config file: .gitattributes
README for additional insights
• Data is fetched from the FinMindTrade API (TSLA 2024).
• Multithreaded grid search is used for parameter optimization.
• Dash provides an interactive UI for position sizing.
• Core metrics: returns, Sharpe Ratio, max drawdown.
a. Main components to represent
1. Data Source (FinMindTrade API)
2. Data Ingestion / ETL module (requests → pandas DataFrame)
3. Strategy Engine(s)
– SMA Crossover logic
– Scalping logic (TradingView)
4. Backtesting Engines
– Custom backtester in notebook
– Backtrader library
5. Optimization Engine (grid search + Python multithreading)
6. Sizing Engine (Kelly Criterion math core)
7. Visualization / UI
– Jupyter notebooks (plots)
– Dash web app
8. Orchestration environment (Jupyter kernel)
b. Relationships and interactions
• The Data Source feeds into the Data Ingestion module.
• Ingestion outputs feed simultaneously into the Strategy Engines.
• Strategy Engines pass “signal” data to Backtesting Engines.
• Backtesting Engines output performance metrics.
• Metrics feed into the Optimization Engine.
• Optimization Engine loops back to Strategy Engine to refine parameters.
• Final trade results feed into the Position Sizing Engine.
• Sizing Engine outputs are fed to the Dash UI for interactive simulation.
c. Architectural patterns & design principles
• Pipeline / modular ETL pattern
• Feedback loop for optimization (grid search ↔ strategy)
• Separation of concerns: Data fetch, strategy logic, backtest logic, optimization, sizing, visualization
d. Key technologies/frameworks/libraries
• Python
• pandas (data manipulation)
• FinMindTrade API (data source)
• Backtrader (backtesting framework)
• Python threading / concurrent.futures (parallel grid search)
• Dash (interactive web app)
• Jupyter Notebook environment
Elements
• Clear labels: e.g., “FinMindTrade API,” “SMA Strategy Engine,” “Backtrader.”