Seven-Module Quant Trading Platform -- 2
Budget: £20 – £250 GBP
I’m assembling a production-ready quantitative trading platform built around seven tightly-integrated modules:
1. Data Pipeline
• Pull equities, options and futures data from Alpaca, Interactive Brokers and Yahoo Finance.
• Prioritise data accuracy over raw ingestion speed, with validation, deduplication and time-zone normalisation baked in.
• This pipeline and the Strategy Library are the only components that need to touch external APIs directly.
2. Backtesting Engine
• A hybrid architecture: vectorised processing for speed, event-driven logic for realism.
• Swappable execution models (market, limit, slippage) and portfolio-level handling.
3. Strategy Library
• Five starter strategies: pairs trading, momentum, statistical arbitrage, regime detection and volatility trading.
• Each strategy must expose clear long/short signals and accept risk parameters.
• It should also be API-ready so I can bolt on new data sources later.
4. Risk Management Layer
• Real-time VaR, CVaR and configurable circuit breakers.
• Position sizing and capital allocation rules that slot seamlessly into both backtests and live trading.
5. Performance Analytics & PDF Reports
• Tear-sheets with common metrics (Sharpe, Sortino, drawdown, hit ratio, turnover, etc.).
• Auto-generated PDF reports with Matplotlib/Plotly visuals.
6. Walk-Forward Optimisation
• Robust anti-overfitting workflow: rolling train/test splits, parameter stability checks and equity-curve validation.
7. Streamlit Dashboard
• One-page overview plus drill-downs for data, strategy metrics, risk and live positions.
• Simple authentication so I can share the link with collaborators.
Acceptance criteria
• End-to-end demo on sample equities universe.
• All modules callable from a single orchestrator script (Python 3.11 preferred).
• Clean, commented code with README and environment.yml.
• Unit tests for core functions; ≥80 % coverage.
If you have experience stitching together Alpaca or IB data feeds, writing vectorised backtests and deploying Streamlit apps, let’s talk about timelines and milestones—my goal is a maintainable, extensible codebase rather than a quick prototype.
1. Data Pipeline
• Pull equities, options and futures data from Alpaca, Interactive Brokers and Yahoo Finance.
• Prioritise data accuracy over raw ingestion speed, with validation, deduplication and time-zone normalisation baked in.
• This pipeline and the Strategy Library are the only components that need to touch external APIs directly.
2. Backtesting Engine
• A hybrid architecture: vectorised processing for speed, event-driven logic for realism.
• Swappable execution models (market, limit, slippage) and portfolio-level handling.
3. Strategy Library
• Five starter strategies: pairs trading, momentum, statistical arbitrage, regime detection and volatility trading.
• Each strategy must expose clear long/short signals and accept risk parameters.
• It should also be API-ready so I can bolt on new data sources later.
4. Risk Management Layer
• Real-time VaR, CVaR and configurable circuit breakers.
• Position sizing and capital allocation rules that slot seamlessly into both backtests and live trading.
5. Performance Analytics & PDF Reports
• Tear-sheets with common metrics (Sharpe, Sortino, drawdown, hit ratio, turnover, etc.).
• Auto-generated PDF reports with Matplotlib/Plotly visuals.
6. Walk-Forward Optimisation
• Robust anti-overfitting workflow: rolling train/test splits, parameter stability checks and equity-curve validation.
7. Streamlit Dashboard
• One-page overview plus drill-downs for data, strategy metrics, risk and live positions.
• Simple authentication so I can share the link with collaborators.
Acceptance criteria
• End-to-end demo on sample equities universe.
• All modules callable from a single orchestrator script (Python 3.11 preferred).
• Clean, commented code with README and environment.yml.
• Unit tests for core functions; ≥80 % coverage.
If you have experience stitching together Alpaca or IB data feeds, writing vectorised backtests and deploying Streamlit apps, let’s talk about timelines and milestones—my goal is a maintainable, extensible codebase rather than a quick prototype.