Trading Strategy Challenges

This repository contains implementations of two quantitative trading strategy challenges focused on signal mapping and options-based forecasting.

Overview

Both challenges involve developing systematic trading strategies with specific risk-adjusted return objectives. The implementations prioritize robustness, interpretability, and risk management over model complexity.

Challenge 1: Signal Mapping Strategy

Objective

Create an optimal mapping from 51 input signals (10 long-term + 41 short-term) to portfolio exposure values that beat a provided baseline in risk-adjusted returns.

Approach

• Signal Processing: Standardized signal transformations with volatility-based filtering
• Feature Engineering: Signal interactions, momentum indicators, and regime detection
• Model: Ridge regression with walk-forward validation to prevent overfitting
• Risk Management: Dynamic exposure limits with volatility-adjusted position sizing

Key Features

• Volatility regime detection for conditional signal weighting
• Signal interaction terms to capture non-linear relationships
• Walk-forward validation with expanding windows
• Aggressive leverage in high-conviction scenarios (up to 3.0x)

Performance Metrics

The strategy is evaluated on:

• Sharpe Ratio: Risk-adjusted returns
• Calmar Ratio: Return relative to maximum drawdown
• Maximum Drawdown: Largest peak-to-trough decline

Files

• Challenge1_Signals/QV_Signal_Strategy.py - Main strategy implementation
• Challenge1_Signals/Signal_Mapping_Analysis.ipynb - Development process and analysis
• Data/QQQ_base - QQQ_base.csv - Baseline signal reference data

Challenge 2: Options-Based Directional Forecasting

Objective

Use end-of-day QQQ options data to forecast next-day directional movement with exposure range -1.0 to +1.5 and target Calmar ratio > 2.0.

Approach

• Data Source: Millions of rows of end-of-day options prices and Greeks
• Feature Engineering:
  • Options sentiment indicators (put/call ratios, skew)
  • Volatility surface metrics (IV percentiles, term structure)
  • Positioning signals (open interest, volume concentrations)
  • Greeks-based risk metrics (gamma exposure, delta imbalance)
• Model: Ensemble approach combining multiple signal types
• Risk Management: Strict exposure limits with downside protection

Key Features

• Comprehensive options Greeks utilization (delta, gamma, vega, theta)
• Put-call ratio analysis across strikes and expirations
• Implied volatility surface analysis for regime detection
• Open interest and volume flow analysis
• Lookahead bias prevention in all feature calculations

Performance Targets

• Calmar Ratio: > 2.0
• Sharpe Ratio: > 1.5
• Exposure Range: -1.0 to +1.5

Files

• Challenge2_Options/QV_Options_Strategy.py - Main strategy implementation
• Challenge2_Options/Options_Strategy_Analysis.ipynb - Development process and analysis
• Data/options_eod_SPY.csv - SPY options end-of-day data (6.5M rows)
• Data/options_eod_QQQ.csv - QQQ options end-of-day data (5.0M rows)

Data Files

Large Data Files (Not Included in Repository)

Due to GitHub's file size limitations, the following large data files are not included in this repository:

• Data/options_eod_SPY.csv (~1.35 GB)
• Data/options_eod_QQQ.csv (~1.04 GB)

Note: These files are required to run the Challenge 2 options strategy. The baseline signal file for Challenge 1 is included.

Included Data

• Data/QQQ_base - QQQ_base.csv - Baseline signal reference for Challenge 1

Options Data Schema

The SPY and QQQ options datasets (when obtained) contain:

• tradeDate: Trading date
• spotPrice: Underlying price at close
• strike: Option strike price
• delta, gamma, vega, theta, rho: Options Greeks
• callPrice, putPrice: Option prices
• callOpenInterest, putOpenInterest: Open interest
• callVolume, putVolume: Trading volume

Date range: 2020-present (depending on data availability)

Requirements

Python Version

• Python 3.8 or higher

Dependencies

pandas>=1.3.0
numpy>=1.21.0
scipy>=1.7.0
scikit-learn>=1.0.0
matplotlib>=3.4.0
jupyter>=1.0.0

Installation

1. Clone the repository:

git clone https://github.com/yourusername/Trading-Strategy-Challenges.git
cd Trading-Strategy-Challenges

2. Install dependencies:

pip install -r requirements.txt

Or install manually:

pip install pandas numpy scipy scikit-learn matplotlib jupyter

Usage

Running Challenge 1

python Challenge1_Signals/QV_Signal_Strategy.py

Running Challenge 2

python Challenge2_Options/QV_Options_Strategy.py

Exploring Analysis Notebooks

jupyter notebook Challenge1_Signals/Signal_Mapping_Analysis.ipynb
# or
jupyter notebook Challenge2_Options/Options_Strategy_Analysis.ipynb

Results

Both strategies demonstrate the ability to generate positive risk-adjusted returns through systematic signal processing and disciplined risk management.

Challenge 1: Signal Mapping

• Successfully beats baseline metrics through robust signal combination
• Demonstrates regime-aware exposure management
• Shows consistent performance across walk-forward validation windows

Challenge 2: Options Strategy

• Achieves target Calmar ratio through comprehensive options analysis
• Utilizes multi-dimensional features from options Greeks and positioning
• Maintains strict exposure discipline while capturing directional opportunities

Methodology Notes

Development Process

The Jupyter notebooks document the iterative development process, including:

• Initial feature exploration and analysis
• Model selection and hyperparameter tuning
• Performance evaluation across multiple metrics
• Identification and resolution of lookahead bias
• Walk-forward validation results

Risk Management

Both strategies implement:

• Strict exposure limits to control portfolio risk
• Volatility-based position sizing
• Maximum drawdown monitoring
• Regime-aware signal adjustments

Acknowledgments

Research conducted with assistance from Perplexity AI.

License

This code is provided for educational and demonstration purposes.

Author

Will Hammond November 2025