Quantitative Developer / Options Trading Researcher needed

Project Type: Fixed-price technical consultation with optional follow-up development Primary Language: Python 3.11+ Market Focus: SPY / QQQ 0DTE Options Current Stage: Research prototype / historical backtesting Long-Term Goal: AI-assisted research, supervised strategy improvement, and automated trading 1. Project Overview I’m developing ZeroDTE Lab v0.6, a personal quantitative trading research platform focused on SPY and QQQ zero-days-to-expiration (0DTE) options. The project already has an existing Python codebase, including a custom event-driven backtesting engine, options Greeks validation, statistical testing, historical-data utilities, and research auditing tools. I’m not looking to hire someone to build a trading bot from scratch. My immediate objective is to have an experienced quantitative researcher or developer independently provide the necessary real options/trading data, evaluate the existing system, identify its most important weaknesses, and recommend affordable, targeted improvements. Longer term, I intend to evolve the project into a brokerage-connected paper-trading system and eventually a risk-controlled automated trader with AI-assisted analysis and supervised strategy optimization. 2. Current Trading Strategy The initial strategy focuses on intraday directional momentum in SPY and QQQ options expiring the same trading day. Current research parameters include: * Relative Volume: A 2× volume threshold intended to identify unusual trading activity. * Options Delta: Minimum absolute delta of 0.75 for contract selection. * Momentum Confirmation: EMA, VWAP, and breakout-based conditions. * Stop-Loss: An initial $0.40 option-premium stop. * Trade Management: Trailing exits and configurable risk parameters. These rules are hypotheses, not validated sources of trading profitability. I’m particularly interested in whether these parameters make statistical and practical sense, whether they should be adjusted, and whether the underlying strategy has any measurable predictive value after realistic trading costs. 3. Existing System and Technology Stack The project is modular and primarily written in Python. Language: Python 3.11+ Hosting: Linux/Digital Ocean Droplet Container: Docker Version Control: Git CI/CD: GitHub Action Market Data: Read Only Alpaca Integration Historical Data: CSV/Parquet optional DuckDB Planned Historical Provider: (CHOICE FROM YOU!) … ThetaData or other OPRA vendor BackTester: Custom event-driven Python simulator Options analytics: Custom Black–Scholes Greeks calculations and validation Statistical research: Walk-forward, CPCV, Romano–Wolf, stationary/block bootstrap, Newey–West variance estimation Risk testing: Adversarial scenarios and execution stress tests AI tooling: Optional LLM-assisted code and research auditing (ChatGPT or Claude) Automated tests: pytest / unittest Code quality: Ruff / Vulture / Bandit Optional research references: QuantConnect LEAN / NautilusTrader (The project does not currently include a production brokerage order-execution system.) Existing Research Capabilities * Historical strategy simulation and trade reporting. * Options quote, timestamp, and Greeks quality checks. * Walk-forward validation. * Combinatorial Purged Cross-Validation (CPCV) split planning. * Romano–Wolf multiple-hypothesis testing correction. * Bootstrap confidence intervals and performance uncertainty analysis. * Sharpe-based diagnostics and drawdown analysis. * Adversarial testing for unfavorable execution and market conditions. * Read-only market monitoring. * Optional AI-assisted research and code auditing. Important: The bundled example results use synthetic data. No statistically validated profitable trading edge has been established. CPCV currently supports split planning rather than a complete automated model-retraining workflow. 4. Scope of Work — Initial Quantitative Audit I would like the consultant to review the following areas. A. Strategy and Options Analytics * Assess the statistical rationale behind the 2× relative-volume threshold, 0.75 delta filter, and EMA/VWAP momentum conditions. * Evaluate whether the $0.40 premium-based stop-loss is appropriate for SPY/QQQ 0DTE options. * Review contract selection, Greeks calculations, implied volatility assumptions, and option liquidity filters. * Identify any missing market features that would materially improve the research methodology. B. Backtesting and Execution Accuracy * Identify potential look-ahead bias, survivorship bias, timestamp misalignment, or data leakage. * Review bid/ask spread handling, slippage, commissions, and fill assumptions. * Assess how the simulator handles rapid price movements, stale quotes, stop-loss gaps, and insufficient liquidity. * Determine whether the custom event-driven simulator is suitable for continued development or would benefit from comparison with an established framework such as LEAN or NautilusTrader. C. Statistical Validation * Review the correctness and appropriateness of walk-forward validation. * Assess CPCV purging, embargo, and split construction. * Verify the Romano–Wolf stepdown implementation and its assumptions. * Evaluate bootstrap methodology, Sharpe-based diagnostics, and multiple-testing controls. * Recommend how to distinguish genuine out-of-sample performance from overfitting or data mining. D. Software Architecture and Reliability * Review Python code quality, modularity, and maintainability. * Identify important bugs, unnecessary complexity, and performance bottlenecks. * Assess test coverage and failure handling. * Identify missing components required for reliable paper trading. * Recommend/Implement improvements without unnecessarily rebuilding the existing system unless proven to be absolutely necessary. E. Risk Management and Future Automation * Evaluate position sizing, maximum daily loss, exposure limits, and emergency shutdown requirements. * Recommend a safe architecture for broker-connected paper trading. * Assess how future AI-assisted analysis and supervised strategy optimization could be integrated. * Identify safeguards needed to prevent unvalidated strategy changes from reaching an active trading environment. 5. Historical Market Data I plan to acquire historical SPY/QQQ options data from ThetaData or a comparable provider, ideally using a short-term subscription to download a reusable research dataset where licensing permits. Desired historical data includes: * Intraday underlying price and volume. * Historical option chains, including expired contracts. * Bid/ask quotes and timestamps. * Options trades and liquidity information. * Delta, gamma, theta, vega, and implied volatility where available. * Sufficient historical coverage to evaluate different volatility and market regimes. I’d appreciate guidance on whether this dataset is sufficient and which additional fields are essential. The initial engagement does not require the developer to purchase or acquire the complete historical dataset. 6. Expected Deliverables 1. Technical Audit Report: A written assessment identifying specific issues in the existing codebase and research methodology. 2. Prioritized Findings: Critical, moderate, and optional improvements, ranked by importance. 3. Reproducible Evidence: Code references, test cases, or examples demonstrating material issues. 4. Statistical Assessment: An opinion on whether the existing validation methods are correctly implemented and appropriate for the strategy. 5. Data Recommendations: A clear explanation of the minimum historical dataset needed for credible testing. 6. Development Roadmap: Recommended steps toward reliable paper trading, supervised strategy improvement, and eventual automated execution. 7. Implementation Estimates: Estimated hours and costs for the highest-priority fixes. The consultant should clearly distinguish between issues verified in the code and potential risks that require further testing. 7. Longer-Term Development Roadmap The following phases are future objectives, not requirements for the initial audit. Phase 1 — Validate Historical Research Acquire suitable historical options data, evaluate strategy performance, and independently verify statistical and execution assumptions. Phase 2 — Broker-Connected Paper Trading Integrate live market data and simulated broker order execution with comprehensive logging, risk limits, and monitoring. Phase 3 — AI-Assisted Research and Supervised Optimization Introduce a feedback loop capable of analyzing trading outcomes, identifying changing market regimes, proposing strategy improvements, and testing alternative configurations. Proposed changes should undergo statistical validation, paper testing, version control, and human approval before deployment. Phase 4 — Automated Execution If the system demonstrates a sufficiently reliable edge and passes operational risk testing, implement automated order placement, position monitoring, broker reconciliation, exposure limits, daily loss limits, and emergency shutdown controls. The long-term objective is a personal trading system that can continuously evaluate its performance and propose improvements without independently deploying unvalidated changes. 8. Preferred Qualifications I’m looking for someone with practical experience in: * Python quantitative development. * US equity/ETF options, preferably SPY/QQQ and 0DTE. * Options Greeks, implied volatility, and market microstructure. * Historical options data and realistic backtesting. * Statistical hypothesis testing and overfitting prevention. * Walk-forward validation, bootstrap methods, and multiple-comparison corrections. * Brokerage APIs and paper-trading infrastructure. * Code review, testing, and technical documentation. Experience with QuantConnect LEAN, NautilusTrader, ThetaData, or Databento is a plus. A quantitative researcher with strong statistical and options-market experience is preferable to a general-purpose AI developer.