I built a Java backtester to reproduce two published options strategies

Javier G — Wed, 26 Aug 2026 17:49:40 +0000

Most options backtests publish a final chart, but often leave out the exact requests, fill rules and execution assumptions needed to reproduce it.

I wanted to build something where the complete process could be inspected, so I created an open-source Java backtester and used it to replicate two publicly described short-duration options strategies: Early Retirement Now and WealthyOption.

Results
Both tests use SPXW data from January 2017 to July 2026.

Strategy	My backtest	Published result
ERN — 0DTE + 1DTE puts	1.55% annualized return	7.06% annualized return
WealthyOption — 3 puts + 3 calls	17.58% CAGR / 21.43% max drawdown	23.1% CAGR / 19.2% max drawdown

These aren’t exact like-for-like comparisons. The ERN result published by its author includes discretionary decisions, while my implementation is entirely rules-based. The periods, data and execution models also differ.
Technical details

The project includes:

A streaming Java backtesting engine
A Spring Boot API
ThetaData five-minute option data integration
SQLite persistence
Configurable commissions, slippage and fill assumptions
Versioned requests and complete backtest results
Documentation covering the methodology and known limitations

The core engine is separated from the ThetaData integration so other data providers can be added later. The published tests include commissions and use documented bid/ask-based fill rules rather than assuming every trade executes at the midpoint.

The repository is here:
github.com/Javier-Garzo/ern-wo-options-backtester

The main goal is to let other developers inspect the assumptions and reproduce the results instead of having to trust a performance chart. Feedback on the architecture, execution model or methodology is welcome.

DEV Community: Javier G

I built a Java backtester to reproduce two published options strategies