DPFounder & Head of Research

Early-career professional exploring practical experience

I have some work experience and I'm actively developing skills and taking on projects to shape my career path.

View my workMessage Me

Work samples

A collection of projects and tasks completed while I build experience. Each entry shows what I worked on and what I learned.

  • USD/JPY ASIAN FX MEAN REVERSION STRATEGY
    USD/JPY ASIAN FX MEAN REVERSION STRATEGY

    Built a systematic mean reversion backtest on USD/JPY using a 20-day z-score signal to identify overbought and oversold conditions, validated across 20.5 years of daily OHLCV data spanning two BOJ normalization cycles. Engineered a dual confirmation filter combining realized volatility regime detection and AUD/USD momentum to capture JPY safe-haven dynamics, reducing active trading to 11.2% of days while maintaining a 52.4% hit rate. Measured six model configurations including trend following, time stops, and CFTC overlays, finding that signal simplicity outperformed complexity and producing a four-panel dashboard of entry signals, z-score zones, cumulative returns, and rolling Sharpe.

    PythonClaude CodepandasNumPymatplotlibyfinance

About me

I have some work experience and I'm actively developing skills and taking on projects to shape my career path.

I'm Dominic Petoral. I have some work experience but consider myself at the start of my career, exploring roles and building skills as I grow.

Experience

Founder & Head of Research

ALPHA GUIDE CAPITAL · July 2026 – Present

Equities Trading Analyst

SMB CAPITAL · May 2026 – July 2026

Investment Analyst Associate

THE GENAU GROUP REALTY ADVISORS · November 2024 – September 2025

Education

THE GEORGE WASHINGTON UNIVERSITY, School of Business

Bachelor of Science in International Business, Concentation in Finance · Class of 2026

Skills

Early-career experienceProject-based learningSkill developmentHands-on practice
Back to work

USD/JPY ASIAN FX MEAN REVERSION STRATEGY

Built a systematic mean reversion backtest on USD/JPY using a 20-day z-score signal to identify overbought and oversold conditions, validated across 20.5 years of daily OHLCV data spanning two BOJ normalization cycles. Engineered a dual confirmation filter combining realized volatility regime detection and AUD/USD momentum to capture JPY safe-haven dynamics, reducing active trading to 11.2% of days while maintaining a 52.4% hit rate. Measured six model configurations including trend following, time stops, and CFTC overlays, finding that signal simplicity outperformed complexity and producing a four-panel dashboard of entry signals, z-score zones, cumulative returns, and rolling Sharpe.

PythonClaude CodepandasNumPymatplotlibyfinance
View all work