Mo Rahmani

Quantitative researcher — systematic FX forecasting, rigorous backtesting, and research infrastructure.

About

I build end-to-end quantitative research pipelines: from raw market and macro data through feature engineering, rolling out-of-sample model evaluation, and portfolio-level analysis. My work emphasizes reproducible backtests, explicit risk overlays, and communication of methodology — not black-box performance claims.

Currently focused on multi-timeframe FX directional forecasting (daily through intraday), expanded-universe backtests, and integrated research dashboards for signal monitoring and production workflows.

Research principles

  • Out-of-sample evaluation first — rolling holdout and walk-forward protocols, not in-sample fit.
  • Transparent data lineage — versioned parquet pipelines, manifest reports, reproducible run scripts.
  • Risk-aware framing — stop-loss overlays, transaction-cost sensitivity, and clear limitation statements.

Selected research

Non-linear FX directional forecasting

Multi-timeframe binary and ternary classification on FX pairs using merged macro fundamentals (FRED), technical features, and cross-pair context columns. Rolling OOS tuning with batch holdout sweeps across daily, weekly, hourly, and 4-hour bars.

Python Rolling OOS Parquet pipeline H1 / H4 / Daily

FX data repository & model inputs

Automated collection of FX OHLC (FXCM + Yahoo tail fill), FRED macro panel, and aligned fundamentals/technicals for modeling. Manifest-driven reporting for series coverage, date spans, and failure diagnostics.

FRED API Yahoo Finance Feature engineering ETL

Portfolio & stop-loss overlay research

Batch export sweeps with stop-loss overlay analysis across expanded FX and non-FX universes. Portfolio preview workflows combining multi-timeframe signals with explicit risk constraints for research review.

Backtesting Risk overlay Multi-asset

Integrated forecasting dashboard

Streamlit-based research workspace unifying FX forecast signals, data refresh controls, and cross-asset analytics modules — designed for daily research workflows and operational monitoring.

Streamlit Research ops Automation

Tools & methods

Python pandas / NumPy scikit-learn Time-series CV Backtesting Parquet / ETL Macro data (FRED) FX microstructure PCA / factor analysis PowerShell automation Streamlit Git

Contact

Open to senior quantitative analyst and quant researcher roles in investment management. Reach me at mr.mh.rahmani@gmail.com or via LinkedIn.