Mehrnoush S.

Quantitative Risk Specialist @Federal Reserve Bank of San Francisco

San Francisco, CA, US

Stress Testing & Loss Forecasting: Developed econometric models to forecast retail and wholesale losses by integrating large bank regulatory data with Federal Reserve macroeconomic stress scenarios. Directed modeling for Dodd-Frank Act Stress Tests (DFAST), projecting losses, revenues, expenses, and capital adequacy under adverse conditions.· Model Validation & Quality Assurance: Led validation efforts with robust performance testing frameworks. Evaluated assumptions, assessed stress performance, and recommended recalibrations. Oversaw disclosure reviews of wholesale and trading models, ensuring compliance and recommending adjustments to maintain sensitivity and comparability.· Predictive Credit Risk & Loan Pricing: Designed a predictive risk assessment framework for credit losses using survival analysis and machine learning on recovery data. Validated models outperforming traditional default-rate methods, improving loan pricing and risk-adjusted lending. Built end-to-end ML pipelines to forecast defaults across retail and wholesale portfolios. Developed scalable tools to analyze borrower behavior and sensitivity to macroeconomic shocks.· Term Structure Modeling & Interest Rate Transition: Estimated dynamic yield curve models for debt valuation by fitting term structures to historical data and generating out-of-sample forecasts. Built models to support the transition from LIBOR to SOFR under macroeconomic scenarios, enabling robust analysis in a changing regulatory environment.

University of Illinois Urbana-Champaign

Doctor of Philosophy - PhD, Economics (Finance)

Alzahra University

Bachelor of Science - BS, Industrial Engineering (Computer Science)

Sharif University of Technology

Master of Arts - MA, Economics