1. Validated Commercial Real Estate and Commercial & Industrial Loan Level Models, including PD, LGD, Balance, Loss Forecast Components for CCAR, CECL, IFRS9 and BAU purposes; 2. Conducted independent testings, built benchmark models, drafted validation reports and prepared exam materials for management and regulators.

1. Built residential mortgage, auto loan and credit card capital models for BASEL III regulation purposes;2. Collected data, built loan level models, and validated models through backtesting, sensitivity tests, etc, using python and SAS.

1. Examined data integrity, accuracy and completeness2. validated model framework including pricing model, marketing model, PPNR model and other risk models3. reviewed model methodology, assumption and limitation. I worked with many different types of models, including machine learning model, time series model and logistic regression model.4. evaluated model development and testing results, replicated model parameters, and checked model implementation platform and performance monitoring plan on Finance models, PPNR models, Compliance models, Fair Lending models, and Marketing Strategy models.5. Drafted model validation report, summoned validation meetings with related counter-parties, and presented validation result to senior management, model developer and model user.

1. Built CRE, C&I and Residential Mortgage loan pricing and production models and tested models trough assumption soundness analysis, in sample testing, back testing, sensitivity analysis and stress testing.2. Construct Mortgage loan Origination Fee and Gain on Sale Revenue models to forecast sellable mortgage profit using waterfall calculation method. 3. Created implementation files and Model Monitoring templates to forecast new origination volume and pricing for PPNR and BAU purposes.

1. Consumer Real Estate Asset Project Generated Residential Mortgage loans’ risk inventory analysis, Quantifiable Support Metrics (QSM) and loan analysis on Modified Loans, Affordable Lending, etc.Analyzed Residential Mortgage and Home Equity Non-performing Loan (NPL) Roll Forward Trends through linking the Other Real Estate Owned (OREO) asset database with Consumer data warehouse loan fact information and performance observation (Default, Prepayment, etc.). Created Residential Mortgage segment level bad rates based on product program, loan purpose, occupancy type, etc. for loan pricing purpose.2. Loan Valuation Project Forecasted Residential Mortgage and Home Equity Expected Loss according to loan characteristics including origination vintage, product types, lien positions, occupancy type, fico score, LTV and other risk factors.Imported expected loss curve, yield curve and outstanding balance curve into QRM cash flow engine to calculate appropriate Real Estate loans valuation.

Collected, cleaned and analyzed Residential Mortgage, Home Equity, Auto Loan and Credit Card data from Data Warehouse.Built Probability of Default (PD) model and Loss Given Default (LGD) model using logistic regression, Generalized Linear Model (GLM), Dual-Time Dynamics (DtD) Decomposition. Validated PD and LGD models through residual test, back testing, sensitivity test and hold-out sample testing.Forecasted portfolio level loss and Non-performing Loans under Baseline, Core Adverse and Severe Adverse economic scenarios.Conducted Bank’s Consumer Credit Loss Forecast and Capital Review under supervisory scenarios for Federal Reserve’s Dodd-Frank Act Stress Test (DFAST, 2014, 2015) and Comprehensive Capital and Review (CCAR, 2014, 2015, 2016), and provided documentation for CCAR and DFAST.