Quantitative Model Analyst: Model Risk Management - Fraud Validation

at  US Bank National Association

At U.S. Bank, we’re on a journey to do our best. Helping the customers and businesses we serve to make better and smarter financial decisions and enabling the communities we support to grow and succeed. We believe it takes all of us to bring our shared ambition to life, and each person is unique in their potential. A career with U.S. Bank gives you a wide, ever-growing range of opportunities to discover what makes you thrive at every stage of your career. Try new things, learn new skills and discover what you excel at—all from Day One.

JOB DESCRIPTION

U.S. Bank is seeking the full-time position of Quantitative Model Analyst: Model Risk Management - Fraud Validation in Charlotte, NC.

BASIC QUALIFICATIONS:

This position requires a Master’s degree or equivalent in Mathematics or Statistics and 1 year of experience applying mathematical and statistical theory to develop complex statistical models.

REQUIRED SKILLS/EXPERIENCE:

Must also have 12 months of experience with each of the following: 1) Store, retrieve, and manipulate data for use in statistical models including regression and machine learning techniques. Utilize Pandas, Numpy and other packages to manipulate and merge data from multiple sources for further evaluation and analysis. 2) Demonstrate data compilation, programming and qualitative analysis skills including reading, comprehending, and critically evaluating written documentation. Translate complex code and results into clear and explainable written text and vice versa. 3) Utilize statistical and econometric analysis methodologies including variable interpretability using PDP and SHAP. Evaluate and interpret model sensitivity analysis. Must also have experience developing and executing model validation tests and methodologies using Python or similar programming languages. Tests include PSI, AUC, PRAUC, accuracy and misclassification rates using Python or similar statistical package. Will accept experience gained before, during or after Master’s program. Employer will accept experience gained concurrently.
This position is with U.S. Bank National Association, a U.S. Bank company.
Base pay range may vary if an offer is made for work in a different location. Pay Range: $66,955 - $129,699.

ESSENTIAL RESPONSIBILITIES:

Responsible for challenging and validating fraud model outputs and results by developing or applying mathematical or statistical theory and methods to collect, organize, interpret, and summarize numerical data. As part of the Fraud Model Validation team within Model Risk Management, perform second line of defense validation and review to ensure the models used at the Bank are conceptually sound, implemented and used correctly, and aligned with the Bank’s model risk appetite. Specific duties include:

• Provide rigorous and effective Second Line of Defense validation of fraud models;
• Conduct comprehensive review of Model Development, Implementation, and Monitoring documents;
• Maintain frequent and effective communication with model owners and users to understand the use and business context of the models under review;
• Develop and execute model validation tests and methodologies using Python or similar programming languages. Tests include PSI, AUC, PRAUC, accuracy and misclassification rates using Python or similar statistical package;
• Store, retrieve, and manipulate data for use in statistical models including regression and machine learning techniques. Utilize Pandas, Numpy and other packages to manipulate and merge data from multiple sources for further evaluation and analysis;
• Demonstrate data compilation, programming and qualitative analysis skills including reading, comprehending, and critically evaluating written documentation. Translate complex code and results into clear and explainable written text and vice versa;
• Utilize statistical and econometric analysis methodologies including variable interpretability using PDP and SHAP. Evaluate and interpret model sensitivity analysis;
• Perform validation and testing of complex statistical models, which includes extensive programming in Python;
• Produce clear, well-written documentation of all outcomes following established templates and Model Risk Management guidelines;
• Effectively communicate conclusions with senior management, model owners, and key business partners;
• Create efficient, interpretable code to produce streamlined and repeatable results; and
• Work with the following tools, technologies, and methodologies: Tools and Technology: Python, including SciPy, Numpy, Pandas Scikit-Learn, Jinja, and pyWin32 packages. Statistical and ML Modeling Algorithms: General Linear Models (Logistic and Linear Regression), Gradient Boosting Machines (GBM), XGBoost, and LightGBM. Databases: Microsoft SQL Server.

Our approach to benefits and total rewards considers our team members’ whole selves and what may be needed to thrive in and outside work. That’s why our benefits are designed to help you and your family boost your health, protect your financial security and give you peace of mind. Our benefits include the following (some may vary based on role, location or hours):

• Healthcare (medical, dental, vision)
• Basic term and optional term life insurance
• Short-term and long-term disability
• Pregnancy disability and parental leave
• 401(k) and employer-funded retirement plan
• Paid vacation (from two to five weeks depending on salary grade and tenure)
• Up to 11 paid holiday opportunities
• Adoption assistance
• Sick and Safe Leave accruals of one hour for every 30 worked, up to 80 hours per calendar year unless otherwise provided by la