Matt Johns Email & Phone Number

New York, New York, Us

• Integral contributor to the development, validation, and implementation of the new Current Expected Credit Losses (CECL) model used to calculate multi-billion dollar credit reserves, based on a Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD) methodology.• Primary liaison between the CCAR and CECL modeling teams and senior leadership, internal and external auditors, and regulators.• Provided guidance on the development of machine learning models for CCAR, including Artificial Neural Networks and Random Forests.

New York, New York, Us

• Forecasted deal term credit losses and approval rates for potential de novo and acquisition partnerships, taking into account complexities such as product design, customer demographics, macroeconomic outlook, etc.• Responsible for identifying potential risks and mitigants for each deal based on a holistic understanding of the deal terms and P&L.• Proactively re-engineered the loss forecasting model to increase flexibility and capabilities while reducing development time and user error. The improved model was integral to the analysis underpinning an upcoming portfolio acquisition.

San Jose, Ca, Us

• Developed a new credit line increase (CLI) program targeting top PayPal customers which generated ~$23M in profit over 24 months. • Expanded the reactive CLI strategy to increase approval rates and optimize credit lines.• Created a new underwriting strategy for two segments of applicants representing ~25% of total applications using a combination of bureau and internal data to increase approval rates while remaining risk neutral. • Pioneered the use of Tableau within my department by creating several interactive dashboards that are now used at all levels, from individual analysts to the Credit VP & GM.

• Performed analysis on the predictive power of three credit bureaus in the UK in preparation for PayPal Credit’s first overseas product.

• Acted as lead developer, designed and edited extensive SAS code used to automate the ETL (Extract-Transform-Load), analysis, and data visualization process to output 150+ graphs directly to PowerPoint on a monthly basis.

Laurel, Maryland, Us

• Developed information-theoretic metrics used to measure and predict control algorithm performance in real time. This work was later used to develop a hybrid control system for Navy ships.

Timonium, Md, Us

Worked as a SAS developer and data analyst in the credit department.