Summary:
3 years’ industry experience in account level CCAR and CECL model development on international unsecured portfolios.
6 years’ academic research experience in econometric modeling (e.g. time series, panel data models, 2SLS, OLS, instrument variable, regression discontinuity, etc.) and statistical modeling (e.g. logistic regression, decision trees, PCA, cluster analysis, non-parametric methods).
Solid knowledge in data science and machine learning techniques, including cross-validation, bootstrapping, cluster analysis, support vector machines, bagging, random forest, gradient boosting etc.
Extensive experience in experimental design and behavioral economics
Proficient in various statistical software including R, SAS, STATA and Eviews.
Familiar with Python, including Pandas, Numpy and Jupyter.
VP, Quantitative Modeler, Risk Modeling Utility, Citigroup, Wilmington DE 10/2017 - Current
Full-cycle account level CCAR & CECL credit risk modeling experience on international unsecured portfolios, including data cleaning, model development, documentation and validation defense.
Continuously monitoring and re-evaluating existing models, redevelop models as needed.
Constructed cohort level recovery models to complete end-to-end NCL predictions, provided team with adoption package including code, manual, templates and presentations.
Research Assistant, University of Delaware, Newark DE 06/2015 - Current
Built agent-based computer simulation models to understand watershed level pollution dynamics.
Designed economic experiments to parameterize agent-based model.
Used various supervised and unsupervised machine learning techniques to draw insights from data.
Presented research and won awards in national and international research conferences.
Intern, Model Development, APM, PNC Financial Services Group, Wilmington DE 09/14-05/15
Conducted CCAR modeling for Student Loan and Auto Loan portfolios.
Performed statistical modeling and econometric forecasting on Probability of Default (PD) of loans under different economic scenarios.
Cleaning, wrangling, graphing and modeling large financial datasets.
Developed model documentation and ensured model met federal requirements.
Teaching Assistant, University of Delaware, Newark DE 09/2013-05/2014
Guided over 100 students on topics related to econometrics and macroeconomics.
Research Assistant, University of Delaware, Newark DE 09/2011-08/2013
Utilized SAS/R/STATA extensively in various studies to perform data extraction, cleaning, analyzing and reporting on different types of data including panel data, survey data and experiment data.
Built a large number of econometric models: e.g. panel data/time series models (mixed-effects, fixed & random-effects, diff-in-diff), categorical response models (Logit, Probit), Tobit, 2SLS, OLS.
Used a variety of statistical methods including classification and regression tree, cross-validation to validate econometric analysis results.
Worked with USDA-ERS (Economic Research Services) economists on various panel/time series data sets to identify the effect of an insurance policy on farmland values over time.
Presented research at high-standard professional conferences (AAEA/NAREA/ESA).