As a model validation risk professional, my research experience provides effective challenge of enterprise bank balance sheet DFAST stress testing models. My role is to help banks understand the model risks associated with key financial variables such as deposit balances, loan growth rates, loan loss and recovery rates. Because of the systemic risks associated with banks, financial institutions must provide a credible estimate of the worst case scenario of their credit related losses so that they can hold sufficient capital reserves to be resilient and withstand stressful episodes. These reserves will heavily depend on the projected values of their loans, deposits, and other modeled balance sheet variable forecasts. This makes model risk awareness especially important for banks.
As a consultant, I conducted research that extends the central findings of my thesis on return predictability. The aim of my applied research is to develop an investment product that taps into systematic, predictable sources of returns in global financial markets that are uncorrelated with global equity returns. In addition, I strive to make the product simple and intuitive to understand. Further, if the promise of the idea delivers in practice, I will help investors gain access to global financial returns that are robust to various investment environments.
AVP, Model Validation - Model Risk Management
Perform validation and annual review of enterprise models
Identify, quantify and evaluate credit risk and its correlation with market risk and macroeconomic factors. Validate bank stress testing econometric models built with macroeconomic aggregate data.
Utilize industry data to enhance peer benchmark analysis and variable selection. Expanded knowledge of intersection between market and credit risk through lens of bank stress test modeled interest rate sensitivity.
Perform validation of enterprise stress testing bank balance sheet models; effective challenge of existing econometrics models using data mining software that statistically over-fit recent history data; prescribed changes to more appropriate variable selection process driven by industry data and research, theory.
Spearheaded peer benchmarking analysis using publicly available data to enhance the quantification of credit risk (expected loan losses) in the following classes: credit card, consumer loan, and home equity loans.
Enhanced model risk oversight in bank credit risk models with required minimum performance metrics tied directly to development sample metrics; aimed at helping modelers understand the bias-variance trade-off in their models.
Consultant, Research
Worked on replication and analysis of global "smart beta" strategies by building a simulation of global macro value strategies.
Confirmed the following insight: Global portfolio construction and performance analysis revealed that smart beta investment strategies combine (1) asset class diversification and (2) additional expected excess returns to consistently outperform the market on a risk-adjusted basis; they behave like 'long-duration' assets in providing steady returns without the equity-like volatility.