Senior Manager, Credit Model Development - Commercial Scorecard Development & Maintenance (Hybrid - see description for potential locations)

# \*\* Work Location/Arrangement: This is a hybrid position requiring in-office work four days every week. Ideally, it will be based in Bridgeport, CT but it may be based in Buffalo, NY, Baltimore, MD, Washington, DC, Wilmington, DE, Iselin, NJ, or possibly in New York, NY or another M&T corporate office. ## *\*\* Depending upon the location of the final candidate, there might be potential for remote work.* # Overview: The Manager, Commercial Scorecard & Risk Rating Modeling is responsible for leading the strategic design, development, implementation, governance, validation support, maintenance, and ongoing enhancement of the Bank’s Commercial Risk Rating and Scorecard Models used for credit risk management, portfolio monitoring, regulatory compliance, capital management, and other enterprise-wide initiatives. Establishes the long-term vision and roadmap for commercial credit risk modeling frameworks, ensuring models remain robust, predictive, compliant, and aligned with evolving business objectives and regulatory expectations. Provides subject matter expertise and leadership for the Bank’s commercial underwriting and risk quantification models, with particular focus on Probability of Default (PD), Loss Given Default (LGD), Exposure at Default (EAD), and risk rating scorecards. Oversees the full model lifecycle, including model development, calibration, performance monitoring, back-testing, testing, implementation, documentation, governance, and continuous improvement. The Commercial Scorecard group is a critical component of the Credit Risk Department. Risk Ratings are utilized in many areas of the bank and are a key driver of many enterprise level decisions such as the level of the Allowance for Loan and Lease losses, determining levels of Approval Authority and Asset Quality Metrics. This is mission critical information that is utilized internally across the organization and externally by the Bank Examiners, Outside Accountants, rating agencies and the investment community. The ratings are also a key input into the loss forecasting models utilized for the CCAR process. Loss Forecasting models are used in Capital Plan submissions that are a critical component of sound Bank management and are subject to regulatory scrutiny under DFAST regulations. This role is highly technical in nature and requires demonstrated attention to detail execution and follow up on multiple initiatives within the Credit Risk department. The ability to identify, analyze, rationalize and communicate complex business problems and recommend solutions is a key factor of success in this role. Success in this role requires the ability to use analytics in a collaborative effort across multiple functions and products to derive optimum solutions to business problems. **Primary Responsibilities:** - Oversee the development, implementation, and maintenance of the framework for Commercial PD and LGD credit underwriting models for the institution using internal/external data/environment, next gen technologies, and agile modeling principles - Develop algorithms and tools for testing overall performance, robustness, stability, and ongoing monitoring of the model to ensure compliance of models to internal/external regulations. - Adapt automation and machine learning techniques, data frameworks, and implementation platforms to build scalable modeling solutions across data mining, segmentation, back testing, reporting and ongoing monitoring areas to speed up the model development process. - Develop credit ratings to structured finance transactions, by performing collateral analysis, cash flow modeling, and structural enhancement assessments - Determine when redevelopment or recalibration is needed based on changes in market conditions/regulations/strategy and guide the redevelopment efforts - Partner with Centralized Technology to ensure that Rating models are fully integrated into the appropriate platform which allows seamless delivery to the end user while providin…