• Data Scientist (Contractor)

    Job Locations
    US-CA-San Francisco
  • Overview

    First Republic Bank seeks a quantitative consultant/contractor for a 3-month project.


    The consultant should have excellent programming skills in R and T-SQL. The consultant will help the Bank convert stress test credit loss models to meet the FASB ASU 2016-13 (aka “CECL”) accounting standards, which fundamentally changes the way U.S. banks calculate and hold allowance for loan losses (aka “ALLL”). The consultant will be responsible for solving the following problem statement:


    Measure and explain changes in the allowance under the CECL framework (i.e, which proportion of the change was due to loan data changes, economic scenario changes, prepayment behavior changes, addition of new loans, additional draws on existing lines of credit).


    The consultant is expected to:

    • Propose and develop a data monitoring program and change management process for Risk Data Mart, Moody’s Data Buffet, and Current Expected Credit Loss models – all of these components are sources of allowance changes
    • Measure, monitor, attribute, and explain allowance changes for each month by portfolio
    • Standardize model inputs and outputs across all the CECL models and automate reports of allowance results and attribution of the month-over-month changes
    • Develop a process for model integration with a technology platform (SQL Server 2016 ML Services)
    • Develop a process for model execution
    • Integrate and implement CECL models into a production system that will run on a scheduled basis

    The consultant will work closely with our Finance team, Accounting team, Credit team, Technology team, and SOX team, but will not build predictive models.


    Job Requirements:

    • PhD or MS degree in a quantitative field (e.g., statistics, computer science, analytics, operations research)
    • 3-5 years of programming experience in R and T-SQL
    • 3-5 years of professional experience collaborating with multiple teams at a company
    • Familiarity with linear and logistic regression methods


    Sorry the Share function is not working properly at this moment. Please refresh the page and try again later.
    Share on your newsfeed