Human beings do not always behave optimally. This is true even with financial decisions, and is especially true with regard to people’s decisions to prepay their mortgages early. Often, a borrower has the opportunity to save a lot of money by refinancing at a lower interest rate, and yet does not. Why not? Who is most likely to prepay their mortgages, and when? In this position we study the behavior of homeowners with mortgages, and build models to predict mortgage prepayments.
1. Have very strong analytic and quantitative skills. Most qualified candidates would have a degree in a quantitative field.
2. A passion for analyzing data and implementing models. As the work involves modeling mortgages, the candidate should have an interest in the economics/finance/psychology associated with mortgage behavior.
3. Have some programming experience, and some basic knowledge of C++
Helpful but not required qualifications:
1. Working knowledge of statistics, especially survival analysis
2. Excel and Visual Basic experience
3. An understanding of finance and economics
4. Experience working with SAS or other statistical software
5. Working knowledge of SQL and databases
A. Study prepayment behavior, focusing on Agency (Fanny, Freddie, FHA, VA) Mortgages
B. Develop insight into borrower, servicer and lender behavior and analyze trends and changes
C. Build predictive models using SAS to predict future mortgage prepayments
D. Develop quality C++ code to implement the predictive models