THE PROBLEM: Private Student Loans are a form of consumer Installment debt with particularly vexing attributes:
- Unsecured, generally high balances
- Long-dated credit extensions (in some cases 15- 25 years)
- Often no repayment performance observed for years after disbursement
- Extended to primary borrower with little or no credit history (e.g. undergraduates)
THE ANALYSIS: There are two distinct phases in the maturation of a Private Student Loan portfolio. The first is dominated by credit risk and can persist for up to seven years after origination given an often long period of unobserved performance. Thereafter, for the balance of the portfolio's life (often 15-20 years) the risk drivers shift to adverse selection in prepayment and credit migration concerns.
THE SOLUTION: Statistically we can construct a compound model which draws upon Logistic Regression to assess early stage performance and then couples this with actuarial-like techniques (Survival Analysis) to project the "tail" performance for the portfolio's remaining decade-long life. Our work has demonstrated a much slower and longer-evolving loss development curve than most portfolio sponsors, holders and rating agencies had been projecting.
Kirk Monteverde published the seminal statistical paper on risk management for private student loans in the academic journal, Research in Higher Education (Vol. 41, No. 3, 2000), a copy of which is available for download HERE.