|
Consumer Mortgage prepayment adversely affects profitability of any Mortgage portfolio and predicting the prepayment behavior either for individual loans or loan pools is an important and non-trivial tasks. From an investment perspective prepayment drives down profitability of a portfolio because investors receive invested capital earlier than expected and now need to re-invest it in an unfavorable interest rate environment (loan prepayment accelerates when interest rates are lower). Even Loans that are sold to Agencies like Fannie Mae and Freddie Mac to be securitized need to analyzed to manage revenue cash-flows associated with Mortgage Servicing Rights (MSR) or to leverage opportunities to market to customers from Sold Loan or MSR portfolio if they are at risk of refinancing with competitors.
From a Marketing perspective, loans are acquired at some premium either through a teaser discount rate or by paying a premium to buy loan pools from other banks and if these loans refinance out of the portfolio before the bank has an opportunity to recover this investment, this investment becomes a sunk cost.
Typically econometric models are employed to predict loan prepayment- individual loan prepayment probabilities are predicted using Logit Models, while prepayment rate for whole loan pools are predicted using Proportional Hazards model. |
