Prepayment modeling is one of the most important elements in predicting a loan’s future cash flow. Cash flow, of course, is the primary value-driver for both mortgage-backed securities (MBS) and mortgage servicing rights (MSR). As prepayment activity increases, interest received on an MBS investment and servicing fees that a servicer collects on payments from borrowers decrease. Logically, the earlier a prepayment risk can be detected, the better it is for a portfolio manager who has more time to assess and address the risk.
The refinance boom of the past couple of years has demonstrated the limitations of traditional prepayment models, which rely on historical prepayment data to predict future prepayment activity. Unusually low interest rates, combined with an increase in tappable equity, significantly increased prepayments on loans with higher rates and raised the bar for early and accurate prepayment modeling. When market conditions changed rapidly historical data was not a reliable indicator of future prepayment activity.
Fortunately, there’s a better way. Mortgage lock data, recorded when a borrower takes the initiative to lock in a mortgage rate 30- to 90-days prior to closing on a new loan, provides one of the earliest and most reliable indications that an existing mortgage is going to be paid off early.
A new white paper published by Black Knight’s Optimal Blue Division examines this innovative new modeling approach and demonstrates how rate lock activity can be used to accurately forecast short-term prepayment speeds. The paper is available for free download here.