Overview
One of the ways to tackle pre-payment modeling is to assume that pre-payments occur at a given pre-specified rate, which is given as a function of the age of the loan. There are multiple standard pre-payment rate assumptions in use in the market.
The standard assumption of a pre-payment rate is that a given percentage of the dollar amount outstanding will be pre-paid. However, these model give no way to determine which loans in a pool are the loans that will pre-pay. Because the loans may have different terms and different rates assigned to them, the choice of which loans pre-pay will affect the overall results.
One way to handle this situation is to just assume that the pre-payment occurs proportionally across all loans. That is, no loan is fully prepaid, but each loan will prepay part of its balance.
A second alternative is to assume that the pre-payment rate expresses a probability of prepayment. (In this case, it would identified as a hazard function in survival analysis). When interpreted as a probability, it can be included in a simulation of the pools cash flows.