Key Takeaways
- The PSA Standard Prepayment Model forecasts prepayment risk for mortgage-backed securities by assuming a gradual increase in prepayment rates to 6% by 30 months.
- Prepayment risk arises when loans are paid off early, impacting the duration and cash flows of mortgage-backed securities, often negatively affecting investors.
- Introduced in 1985, PSA remains a prevalent model among several types used to evaluate prepayment in asset-backed securities.
- The PSA Model, now under SIFMA, initially assumed a 0% prepayment rate increasing by 0.2% monthly, peaking at 6% after 30 months.
- A prepayment rate exceeding the PSA forecast can shorten a security’s life, returning capital to investors in potentially unfavorable yield environments.
What Is the PSA Standard Prepayment Model?
The Public Securities Association Standard Prepayment Model (PSA) is a tool used to estimate the rate at which borrowers will pay off their mortgage loans early, impacting mortgage-backed securities. Understanding PSA is crucial for investors in mortgage-backed securities (MBS) and collateralized mortgage obligations (CMO) to assess how early loan repayments can affect the lifespan and yield of their investment. The PSA model is pivotal in estimating that prepayments rise gradually and peak at around 30 months, helping investors manage their risks and returns effectively.
How the PSA Model Works in Mortgage-Backed Securities
The Public Securities Association Standard Prepayment Model (PSA) acknowledges that prepayment assumptions will change during the life of the obligation and may affect the yield of the security. The model assumes a gradual rise in prepayments, which peaks after 30 months. The standard model, called 100% PSA, starts with an annualized prepayment rate of 0% in month zero, with 0.2% increases each month until peaking at 6% after 30 months.
Prepayment assumptions are based on homebuyer data that shows that, during the first few years, a borrower is less likely to be willing or able to prepay the mortgage. This data makes sense, as a new homeowner is unlikely to move to a different home or immediately refinance, and the costs of buying a house generally don’t leave a lot of free cash flow for a new owner to make additional payments.
It is important to note that the PSA is just the most common prepayment model. There are different models, including proprietary ones, that can be used to model and evaluate prepayment in mortgage-backed investments. The Public Securities Association Standard Prepayment Model is also referred to as the PSA prepayment model.
Why Investors Rely on the PSA Prepayment Model
If the single monthly mortality (SMM) for a given MBS or CMO is above what was projected according to the PSA, then the security may see its overall lifespan shortened. This can result in capital being returned to the investors sooner than planned.
The return of capital through prepayment is generally a negative for investors, as prepayment tends to increase in low-interest environments, meaning the investors receive capital back that they must reinvest in a less favorable yield environment. So there is a negative impact on the trading value of a security that is exceeding the PSA. In the opposite case, the life of an MBS may be lengthened if the prepayment rates are below the PSA, assuming the PSA was used in the creation and marketing of the security.
The History and Evolution of the PSA Model
The Public Securities Association Standard Prepayment Model was developed by the Public Securities Association in 1985. The Public Securities Association eventually became the Bond Market Association and, in 2007, it merged with the Securities Industry Association to become the Securities Industry and Financial Markets Association (SIFMA).
The prepayment model is still referred to by its original name, but due to the subsequent name changes of the association, it is sometimes called the Bond Market Association PSA. It is also quite common for the acronym for the model to be confused with the identical acronym for former Public Securities Association as well as an acronym for the function of the model, that is, providing a prepayment speed assumption (PSA).
