Prepayment Model
MBS Prepayment Model
Plutus Analytics
Introduction
An MBS prepayment model is essential to MBS valuation and analysis. It predicts the level of prepayments on an MBS pool or a portfolio of mortgage whole loans, assuming the loans can be easily securitized into MBS. In the US, most mortgages come with a prepayment option which allows borrowers to pay off their mortgage earlier. The prepayment option makes it quite challenging to price mortgages as it introduces uncertainty to the future cash flow of a mortgage, even at the pool level. To that end, the prepayment model was created so that the true value of a mortgage can be more accurately estimated.
In a nutshell, a prepayment model is a long and often very complex mathematical equation which takes interest rates, home prices, mortgage attributes and other macroeconomic variables as input and computes a quantity named single monthly mortality or SMM. People are probably more familiar with the annualized version of SMM – CPR.
Main Structure of a Prepayment Model
A prepayment model usually consists of several additive components. Each component captures a typical reason for mortgage prepayment. In this section, we will discuss some of the main reasons why mortgages are prepaid.
Turnover
When mortgage borrowers sell their home, they normally need to pay back their mortgage. One exception is the mortgage is assumable. Assumable mortgages have become rare in recent years because to assume a mortgage the home buyers would need to make a sizable down payment that covers the seller’s original down payment, the amortized principal payment over time, and any home value appreciation.
The turnover component of a prepayment model aims to give an accurate estimate of the prepayment due to housing turnover. There could be many factors affecting the level of turnover. Following are some of the main ones.
Mortgage age is very often considered as a factor affecting the turnover level. When homeowners move into a new home, they usually live there for a while. For that reason, if we know a mortgage is used to purchase a house, then it is safe to assume that the turnover rate will be low at least in the next couple of years. On the other hand, if a mortgage is used to refinance another mortgage, then it is less clear. In that case, we may need to first estimate how long the homeowners have lived there and then adjust the turnover prediction accordingly.
It is well known that the housing market exhibits a very strong seasonal pattern. People usually buy homes in the spring. By the time winter arrives, the housing market becomes much less active.
The mortgage rate also affects the turnover level through a so-called lockin effect. When the market prevailing mortgage rate is significantly higher than the current note rate of homeowners’ mortgage, the homeowners may become reluctant to list their home for sale and therefore drive down the turnover level. On the demand side, a higher mortgage rate makes a home less affordable and reduces the number of buyers out there. The summer of 2022 is a perfect example of that. As the Fed kept raising rates, the mortgage rate skyrocketed to a 14-year high. The high mortgage slowed down the housing market significantly.
When home value appreciates fast, some homeowners may be tempted to sell their home if they have considered selling. This is especially true for those mortgages used to purchase investment properties. As a result, we usually see an elevated turnover level in a hot housing market and vice versa.
Sometimes, mortgage borrowers’ choice of the mortgage product also gives an indication of how likely they will move. For example, some homeowners choose to use an adjustable rate mortgage to finance their home purchase because they know they will move within a certain time horizon.
There are other factors affecting the turnover as well, such as geographical location, number of days in a particular month, to name a few.
Curtailment
Curtailment is another form of prepayment. There are mainly two types of curtailment. The first one is that some homeowners like making extra principal payments whenever they can so that they can save on their mortgage interest and pay off the mortgage early. The second type of curtailment happens when the mortgage approaches its maturity. Some mortgage borrowers simply prefer to pay off their mortgage when there are only a few payments left so that they can get writing a check for the mortgage off their chores list.
Refinance
Refinance is arguably the most important component of prepayment, not only because it accounts for more prepayments than the other types but also because it is heavily driven by the interest rate and exposes mortgage lenders and investors to great interest rate risk if not properly managed.
Many homeowners refinance their mortgage to lower the interest rate on their mortgage. This is known as the rate-term refinancing. There is another type of refinance, where mortgage borrowers cash out some of the equity they have in their home by refinancing their current mortgage. This type of mortgage is also called the cash-out mortgage.
For a rate-term refinance, the most important factor is the prevailing market mortgage rate and the current note rate on the homeowners’ mortgage because homeowners refinance their mortgage only if they can lower their mortgage rate. The lower the prevailing rate compared to the homeowners’ current mortgage rate, the more they can save by refinancing their mortgage.
In addition to the mortgage rates, there are other factors that affect how likely homeowners would refinance their mortgage or how easily they can refinance. Among these factors are their mortgage balance, credit score, debt to income ratio, loan-to-value ratio, to name a few.
Some factors that have significant impact on refinance are not directly observable. Examples of such factors include the so-called burnout effect and media effect. Other factors, though observable, affect the refinance in less obvious ways. For example, mortgages in certain states are known to refinance less frequently than other states for some reasons. Another example is that mortgages originated at a bank’s branch office appear to refinance differently than those originated through a mortgage broker. To learn more about how these factors affect prepayments and how they are modeled in the Plutus Analytics prepayment model, please contact the customer support team at sales@plutusanalytics.com.
A cash out mortgage is less sensitive to mortgage rates than a rate-term refinance mortgage as its primary purpose is to cash out the equity in the home. While a cash out mortgage shares a number of similarities with a rate-term refinance mortgage, there are some differences between the two as well. One such difference is that loan-to-value ratio, while not necessarily among the most important drivers for a rate-term refinance, turns out to be one of the most important factors affecting the cash out refinance. This is understandable in that for homeowners to cash out they need to have a meaningful amount of equity in their home to begin with. Another difference in the opposite way is that though mortgage balance is one of the most important drivers for a rate-term refinance it is less important to a cash out mortgage.
Default
So far, we have discussed several types of prepayment: turnover, curtailment and refinance. These prepayments are also known as voluntary prepayment. In other words, these prepayments are initiated by mortgage borrowers voluntarily.
In a way, default can be viewed as a type of mortgage prepayment as well, albeit involuntary. A mortgage becomes delinquent when the mortgage borrowers stop making payment. If enough payments are missing, we say the mortgage is in default. A mortgage, by definition, is collateralized by the property which the mortgage is used to purchase. When a mortgage defaults, the mortgage lender has the right to repossess and liquidate the underlying property to cover what the mortgage borrowers owe.
For agency MBS, the underlying mortgages are guaranteed by agencies such as Fannie Mae, Freddie Mac, and Ginnie Mae (through FHA, VA, etc.). If a mortgage in an agency MBS pool defaults, the agencies will step in and make timely payments of principal and interest to the MBS investors. And if enough missing payments are accumulated, the agencies will buy the default mortgage out of the MBS pool. From MBS investors’ perspective, prepayment due to default is no different from other types of prepayment.
For non-agency MBS or private label securities, the investors have to wait until the underlying property of the default loan is liquidated to receive what they are owed, in which case they may end up with a loss. In a PLS, the loss will be absorbed by the subordinate tranche first and work its way up the tranche structure.
Mortgage default is heavily driven by credit-related factors, such as borrowers’ credit score, loan-to-value ratio, etc. For non-agency MBS, we also need to estimate LGD to predict the loss.
The total prepayment is the sum of these components. To learn more about Plutus Analytics prepayment model or other models, please contact the customer support team at sales@plutusanalytics.com.
©Plutus Analytics 2022
No part of this article may be reproduced, distributed, stored in a retrieval system, or transmitted in any form or by any means, without the prior written permission of Plutus Analytics, except in the case of brief quotations embodied in reviews and certain other non-commercial uses permitted by copyright law.