What it means
A mortgage is a 30-year promise that borrowers keep for about seven to twelve years on average. People move, refinance, and pay off early, and every early payoff changes the cash flows of anyone holding the loan.
Mortgage-backed securities pool thousands of those loans, so their value hinges on predicting the speed of early repayment. Prepayment models are the machines that make the prediction.
The industry's reference convention is the PSA model, created by the Public Securities Association, now part of SIFMA. Its standard formulas define 100% PSA as prepayment rates starting at 0.2% CPR in the first month, rising 0.2% each month until reaching 6% annually from month thirty.
Actual speeds are quoted as multiples of that curve: 150 PSA means one and a half times the standard speed, and dealers compare realised prepayments against it constantly. The single biggest driver is interest rates.
When rates fall, refinancing surges and prepayments accelerate, which is why a mortgage security behaves unlike a normal bond: falling rates shorten its life just when the holder wants the high coupons to last. Other inputs layer on: loan age, season, home price moves, loan size, and borrower credit, each shifting the speed in ways modern models fit on mountains of loan-level data.
Getting the model wrong is expensive. Prepayment surprises reprice trillions of dollars of mortgage securities, and entire hedging programmes rest on the model's assumptions.
For a non-finance reader, the prepayment model is the weatherman of the mortgage market: the forecast is always wrong in detail, but everyone with money at stake must still plan around it. Sophisticated models go further than the PSA convention, fitting borrower behaviour on millions of loans.
They capture burnout, the way a pool's refinancing response weakens after the most rate-sensitive borrowers have already left. Hedging depends on the model's output as much as trading does.
The duration of a mortgage security is a model artefact, changing every time prepayment expectations change, so risk managers re-run the models nightly.
In practice
Real-world examples.
Example
A dealer quotes a mortgage pool at 180 PSA after a rate rally, pricing in faster refinancing than the standard curve. The multiple is shorthand; the underlying bet is on borrower behaviour. If rates then rise, the same pool may be re-quoted at 100 PSA or slower.
Example
An investor paying 104 for a mortgage security loses money when prepayments accelerate and principal returns at 100 years earlier than modelled. On $1,000,000 of face value, the $40,000 premium paid above par is only recovered through coupons, and faster repayment cuts the time those coupons are received.
Example
Winter slows home sales, and seasonal adjustments in the model trim expected prepayment speeds for the quarter. The desk expects speeds to rise again in spring, so it does not change its longer-term view. The adjustment is small compared with the effect of a one-point move in mortgage rates.
Formula
Calculation
CPR, the conditional prepayment rate, is the annualised percentage of the remaining pool expected to prepay. Under 100% PSA, CPR starts at 0.2% in month one, rises 0.2% monthly, and plateaus at 6% from month thirty; multiples scale the whole curve. The single monthly mortality rate (SMM), the share of the pool prepaying in one month, is SMM = 1 - (1 - CPR)^(1/12).
Worked example: in month 12, the 100% PSA curve gives CPR = 12 x 0.2% = 2.4%. At 150 PSA, CPR = 1.5 x 2.4% = 3.6%. Then SMM = 1 - (1 - 0.036)^(1/12) = about 0.305%, so on a pool with a $200,000,000 balance, expected prepayments that month are about 0.305% x $200,000,000 = $610,000. If speeds jump to 320 PSA, CPR = 3.2 x 2.4% = 7.68%, SMM is about 0.664%, and expected prepayments rise to about $1,327,000, more than double. That doubling of principal returned at par is exactly the surprise that hurts an investor who paid a premium.Case study
Seen in the real world.
This case study is fictional and illustrative. A made-up asset manager buys a mortgage-backed security yielding 5.5%, priced at 150 PSA, expecting the premium it paid to amortise slowly over a decade of high coupons. Six months later, rates fall sharply and refinancing waves through the pool; realised speeds jump to 320 PSA. The consequences arrive fast: the expensive premium is repaid at par within two years instead of ten, crystallising a loss on the price paid above par, and the portfolio's yield collapses as the high-coupon loans vanish.
The fund's hedges, calibrated to the old model speeds, over-hedge the now-shorter security and bleed money as rates keep falling. The post-mortem finds no villain, only a reminder: in mortgage land, the borrower's right to refinance is the investor's central risk, and the prepayment model is the map of that risk, redrawn every time rates move. The manager now stress-tests every purchase at both half and double the expected speed before buying.
Watch out
Common mistakes.
- Valuing mortgage securities like ordinary bonds; prepayment risk means their cash flows reshape themselves every time rates move.
- Assuming historical speeds persist; prepayment behaviour shifts with rates, housing markets, and lending standards faster than backtests suggest.
- Paying a premium for mortgage paper without modelling faster prepayment, since early repayment at par punishes anything bought above it. Discount purchases carry the mirror-image benefit.
Questions
People also ask.
What is a prepayment model?
A statistical model forecasting how quickly mortgage borrowers repay early, used to value and hedge mortgage-backed securities.
What does 100 PSA mean?
The industry-standard baseline prepayment curve, rising from 0.2% CPR in month one to 6% annually from month thirty, with actual speeds quoted as multiples of it.
What drives prepayments most?
Interest rates: falling rates trigger refinancing waves that accelerate prepayments, while rising rates slow them to a crawl. Housing turnover and loan age adjust the baseline too.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%