Pass-Through Prepayment Behavior

Pass-through securities channel principal and interest payments from underlying loans directly to investors—and prepayment behavior is the single largest source of cash flow uncertainty in these structures. When borrowers pay off loans early (through refinancing, home sales, or voluntary curtailment), the timing and magnitude of your cash flows shift in ways that standard yield calculations don't capture. For mortgage-backed securities alone, a 1% misjudgment in prepayment speed assumptions can move portfolio valuations by 50-150 basis points, depending on tranche position and remaining term.

TL;DR: Prepayment behavior determines when (not just whether) you receive principal back from pass-through securities. Understanding CPR, SMM, and PSA benchmarks—and stress-testing your assumptions—is the difference between accurate duration management and unexpected reinvestment risk.

What Pass-Through Prepayment Behavior Actually Means

A pass-through security pools loans (mortgages, auto loans, student loans) and distributes borrower payments to certificate holders on a pro-rata basis. Every monthly payment has two components: scheduled principal and interest (predictable from the amortization schedule) and unscheduled principal (prepayments). The second component is where the complexity lives.

Prepayment means any principal payment that arrives ahead of the original amortization schedule. This includes full payoffs (refinancing into a new loan, selling the property) and partial curtailments (extra principal payments). From your perspective as an investor, prepayments return your principal earlier than expected—which sounds harmless until you realize it forces reinvestment at potentially lower yields.

Why this matters: prepayments are not random. They cluster around rate environments, borrower demographics, and loan seasoning. A pool of 30-year fixed-rate mortgages with a 6.5% coupon will prepay at dramatically different speeds depending on whether prevailing rates are 5%, 6.5%, or 8%. Your job is to model these speeds accurately enough to price the security and manage duration.

Core Terms You Need to Know

Conditional Prepayment Rate (CPR): The annualized percentage of remaining principal expected to prepay in a given period. A CPR of 10% means that, over a full year, 10% of the outstanding principal balance would prepay (assuming the rate holds steady each month).

Single Monthly Mortality (SMM): The monthly equivalent of CPR. The conversion formula:

SMM = 1 − (1 − CPR)^(1/12)

For a CPR of 10%, this yields an SMM of approximately 0.874%—meaning roughly 0.874% of the remaining balance prepays each month.

PSA Benchmark (Public Securities Association): A standardized prepayment model that assumes prepayment rates ramp linearly from 0.2% CPR in month 1 to 6% CPR in month 30, then hold flat at 6% CPR for the remaining life. This is called 100% PSA. A pool running at 200% PSA doubles those speeds at every point along the curve (reaching 12% CPR by month 30).

The point is: these aren't just academic definitions. CPR, SMM, and PSA are the inputs that drive every cash flow projection, duration estimate, and relative value calculation for pass-through securities. Get them wrong, and your entire analysis breaks down.

How Prepayment Behavior Works in Practice (The Three Drivers)

Borrower prepayment decisions cluster around three forces, each with distinct characteristics and predictability.

1. Refinancing Incentives (Rate-Driven Prepayments)

This is the dominant driver for residential MBS. When market mortgage rates fall 100-150 basis points below a loan's coupon rate, borrowers have sufficient economic incentive to refinance (after accounting for closing costs, typically $3,000-$8,000). The result is a sharp, nonlinear increase in prepayment speeds.

The relationship is asymmetric. A 200 bps rate decline might push CPR from 8% to 25% or higher, but a 200 bps rate increase doesn't reduce CPR proportionally—it merely compresses prepayments toward a floor set by housing turnover and curtailments. This asymmetry is what makes pass-through securities behave differently from plain-vanilla bonds.

Why this matters: you face "negative convexity." When rates fall, your high-coupon principal comes back early (and you reinvest at lower rates). When rates rise, prepayments slow down and your capital stays locked in a below-market coupon for longer. You lose in both directions relative to a non-callable bond.

2. Housing Turnover (Economically Driven Prepayments)

Regardless of the rate environment, people move. Job relocations, divorces, upsizing, downsizing—these life events generate 15-25% of annual prepayments in a typical residential mortgage pool. This "turnover component" creates a baseline prepayment speed that persists even when refinancing incentives are zero.

Turnover-driven prepayments are more stable and more predictable than rate-driven prepayments. They correlate with employment conditions, regional housing markets, and borrower demographics (younger borrowers move more frequently than retirees, for example).

3. Curtailments and Partial Prepayments

Some borrowers make extra principal payments without fully paying off the loan. These curtailments are smaller individually but can add 2-5% CPR in aggregate, with seasonal patterns tied to tax refund season (February–April) and year-end bonuses. For auto loan and student loan pass-throughs, curtailments can represent a larger share of total prepayments than they do for mortgages.

The practical point: model all three drivers separately. A single CPR assumption blends these distinct behaviors into one number, which works until the rate environment shifts and the refinancing component moves independently of turnover. Separating them gives you better stress-test scenarios.

Worked Example: Pricing a $100 Million MBS Pass-Through

Here's how prepayment assumptions flow through to cash flows and yield. Walk through this carefully—it illustrates why small CPR differences create large valuation differences.

Your situation: You're evaluating a $100 million pass-through backed by 30-year fixed-rate mortgages with a weighted average coupon (WAC) of 5.75% and a weighted average maturity (WAM) of 348 months (the pool is 12 months seasoned). The pass-through rate to investors is 5.25% (the 50 bps difference covers servicing and guarantee fees).

Scenario A: 150% PSA (Moderate Prepayment Speed)

At 150% PSA, the pool reaches a plateau CPR of 9% by month 30 of the original loans (month 18 from your purchase, since the pool is already 12 months seasoned).

At this speed, roughly half your principal returns within the first 8 years, and the security behaves like an intermediate-duration bond.

Scenario B: 300% PSA (Fast Prepayment Speed)

Now assume rates drop 150 bps after purchase, triggering a refinancing wave. CPR jumps to 18% at plateau.

The WAL compresses from 8.2 years to 4.5 years. You're receiving principal back nearly twice as fast, and every dollar returned must be reinvested at rates that are now 150 bps lower. If you purchased the pass-through at a premium (above par, because you expected that 5.25% coupon in a lower-rate world), the accelerated return of par-value principal also generates a premium amortization loss.

Scenario C: 75% PSA (Slow Prepayment Speed)

Rates rise 100 bps after purchase. Refinancing incentive disappears, and prepayments slow.

Now your capital is locked in for an average of 13.1 years at a 5.25% coupon while new-issue pass-throughs offer 6.25%+. Your security trades below par, and your effective duration has extended precisely when you'd prefer shorter exposure.

The rule that survives: a single security can behave like a 4-year bond, an 8-year bond, or a 13-year bond depending on prepayment assumptions. This is why pass-through investors spend more time on prepayment models than on credit analysis—for agency MBS (where credit risk is guaranteed), prepayment risk is essentially the only risk.

Risks, Limitations, and Common Pitfalls

Pitfall 1: Using a Single Static CPR

The most common modeling error is plugging in one CPR and treating it as constant over the security's life. Prepayment speeds are path-dependent. A pool that seasons for 24 months in a stable rate environment will behave differently from one that experiences a rate shock in month 6. Use dynamic prepayment vectors (month-by-month CPR projections) rather than single-speed assumptions.

Pitfall 2: Ignoring Burnout Effects

After a sustained period of low rates, the borrowers most sensitive to refinancing incentives have already prepaid. The remaining pool is "burned out"—composed of borrowers who face credit issues, high loan-to-value ratios, or simply inertia. A pool at 250% PSA in year 1 of a rate decline might slow to 150% PSA in year 3 even if rates stay low. Failing to model burnout leads to overestimating prepayment speeds in later periods.

Pitfall 3: Conflating MBS and Non-Mortgage Pass-Throughs

Auto loan ABS, student loan ABS, and CLO structures have fundamentally different prepayment profiles. Auto loans rarely refinance (the savings don't justify the effort on a $25,000 loan), so prepayments are driven almost entirely by vehicle trade-ins and voluntary curtailments. Applying MBS-style prepayment models to auto loan pass-throughs will systematically overestimate rate sensitivity.

Pitfall 4: Not Stress-Testing the Tails

Your base case might assume 150% PSA, but what happens at 400% PSA or 50% PSA? These tail scenarios are where portfolio losses concentrate. Stress-test your pass-through holdings against at least three scenarios: base case, fast (2x base CPR), and slow (0.5x base CPR). For each scenario, recalculate WAL, effective duration, and reinvestment yield.

The point is: prepayment modeling isn't about getting the right answer—it's about understanding how wrong you can be and what it costs. The best practitioners don't predict prepayments more accurately; they build portfolios that perform acceptably across a range of prepayment outcomes.

Pitfall 5: Misreading Prepayment-to-Default Ratios in Lower Tranches

For structured pass-throughs (particularly CMBS conduits and CLOs), the interaction between prepayment speeds and default timing determines tranche performance. If prepayment rates exceed default rates by at least 2:1 in the collateral pool, senior tranches maintain adequate credit enhancement. But when that ratio compresses—defaults rising while prepayments slow during a recession—subordinate tranches with 5-10% thickness can breach loss triggers quickly.

Connecting to Related Structures

Pass-through prepayment behavior is the foundation for understanding more complex securitized products. Collateralized Mortgage Obligations (CMOs) exist specifically to redistribute prepayment risk—some tranches absorb faster prepayments (PAC companion bonds) while others are protected up to defined PSA bands (PAC bonds). If you understand why prepayment uncertainty matters for pass-throughs, you understand why CMO structuring exists.

Similarly, Commercial Mortgage-Backed Securities (CMBS) manage prepayment risk through lockout periods, defeasance requirements, and yield maintenance provisions. These structural features exist because commercial borrowers (unlike residential borrowers) face fewer barriers to refinancing large loans, and a single prepayment on a $50 million loan can dramatically alter pool cash flows.

Prepayment Analysis Checklist (Tiered)

Essential (High ROI)

These four steps prevent the most common analytical errors:

• Calculate CPR and SMM for your pool using the conversion formula, and verify against reported pool factors from the trustee
• Benchmark against PSA speeds—know whether your pool is running at 100%, 150%, or 250% PSA relative to the standard curve
• Identify the dominant prepayment driver for your collateral type (rate-driven for residential MBS, turnover-driven for auto ABS)
• Compare weighted average life under at least three prepayment scenarios (base, fast, slow) before committing capital

High-Impact (Systematic Process)

For investors managing multiple pass-through positions:

• Backtest your prepayment model against 10 years of historical SMM data from comparable vintages and coupons
• Monitor monthly pool factors and CPR reports to detect speed changes early (a 3-month moving average smooths noise)
• Map prepayment sensitivity to your portfolio's duration target—know which positions extend or compress most under rate shocks

Advanced (For Active Traders)

If you're actively trading pass-throughs on relative value:

• Model burnout effects by segmenting pools by seasoning, FICO distribution, and LTV bands
• Track the "refinancing efficiency" metric—what percentage of in-the-money borrowers actually prepay each month (historically 30-50% in the first year of incentive)
• Cross-reference SIFMA prepayment data and SEC Regulation AB II pool-level disclosures to validate model inputs against market-wide trends

Prepayment behavior is not something you solve once. It requires ongoing monitoring, model recalibration, and scenario discipline. The investors who manage it well don't predict better—they prepare better.