Estimated study time: 60 minutes
Content:
Mortgage-backed securities (MBS) are fixed income instruments whose cash flows are derived from pools of mortgage loans. At CFA Level 2, the primary focus is on agency MBS (backed by Fannie Mae, Freddie Mac, or Ginnie Mae), non-agency residential MBS (RMBS), and commercial MBS (CMBS). The defining analytical challenge of MBS is prepayment risk — the right of mortgage borrowers to repay their loans early. Prepayments return principal to investors earlier than scheduled, which is harmful when interest rates have fallen (because investors must reinvest at lower rates) and beneficial when rates have risen (because investors get principal back at par in a declining-price environment). This optionality is the fundamental difference between MBS and bullet bonds.
The prepayment process is measured using two conventions. The Single Monthly Mortality (SMM) rate is the monthly prepayment rate as a fraction of the outstanding balance: SMM = prepayments in month / (beginning balance - scheduled principal). The Conditional Prepayment Rate (CPR) is the annualized equivalent: CPR = 1 - (1 - SMM)^12. The Public Securities Association (PSA) benchmark assumes prepayments ramp up from 0% in month 1 to 6% CPR over 30 months, then remain at 6% CPR. A pool prepaying at 150% PSA is prepaying at 1.5 times the PSA standard model (more aggressive); at 50% PSA, it is prepaying half as fast (slower). PSA provides a common language for comparing prepayment behavior across MBS pools.
The pricing of MBS requires accounting for the embedded prepayment option. Because borrowers hold a refinancing option, MBS cannot be priced using standard discount bond formulas — the cash flows are uncertain. The option-adjusted spread (OAS) is the spread over the benchmark curve after stripping out the value of the embedded option. OAS = Z-spread minus the option cost. For a callable bond or MBS with embedded options: OAS reflects the risk premium after paying for the optionality. The Z-spread is the constant spread over the spot curve that equates the present value of scheduled (optionless) cash flows to the market price. For MBS, the Z-spread is always larger than OAS because the Z-spread does not account for prepayment optionality — it would overestimate the true risk premium if used naively.
MBS structural features create different securities from a single pool of mortgages (the collateral). Sequential-pay CMOs (Collateralized Mortgage Obligations) divide the pool's principal payments into tranches that receive payments sequentially — earlier tranches receive all principal until paid off, then the next tranche receives principal, and so on. This creates a front tranche with shorter effective duration (lower prepayment risk) and a later tranche with longer effective duration. PAC (Planned Amortization Class) bonds and Companion (Support) bonds are another CMO structure: PAC tranches receive principal within a predefined schedule over a range of PSA speeds (the PAC band), while Companion tranches absorb all prepayment variability outside the PAC band. PAC bonds have more predictable cash flows; Companions have high prepayment and extension risk.
Commercial MBS (CMBS) are backed by pools of commercial real estate loans — office buildings, retail centers, hotels, multifamily properties. CMBS differ from residential MBS in several important ways: commercial loans are typically non-recourse (lender's only remedy is the property), most CMBS loans have balloon payments at maturity (creating refinancing risk — the risk that the loan cannot be refinanced), prepayment is often restricted by lockout periods and yield maintenance provisions that make voluntary prepayment very costly. Key metrics for CMBS analysis include: loan-to-value (LTV) ratio (loan amount / property value), debt service coverage ratio (DSCR = property net operating income / annual debt service), weighted average LTV across the pool, and the concentration of loans in specific property types or geographic markets.
Key Terms:
Quiz Questions:
Q1. A mortgage pool has a beginning outstanding balance of $200M. In the current month, scheduled principal payments are $500,000 and actual prepayments are $1,500,000. Calculate the SMM and CPR.
A) SMM = $1,500,000 / $200,000,000 = 0.75%; CPR = 1 - (1-0.0075)^12 = 1 - 0.9140 = 8.60%. B) SMM = $1,500,000 / ($200,000,000 - $500,000) = $1,500,000 / $199,500,000 = 0.752%; CPR = 1 - (1-0.00752)^12 ≈ 8.64%. C) SMM = ($1,500,000 + $500,000) / $200,000,000 = 1.0%; CPR = 1 - (0.99)^12 = 11.36%. D) SMM = $500,000 / $200,000,000 = 0.25%; CPR = 2.97%.
Answer: B — SMM = Prepayments / (Beginning Balance - Scheduled Principal) = $1,500,000 / ($200,000,000 - $500,000) = $1,500,000 / $199,500,000 = 0.00752 = 0.752%. CPR = 1 - (1 - SMM)^12 = 1 - (1 - 0.00752)^12 = 1 - (0.99248)^12 ≈ 1 - 0.9136 ≈ 8.64%. Scheduled principal is subtracted from the beginning balance in the denominator because SMM measures voluntary prepayments as a fraction of the pool that could have prepaid (not the total beginning balance).
---
Q2. A 30-year residential MBS pool has an average coupon of 6.5% and is priced at 105% of par (at a premium). The pool is currently prepaying at 200% PSA. If interest rates decline further and PSA speeds increase to 350%, what is the impact on the MBS investor?
A) Higher prepayments increase value because investors receive principal faster. B) Higher prepayments are harmful to this premium-priced MBS investor — faster prepayments shorten the effective life of the bond, reducing the investor's ability to earn the above-market 6.5% coupon for as long as expected. This is contraction risk (faster-than-expected prepayments on a premium MBS). C) Higher prepayments have no impact because the investor receives par on all principal returned. D) The investor benefits because higher PSA reduces extension risk.
Answer: B — For a premium-priced MBS (price > par), the investor paid above par to receive the above-market coupon for the full expected term. If prepayments accelerate, principal is returned at par — not at the premium price — truncating the high-coupon cash flows earlier than expected. The investor must reinvest returned principal at the now-lower current market rates. This is contraction risk (the opposite of extension risk). Negative convexity of MBS arises precisely from this dynamic: when rates fall and you'd normally expect a bond price to rise significantly, MBS price appreciation is limited because the expected call (prepayment) shortens duration.
---
Q3. Two MBS tranches are described: Tranche A (PAC bond) and Tranche B (Companion bond) from the same CMO structure. The PAC band is 100%-250% PSA. If prepayments run at 75% PSA (below the band), which tranche bears more risk and why?
A) Tranche A (PAC) bears more risk because it has lower yield. B) Tranche B (Companion) bears the excess extension risk when prepayments fall below the PAC band — when prepayments are slow, Companion tranches extend further than expected while the PAC tranche's schedule is still met from the Companion's support. C) Both tranches bear equal risk because they come from the same collateral pool. D) Tranche A bears more risk because PAC bonds are always higher duration.
Answer: B — The Companion (Support) tranche is specifically designed to absorb prepayment variability outside the PAC band. When prepayments fall below the band (75% PSA vs. minimum band of 100% PSA), there is less principal coming in than the PAC schedule requires. In practice, the PAC schedule is still protected by having the Companion tranche extend in duration (receive even less principal than normally scheduled), while the PAC tranche continues to receive its planned amortization. The Companion thus bears extension risk when prepayments are slow. When prepayments are very fast (above the upper band), the Companion bears contraction risk.
---
Q4. A CMBS pool's underlying commercial mortgage has a DSCR of 1.10x and an LTV of 85%. An analyst is assessing this loan's credit risk. How should she interpret these metrics?
A) DSCR of 1.10x and LTV of 85% indicate a high-quality commercial loan suitable for investment-grade CMBS. B) DSCR of 1.10x is low (minimal cushion above debt service — standard investment-grade requires 1.25x or higher), and LTV of 85% is elevated (standard is 65-75%). Together, these metrics indicate an aggressive underwrite with limited cushion for property income decline and limited equity cushion in the property. This loan would be consistent with a lower credit quality CMBS tranche. C) Only DSCR matters for CMBS; LTV is irrelevant. D) An 85% LTV is acceptable for retail properties with long-term leases.
Answer: B — Both metrics indicate elevated credit risk. Investment-grade CMBS loans typically target DSCR of 1.25x or higher (providing buffer for income decline before debt service is impaired) and LTV of 65-75% (providing equity cushion for property value declines before lenders suffer losses in foreclosure). A DSCR of 1.10x leaves only 10% cushion — a modest income decline breaches debt service. An LTV of 85% means lenders would take losses if the property value falls more than 15%. This combination would suggest a subordinated CMBS tranche rating (BB or below) rather than an investment-grade rating.
---
Q5. The OAS on an agency MBS is 25 bps and the Z-spread is 55 bps. What does this imply about the cost of the embedded prepayment option, and which metric is more appropriate for comparing the risk premium of this MBS to a comparable corporate bond?
A) Option cost = Z-spread - OAS = 55 bps - 25 bps = 30 bps. The OAS (25 bps) is the appropriate risk premium comparison to a corporate bond, because it has removed the embedded option value from the spread. B) The OAS and Z-spread should be averaged: (55+25)/2 = 40 bps is the correct comparison. C) The Z-spread (55 bps) is the better comparison because it is higher and therefore more conservative. D) Option cost = OAS - Z-spread = 25 - 55 = -30 bps, implying negative option cost.
Answer: A — Z-spread = OAS + Option Cost. Option Cost = 55 - 25 = 30 bps. This 30 bps represents the value of the prepayment option owned by mortgage borrowers — it is a cost to MBS investors. The OAS (25 bps) is the appropriate "apples-to-apples" comparison with a corporate bond's credit spread because both measures represent the risk premium above the benchmark after accounting for embedded optionality. Using the Z-spread (55 bps) would overstate the risk premium because it includes the option cost. The OAS framework makes MBS comparable to option-free securities on a risk-adjusted basis.
---