| Friday, January 24, 2025 |
||||||||
|---|---|---|---|---|---|---|---|---|
|
MANAGING DIRECTOR: |
||||||||
|
US Treasury Market |
|||||||||||
| Date | 1 mo | 3 mo | 6 mo | 1 yr | 2 yr | 3 yr | 5 yr | 7 yr | 10 yr | 20 yr | 30 yr |
| 01/16/25 | 4.30 | 4.30 | 4.30 | 4.17 | 4.23 | 4.30 | 4.40 | 4.51 | 4.61 | 4.93 | 4.86 |
| 01/17/25 | 4.31 | 4.30 | 4.31 | 4.21 | 4.29 | 4.35 | 4.43 | 4.53 | 4.63 | 4.93 | 4.86 |
| 01/21/25 | 4.31 | 4.31 | 4.31 | 4.19 | 4.28 | 4.32 | 4.40 | 4.49 | 4.58 | 4.89 | 4.81 |
| 01/22/25 | 4.31 | 4.32 | 4.32 | 4.20 | 4.29 | 4.35 | 4.45 | 4.53 | 4.65 | 4.94 | 4.87 |
| 01/23/25 | 4.31 | 4.32 | 4.30 | 4.18 | 4.30 | 4.35 | 4.44 | 4.55 | 4.61 | 4.90 | 4.83 |
The data in the table above is static as of the time it was pulled, so rates may have changed. Treat all data in this table and PMR as indications only and availability is always subject to change. This information was pulled manually from sources we believe to be reliable. New source, as of 12/12/2022, Bloomberg, L.L.P. As of: close of business 1/23/2025.
Mortgage Prepayment Models (The Kitchen Sink of Options)
We often have a conversation about which prepayment model or which way to look at a single-family MBS pool is “correct”. For the sake of this discussion, we are going to focus on fixed rate pools, not hybrid arms or floaters, with a fixed coupon. We find that all models are a helpful tool, some more than others, but understanding how they derive their prepayment forecast is important.
Variables That Drive Yield & Returns for Fixed Rate MBS
1) Price you paid (book price)
2) Coupon (fixed when you buy)
3) Prepayments (typically quote in CPR)
Both the price and coupon are fixed and known when you buy a bond, however the prepayments are unknown and typically forecasted. This prepayment forecast will change over the life of the bond and will change with the overall economy and interest rate cycle.
Variables That Drive Yield & Returns for Fixed Rate MBS
1) Price you paid (book price)
2) Coupon (fixed when you buy)
3) Prepayments (typically quote in CPR)
Both the price and coupon are fixed and known when you buy a bond, however the prepayments are unknown and typically forecasted. This prepayment forecast will change over the life of the bond and will change with the overall economy and interest rate cycle.
The above models all attempt to forecast a prepayment speed for a single-family MBS pool. The two most prevalent models used by community bankers are Bloomberg Median and BAM. We have found over time that the BAM model is the “best” model as it will incorporate pool level details such as loan size.
Bloomberg Median assumes that a specific pool is similar to the cohort average (doesn’t adjust for specific pool attributes). This at times can drastically over or understate prepayment forecast which impacts the pool’s yield. At times, we find the contributors to be stale or one contributor can throw off Bloomberg Median if we have limited contributors and one of the contributors is an outlier.
CPR/YTH/PSA models are all good gut checks a user can run against a BAM or Bloomberg Median to understand how yields may vary at various CPR shocks. In an ideal world we will find a pool with a consistent YTH (historical yield) that closely syncs to the BAM model and typically the Bloomberg Median.
The day we purchase a fixed rate MBS pool we tell every client, no matter the model, it will not pay exactly as forecasted. No model is perfect. The best rule of thumb to decrease yield volatility is to have a predefined dollar price threshold which helps to smooth the yield profile. If you buy between $97 to $103, typically the impact of CPR changes is more muted. If you never want to have yields move buy the MBS at par and your yield = coupon.
We generally discourage >$103-dollar prices as the volatility from a small change in CPRs can be high and determinately impact budgets. Deep discount, <$97-dollar price, we do advocate as discounted callable or bullet alternatives, but would strongly suggest shocking a slow CPRs (4-6 CPR) to make sure the slower CPR assumption is still a good value. Don’t buy purely on BAM or Bloomberg Median base case for deep discount MBS pools. Typically, we find over the life of a bond it will not pay less than 6 CPR so a conservative 4-6 CPR we view as a worst case if you held to maturity for deep discounts.
Please reach out to your Country Club Bank representative should you have questions or wish to discuss these models further.
CPR/YTH/PSA models are all good gut checks a user can run against a BAM or Bloomberg Median to understand how yields may vary at various CPR shocks. In an ideal world we will find a pool with a consistent YTH (historical yield) that closely syncs to the BAM model and typically the Bloomberg Median.
The day we purchase a fixed rate MBS pool we tell every client, no matter the model, it will not pay exactly as forecasted. No model is perfect. The best rule of thumb to decrease yield volatility is to have a predefined dollar price threshold which helps to smooth the yield profile. If you buy between $97 to $103, typically the impact of CPR changes is more muted. If you never want to have yields move buy the MBS at par and your yield = coupon.
We generally discourage >$103-dollar prices as the volatility from a small change in CPRs can be high and determinately impact budgets. Deep discount, <$97-dollar price, we do advocate as discounted callable or bullet alternatives, but would strongly suggest shocking a slow CPRs (4-6 CPR) to make sure the slower CPR assumption is still a good value. Don’t buy purely on BAM or Bloomberg Median base case for deep discount MBS pools. Typically, we find over the life of a bond it will not pay less than 6 CPR so a conservative 4-6 CPR we view as a worst case if you held to maturity for deep discounts.
Please reach out to your Country Club Bank representative should you have questions or wish to discuss these models further.
This information is intended for institutional investors only. The material provided in this document/presentation is for informational purposes only and is intended solely for private use. Past performance is not indicative of future results. This material is not intended as an offer or solicitation for the purchase or sale of any financial instruments.
•Not FDIC Insured •No Bank Guarantee •May Lose Value