# MBS / Agency Mortgage Resources — Block 011 reference

Live data and resource stack for the MBS-Treasury basis / prepayment risk topic area. Updated 2026-07-06 (Block 011 main material).

## Current state (2026-07-06)

| Series | Level | Source | Use |
|---|---|---|---|
| 30y mortgage rate (Freddie PMMS) | 6.43% (2026-07-02) | freddiemac.com/pmms | Primary-secondary spread denominator |
| 10y UST yield | 4.48% (2026-07-01) | FRED H.15 | Treasury comparison for UMBS hedging |
| 30y UST yield | 4.97% (2026-07-01) | FRED H.15 | Same-tenor Treasury comparison (preferred for 30y UMBS) |
| Current 30y UMBS TBA coupon | ~5.58% (implied) | MBS pricing desks | Current production TBA coupon |
| 30y mortgage → 30y UST gross spread | +146bp | derived | Primary-secondary, full picture |
| Current coupon → 30y UST gross spread | +61bp | derived | Secondary market, option-unadjusted |
| Net basis (after g-fee + servicing) | negative (option-overpriced) | derived | Trades < fair value = rich market |
| 5.5% coupon pool PSA speed | ~200-250 PSA | UMBS dealer reports | Current environment refi wave indicator |

## Prepayment thresholds (recall)

| Speed | Read |
|---|---|
| 100 PSA | Baseline (6% CPR), no refi wave |
| 150-200 PSA | Mildly elevated, some refi incentive |
| 300-500 PSA | Clear refi wave |
| 500-800 PSA | Aggressive refi (2016-2017, late 2020) |
| 800+ PSA | Peak refi (2003, 2020) |

## Free data sources (Tier 1)

| Source | Coverage | URL | Cost |
|---|---|---|---|
| **Freddie Mac PMMS** | 30y/15y mortgage rate, weekly | https://www.freddiemac.com/pmms | Free |
| **FRED H.15** | Treasury yields (full curve), daily | https://fred.stlouisfed.org/series/DGS10 | Free |
| **UMBS prepay data (eNAMIC)** | Current-coupon PSA speeds | https://www.enhancemortgage.com | Free with registration |
| **MBA Weekly Applications Survey** | Mortgage application index, refi share | https://www.mba.org/news-and-research | Free |
| **Fannie Mae / Freddie Mac** | Pool-level disclosure, prospectuses | https://www.fanniemae.com, https://www.freddiemac.com/mbs | Free |
| **Treasury.gov daily yield curve** | Full curve 1mo-30y | https://home.treasury.gov/resource-center/data-chart-center/interest-rates | Free |

## Institutional data (Tier 3)

| Source | Coverage | Typical cost |
|---|---|---|
| **Bloomberg MBS function** | Live TBA prices, current coupon, specified pool payups, PSA speeds | Bloomberg terminal required |
| **Intex cashflow engine** | Pool-level cashflow modeling, OAS, option-adjusted spreads | $50-100k+/year |
| **Bloomberg PREPAY** | Historical PSA by cohort, custom analytics | Bloomberg terminal |
| **eMBS (Empirica)** | Specified pool payup, current coupon, refi incentive | $5-15k/year |
| **Mortgage Analytics / Wall Street Analytics** | Pool-level analytics, OAS, convexity | Enterprise pricing |

## Books and primers (Tier 5)

- **Fabozzi "The Handbook of Mortgage-Backed Securities"** — comprehensive, encyclopedic
- **Hull Ch. 24 (Securitization)** — vocabulary orientation (Block 011 pre-read)
- **SIFMA MBS primer** — trade-association perspective
- **Fannie Mae MBS Basics** — agency perspective on structure

## Glossary of terms (Block 011 vocabulary)

| Term | Definition | Block 011 probe |
|---|---|---|
| **Pass-through** | Single-class security with pro-rata share of underlying mortgage pool cash flows | Q1 |
| **Pool factor** | Current outstanding / original face (stock measure) | Q2 |
| **WAC** | Weighted average coupon of pool; drifts with selective prepays | Q3 |
| **WAM** | Weighted average maturity; extends on no-prepay, contracts on refi | Q4 |
| **PSA** | 100 PSA = 6% CPR baseline, 0.2%/month ramp month 1-30 | Q5 |
| **CPR** | Annual conditional prepayment rate (the flow measure PSA proxies for) | (referenced in Q2) |
| **TBA** | To-be-announced, generic forward-settling MBS — the maximally prepayment-exposed benchmark | Q1 + Q8 |
| **Specified pool** | Pool with specific characteristics (high WAC, low WAM, geography, balance) that prepay slower than TBA; trades at payup | Q1 + Q4 (reframed) |
| **Negative convexity** | MBS price-yield curve is concave (bowed *away* from investor) — underperforms Treasury in both directions | (main material) |
| **Prepayment option** | Free American call held by borrower, strike = par; MBS investor short the option | (main material) |

## Cross-references

- Block 010 (yield vol / swaption surface) — agency MBS option-adjusted spreads use vol-surface inputs; the rate-cut-probability framework in Block 010 also drives mortgage rate forecasting
- Block 012 (SOFR-OIS-repo mechanics) — the SOFR-OIS basis is the cousin of the MBS-Treasury basis; both are signal layers for bank-funding stress and rate path expectations
- `~/Obsidian-Macro/03-research/deep/daily-vol-routine.md` — Block 010 5-minute daily routine (vol surface)
- `~/Obsidian-Macro/03-research/deep/daily-mbs-routine.md` — Block 011 5-minute daily routine (MBS signals)
- Block 009 (carry & roll-down) — MBS carry dynamics are dominated by prepayment, not just rolldown
- `~/Obsidian-Macro/00-methodology/risk-framework.md` v0.3 — for size and stop on basis trade
- `~/Obsidian-Macro/00-methodology/anti-patterns.md` — "narrative without pricing" applies to basis trades (e.g., "I think mortgages will underperform" without specifying convexity breakeven)

## Anti-patterns surfaced in Block 011

1. **Narrative-without-pricing on the basis.** "MBS will underperform" without specifying convexity-adjusted breakeven. The breakeven in PSA-speed terms (refi wave doubling prepays) is the tradeable question.
2. **Quoting off-the-run coupon as "current."** The 5.0% pool is legacy 2020-2021 vintage, deep out-of-the-money, prepay-impaired. The current production TBA coupon (~5.58%) is the right reference.
3. **Treating MBS as a Treasury substitute.** MBS is short a prepayment option, has negative convexity, requires PSA-adjusted analytics. The duration number alone misleads.
4. **Confusing pass-through and CMO.** CMO/REMIC = multi-class waterfall; pass-through = single class pro-rata. Different risk profiles, different analysis.
5. **Conflating extension and contraction risk.** Both are prepayment risk. Extension is the bigger risk for most holders (long-duration profile). Contraction is the bigger risk for negative-carry carry trades.

## Update protocol

- Refresh "Current state" table weekly after PMMS release (Thursdays) and FRED H.15 update (daily)
- Update PSA speed estimates when 5.5% coupon cohort's prepay behavior shifts materially
- Add new free data sources as discovered (this is the practical-resource-list pattern from Block 010)