Assessing MBS Coupon-Stack Value with Spread Curves

Summary

The document asks whether a standard parametric model exists for mortgage-backed securities coupon stacks, analogous to familiar term-structure curve models. The response reports no established model of that kind for the TBA coupon stack, and describes a practical market approach based on comparing option-adjusted and zero-volatility-adjusted spread curves across coupons, issuers, and maturities.

Traders can inspect these curves and compare them with seasoned loan pools to identify points that appear relatively rich or cheap. The answer notes that the curves can be irregular and that market valuation remains substantially judgment-based. It mentions general TBA references from several institutions but does not identify a formal coupon-stack model or present data, calibration details, or evidence that one approach consistently predicts value. The guidance is therefore descriptive of market practice, with limited support for treating it as a standardized quantitative framework.

Key ideas

• The response identifies no widely accepted parametric model for the TBA coupon stack.
• Market participants compare option-adjusted and zero-volatility-adjusted spread curves across coupons, issuers, and maturities.
• Seasoned loan pools can provide additional relative-value comparisons.
• Coupon-stack curves may be irregular, and valuation often relies on trader judgment.

Tags

# Are there any standard MBS coupon stack models?


# Are there any standard MBS coupon stack models?












I need to model MBS coupon stack prices. It would not be difficult to create something from scratch, but I don't want to re-invent the wheel (and explain why I did) if a somewhat standard model already exists. By "somewhat standard", I mean models like those for the term structure (e.g. Nelson-Siegel / Nelson–Siegel–Svensson or some newer stuff by Diebold et al).

I would appreciate any pointers, no matter how obscure.

## Answer by Tal Fishman (score 1, accepted)

https://quant.stackexchange.com/a/1485

I am not familiar with any formal "models" of the TBA coupon stack similar to Nelson-Siegel, but we typically compare the TOAS and ZV (Z-spread) curves as a function of coupon and issuer/maturity. We also examine relative value between these curves and various seasoned loan pools. The curve itself is very erratic, and plotting it often gives a good impression of which particular points on the curve may be over/under-valued. I have dealt with a number of market participants (traders) in this area over the years, and it seems like much is still done by guess-timating.

That said, there are a few references on TBAs generally by Citigroup, JPMorgan, the Fed and probably others. I couldn't find a reference to a model of the TBA coupon stack in any of these.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.