MBS: Relative value shifts in prime jumbo private MBS
admin | May 3, 2019
This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.
A drop in prime jumbo private MBS supply of nearly a third this year has driven spreads much tighter in ‘AAA’ front sequentials and pass-throughs to the point of looking rich to agency MBS. But for relative value investors, last cash flow sequentials have lagged the tightening and still offer attractive spread.
Spreads on prime jumbo super-senior pass-throughs have tightened sharply this year. The JPMMT shelf that in January priced pass-throughs $2-00 back of TBA, for instance, in April priced just $1-08 back of the FNCL 4.0% benchmark. The move marks the tightest nominal print for private 4.0% super-senior pass-throughs since the beginning of last year. It is also worth noting that when the April deal priced, agency jumbo pass-throughs traded just $1-04 back of TBA, leaving the priced difference between private and agency jumbo pass-throughs at only $0-04. The price basis between 4.0% private pass-throughs and agency jumbo pools has averaged roughly $0-24 over the past year and was as wide as $0-18 as recently as late March. (Exhibit 1)
Exhibit 1: Jumbo pass-throughs tighten to agency TBA and jumbo pools
Spread tightening has not been contained to pass-throughs. Super-senor front sequentials have seen pronounced spread tightening as well. Front sequentials are by and large quoted on a spread behind the price of their comparable coupon Fannie Mae 15-year pass-through. Front sequentials with 4.0% coupons traded as wide as $2-00 back of 15-year TBA and have tightened by $0-20, with the most recent new issue deal pricing $1-12 back of TBA (Exhibit 2)
Exhibit 2: Front sequentials tighten as the benchmark widens
One part of the cash flow structure that widened substantially into the end of last year and has remained wide is the last cash flow sequential. Those bonds traded to a spread of 100 bp over interpolated swaps this time last year, widened by 40 bps into the end of the year and have continued to trade in that context throughout this year. A relatively weak bid for duration is not unique to the private MBS market as last cash flow sequential CMOs have widened by roughly 15 bp since the end of the year and have remained at the wides, with spreads in the context of 105 over the interpolated Treasury curve. Private LCFs have widened significantly more than their agency counterparts, suggesting that they are not only the relatively inexpensive cash flow within the confines of private MBS but relative to agency CMOs, too. (Exhibit 3)
Exhibit 3: Private last cash flow sequentials offer significant relative value
Private prime jumbo supply and spreads likely to remain tight
Prime jumbo issuance in the first quarter of 2019 dipped to just $6.25 billion, comprising 30% of overall PLS issuance. This is a fairly significant drop from the first quarter of 2018 where prime jumbo volumes approached $10 billion and comprised 40% of total issuance. There are likely a number of factors that may be pushing issuance volumes lower. Banks have increased whole loan holdings by $50 billion over the past 12 months. It’s highly likely that a significant portion of those holdings are either conforming or non-conforming jumbo loans that could have otherwise been sold to the GSEs or gone to private execution. A slowdown in home price appreciation in high cost areas and recent changes to the tax code capping mortgage interest and property tax deductions have stifled demand for high balance loans.
The absolute level of conforming jumbo issuance from the GSEs is also down year over year. Through the first four months of last year, the GSEs pooled roughly $4 billion of conforming jumbo loans representing 1.6% of total originations. That number fell to $2.5 billion through the first four months of 2019 representing 1.2% of overall volumes. Holding the percentage of overall production constant year over year would suggest that conforming jumbo production should have been closer to $3.5 billion. It’s likely that the decreased share of overall production may have flowed into PLS. Even if all those loans flowed into non-agency deals, it would have only contributed an additional $1 billion and even with that, first quarter PLS production was off by a third year over year. Supply constraints should keep a lid on spreads. Given the historically tight relationship between pass-throughs and conforming jumbo pools, it seems unlikely that spreads on pass-throughs could rally materially absent a significant rally in spreads on conforming jumbo pools. Given the overall widening in the benchmark, front sequential spreads may have some room to tighten while last cash flow sequentials appear to offer significant relative when compared to PLS pass-throughs and front sequentials as well as comparable agency last cash flows.