Uncategorized
Issuer calls come home to roost in non-QM MBS
admin | March 6, 2020
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.
Investors in non-QM MBS will likely see $400 million in par principal returned to them this month and more than $2 billion in the months to come from issuers executing calls embedded in the trusts. The loans backing virtually all non-QM trusts currently trade well above par, and low market rates and tight liability spreads on new non-QM deals have likely created significant incentive for sponsors to execute.
Based on current market levels and a few reasonable assumptions, the entire universe of non-QM collateral trades at least a three point premium to par, and all trusts look likely to be called at the first available date call if this rate environment persists or retraces modestly. (Exhibit 1)
Exhibit 1: Estimating the volume and timing of non-QM trust issuer calls
Source: Bloomberg, Amherst Pierpont Securities
Non-QM MBS trusts feature call options generally ranging from two to three years from the issuance date that allow issuers to collapse the trusts by executing a par priced call on the underlying loans. A rate drop, loan prices rise and the call option investors have sold to the issuer goes deeper into the money. Assuming bond spreads are not widening significantly at the same time, it creates significant incentive for the issuer to call the deal, potentially leaving investors with large amounts of cash to reinvest into lower rates.
Attempting to price out the universe of loans securitized in non-QM trusts requires a couple of key assumptions. Calculating the weighted average spread on liabilities of newly issued non-QM deals should provide a reasonable benchmark for pricing the loans backing existing non-QM trusts. Pretium’s DRMT 2020-1 deal priced on February 13, for example, with the issuer selling ‘AAA’ through ‘B’ rated bonds, achieving a 98.2% advance rate on the collateral. The deal’s high advance rate should provide a reasonable benchmark net of pricing on the first loss, interest only and residual classes. The calculation also needs to include any costs associated with securitization such was warehouse financing and transaction fees. The weighted average spread on the deal was slightly less than 125 bp on bonds sold (Exhibit 2).
Exhibit 2: Calculating the spread on the universe of non-QM collateral
Source: Bloomberg, Amherst Pierpont Securities
Pricing the collateral tranches of 80 outstanding non-QM trusts at a spread of 125 bp vgives a weighted average price of roughly $107-06 with all collateral trading above $103-00. It’s likely that twelve 2017 vintage trusts totaling nearly $400 in principal balance across the AOMT, DRMT and VERUS shelves may be called this month as the deals collectively have a weighted average collateral price of roughly $108-19. If the current rate environment persists, or even backs up modestly for that matter, 2018 trusts will begin to become callable in April.
The overwhelming majority of the non-QM universe appears susceptible to call even if rates rise. Assuming a spread on the collateral of 125 bp, roughly 75% of the collateral backing the non-QM universe has a price between $105 and $109 and given the relatively short effective duration of the collateral, it would take a meaningful retracement of interest rates before the loans would trade at a discount. (Exhibit 3) Certain trusts fully backed by investor loans have the highest estimated collateral prices, with 2018 and 2019 vintage VERUS-INV deals all pricing at $110 or higher. While these trusts will in all likelihood get called, investors may, at a minimum, experience slower prepayments until they get called as investor collateral has afforded non-QM investors better convexity than other stories in non-QM.
Exhibit 3: Distribution of collateral prices across the non-QM universe
Note: Prices as of 3/2/2020 COB. Source: Bloomberg, Amherst Pierpont Securities
The move lower in rates certainly looks to cap total return across the stack as bonds, especially those in deals with short dated calls, will in all likelihood not trade materially above par. This may push demand from short duration investors into front sequentials backed by collateral with some prepay protection like seasoned RPLs or agency-eligible investor loans, which will trade to longer durations and as a result would have better total return into a rally. Additionally, this move may drive investors in non-QM trusts to re-evaluate the requisite compensation associated with shorting a call to the issuer.