Balancing price upside and downside in CRT

admin | April 17, 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.

The market in credit risk transfers has gone on a roller coaster ride since mid-February with big ups and downs in prices and changes in the outlook for residential credit. Investors can still get a handle on relative value by framing a base case, a bullish case and a bearish case for CRT. By looking at plausible security prices in those instances, a clear picture emerges of securities that have much more upside than downside. Seasoned CRT that absorbs actual rather than fixed losses comes out looking the best.

Prices on the longest and last cash flows of CRT have rallied meaningfully in recent weeks, but the universe of those bonds still trades below par. These discount dollar price likely reflect changes to both financing and fundamentals. As available financing for the asset class has dried up, investors have likely required more spread compensation to meet their unlevered return bogeys. Discount dollar prices also certainly reflect uncertainty and new fundamental risks from the continued effects of COVID-19 on the economy. While it’s impossible to peg what virus-related delinquency and default rates will be, reasonable assumptions can frame baseline, bullish and bearish possibilities. And looking at the relative performance of these bonds to those scenarios may serve as a reasonable way to frame relative value.

 Pricing base case, bullish and bearish possibilities

One thing we can say with certainty is that COVID-19 is already having an impact on borrower performance. Business disruption and loss of income has driven affected borrowers to reach for payment forbearance offered by the enterprises, which will, in turn, drive delinquency rates higher on these deals. And some of those loans will in all likelihood roll to default. In addition to delinquencies and defaults, the effects of the virus could put downward pressure on regional or national home prices. Other factors also will influence fundamental performance in CRT including prepayment rates, incidence and type of permanent modification and a handful of others. These three can be used to begin to construct base, bullish and bearish scenarios

Given the current velocity of applications for forbearance from GSE borrowers, the base case analysis assumes a monthly 3% rise in delinquency rates for the next six months. It also assumes implied forward interest rates and corresponding prepayments. From there, current market prices correspond to a spread for each bond, and this analysis can use that spread to solve for a price in scenarios that assume bullish lower delinquency rates or bearish higher delinquency rates coupled with a decline in home prices to the downside.

The pricing in base, bullish and bearish scenarios helps produce a pair of important measures: the percentage price gain in the bullish case, and the percentage price loss in the bearish case. For example, the price of a CRT might rise by 3% in the bullish case and drop by 1% in the bearish case. That security would have an upside-to-downside ratio of 3-to-1, or 300%.  That would look more attractive than a second security that might have 2% upside and 1% downside, or a ratio of 200%.

Exhibit 1: Evaluating price convexity across CRT LCF

Note:  Pricing and curve as of COB 4/14/20 Analysis solves for a Discount Margin assuming a monthly 3% increase in delinquency rates and a 50% roll-to-default along a single base case assumption for both prepayments and HPA. Each bond’s DM is then used to solve for a price in the ‘bull’ case which assumes 1% DQ per month for six months with a 50% roll-to-default with no change to HPA assumptions and a ‘bear’ case where delinquencies spike to 4.5% per month for six months concurrent with an immediate 10% downward shock in home prices. Price convexity is calculated as [(% price change bull)/(% price change bear)] -1. Deals that solved for a negative discount margin under the base line stress were excluded from the analysis. Source: Vista, Amherst Pierpont Securities

Using this approach, seasoned actual-loss deals fare better than others for a few reasons (Exhibit 1). First, 2015 through 2017 vintages have deleveraged fairly substantially from the perspective of both the borrower and the structure. The steady rise in HPA since issuance has built substantial borrower equity, which should protect bond holders from losses against the backdrop of moderate home price stresses. Both high- and low-LTV seasoned deals appear to offer better price performance relative to other cohorts. In fact, seasoned high LTV last cash flows like the M3 classes of 2015 vintage STACR HQA deals offer some of the most attractive relative price performance. Additionally, certain 2018 vintage CAS high LTV deals appear to offer attractive relative price performance likely driven by lower loss severities as a function of deep MI coverage coupled with some seasoning and deleveraging. By contrast, newer vintage actual loss deals and earlier vintage fixed severity deals fare worse under this framework. The market seems to be relatively enamored with higher WAC 2019 last cash flows based on the view that they will continue to prepay faster than other cohorts. However, a steady increase in payment forbearance, frictions to refinancing from shelter-in-place orders and the tightening of credit overlays against the backdrop of an economic downturn all could materially suppress prepayment rates leaving these deals exposed to more fundamental tail risk than investors might otherwise anticipate. Fixed severity deal also fare poorly under this methodology as seriously delinquent loans will become credit events after they are more than 180 days delinquent. At that point, loans get removed from the pool with a fixed severity that will increase based on the number of total credit events. Some fixed severity deals traded to a negative discount margin to the base case delinquency scenario and were not run to the bullish and bearish scenarios. As it stands, loans in forbearance will be treated as delinquent for the purposes of calculating credit events in fixed severity deals and delinquency trigger calculations in all CRT. And while investors have petitioned for those loans to get the same treatment that they would get in the case of a natural disaster, FHFA has remained silent on any potential resolution to date.

With the caveat that assuming ‘bull’ and ‘bear’ cases may in reality not be equally likely, relative price performance can help frame relative value across the universe of CRT last cash flows. While dollar prices remain somewhat depressed on these assets, a majority of that may just reflect new fundamental risks to the asset class. Setting a new normal and evaluating relative price performance in better or worse scenarios should help re-frame the CRT market.