An early look at how forbearance is playing out in CRT
admin | April 30, 2021
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 is just starting to get cure and prepayment rates in Fannie Mae and Freddie Mac loans leaving forbearance, and the numbers look positive for CRT. More than half of previous seriously delinquent borrowers have cured or prepaid, a strong positive for credit fundamentals. But borrowers exiting forbearance sooner rather than later are more likely able to reperform, and cure rates on these pools could drop in the future.
Current faster cure rates could cause CRT principal windows to re-open sooner than anticipated on certain structures, particularly pre-REMIC actual-loss deals. The recent rally in CRT has been partially fueled by a repricing of CRT to slower prepayment assumptions. However, the extension that has been priced into premium bonds may be somewhat offset by delinquency triggers passing sooner than previously anticipated.
A look at early stage cure rates
More than half of Fannie Mae and Freddie Mac CRT loans greater than 60 days past due in June 2020 have subsequently either cured or prepaid as of the January remittance. Cure rates between the enterprises were substantially similar. Fifty-six percent of borrowers seriously delinquent in Freddie Mac CRT in June have subsequently cured or prepaid while 53.6% of those borrowers in Fannie Mae CRT have done so.
A look at the breakdown of cure and prepayment rates shows that just less than 13% of previously seriously delinquent borrowers have prepaid after exiting forbearance in both Fannie Mae and Freddie Mac CRT. Roughly 40% of borrowers in Freddie Mac CRT either self-cured or re-performed after their arrearages accumulated during forbearance were capitalized and deferred in non-accrual status. An additional 1.3% of borrowers cured through another form of loan modification.
Given slight reporting differences between the two agencies, a bit more insight can be gleaned from the manner in which loans in Fannie Mae CAS deals re-performed. Just less than 31% of previously seriously delinquent loans in Fannie Mae CRT cured through capitalization modification, and additional 8.8% self-cured with 1.2% of borrowers curing through another form of modification (Exhibit 1).
Exhibit 1: Cure and prepayment rates on previously DQ loans in Freddie Mac CRT
Thoughts on timing of cures
The overwhelming majority of borrowers who entered forbearance appear to have done so in April of last year. Using loans referenced in 2018 vintage Fannie Mae CAS deals as a proxy shows that just over $15 billion of UPB went into forbearance between March of last year and January of this year. Of that $15 billion, $8.6 billion, or 55% of that balance, went into forbearance in April of last year. After netting out prepayments, loans that rolled back to current and borrowers who took forbearance but continued to make payments, it is estimated that roughly $2.5 billion of those loans could exit forbearance plans in October of this year. (Exhibit 2)
Exhibit 2: Estimating timing and magnitude of borrowers exiting forbearance
While both Freddie Mac and Fannie Mae issued revised guidance that borrowers may be eligible for up to 18 months of forbearance, a 6-month extension over the originally proposed 1-year term, not all borrowers will qualify for that extension. For borrowers to be granted forbearance greater than one year, they must work with their servicer to establish Quality Right Party Contact to demonstrate that they are eligible for additional hardship forbearance. Given the fact that this was not required for the first year of forbearance as a result of the CARES Act, it seems likely that borrowers who are no longer in need of deferrals will have to resume making scheduled monthly payments in the near term as these plans reach their one year expiry.
Looking at cure rates by loan attributes
Somewhat surprisingly, cure and prepayment rates were generally consistent across loan attributes. Lower FICO loans have exhibited modestly lower cure rates than higher FICO ones but were still likely higher than many may have anticipated. Forty-four percent of loans seriously delinquent in June of last year with original FICO scores between 300 and 640 have subsequently cured or prepaid. By comparison higher FICO borrowers, those with original credit scores greater than 760, have cured at a rate of 60%. Looking at cure rates across the spectrum of loans’ current LTV ratios shows an almost remarkably consistent cure rate as loans with LTVs less than 60 have cured at a rate of 55% while loans with LTVs greater than 90 have cured at a rate of 51%. While cure rates appear elevated across most measures of borrower attributes, the analysis does not control for compensating credit characteristics and pools with larger populations of loans with layered risk may still experience higher default and lower cure rates than those without. Additionally, while cure rates are currently tracking ahead of where the market may have anticipated, loans that have re-performed or prepaid in recent months may be more positively selected and cure rates may slow over time as stronger credits re-perform and prepay while weaker ones remain in the pools.
The confluence of the recent rally in CRT, rising cure rates, slowing prepayments and borrowers exiting forbearance sooner rather than later could impair valuations on certain premium, earlier vintage, non-REMIC CRT. Extremely elevated prepayments have kept delinquency rates elevated as the population of delinquent loans have been falling at a slower pace than the reference pool balance have been prepaying, keeping the percentage of delinquencies elevated. As prepayments begin to normalize to some degree against the backdrop of higher rates, the ratio of delinquent loans relative to the overall size of the pool should start to fall, especially if cure rates remain elevated. The combined impact could drive delinquency rates below trigger levels, reopening principal windows at the top of the capital structure in these deals shortening the cash flow. However, given that these triggers are calculated based on a 6-month rolling average the impact of these factors may not be felt for some time.