Losses occur when a lender forecloses on a home and the sale proceeds don’t cover the mortgage balance.3 Lower-tiered tranches absorb losses first, followed by higher-rated tranches. A tranche begins to experience losses at its “attachment point” and cannot exceed losses beyond its “detachment point.”
For instance, if a tranche’s attachment point is 2.5%, as in the Last Cash Flow (LCF) tranche in the above graphic, then losses on the underlying pool will not affect the tranche until they surpass 2.5%. If its detachment point is 3.5%, the tranche experiences full principal loss when the pool’s losses reach 3.5%. Further losses on the pool cannot impact the tranche.
For prepayments, the direction of flows reverses. The most senior tranches receive prepayments first. Thus, in the display above, tranche A-H, which is held by the GSE, would receive 95% of prepayments on a pro rata basis (that is, proportional to the 95% of the capital structure retained by the GSE). The remaining 5% of prepayments would go to the subordinated tranches, where they cascade down (that is, not pro rata); the First Cash Flow (FCF) tranche would receive the remaining prepayments until it is paid off, with any prepayments not absorbed by the FCF tranche cascading down to the next tranche (LCF) until it is paid off, and so on. This structure usually creates an M1 bond that has a relatively short weighted average life of one to three years.
When a homeowner prepays their mortgage, CRT investors—especially those in the more subordinated tranches—tend to be happy, so long as their bonds are priced at or below par. After all, that’s one less borrower who might default. But when mortgage rates fall and homeowners refinance en masse, the payoff profile of the securities is affected, as their weighted average life may become shorter than expected. Further, if the tranches are prepaid, investors may find they need to reinvest at lower yields.
What could go wrong?
If the US were to experience a sudden and deep recession, a massive spike in unemployment and an influx of housing supply, that could precipitate a steep decline in home prices. (We do not currently expect any of these events.) Historically, high unemployment and deep recessions have led to high defaults on mortgages as homeowners were forced to sell.
However, today’s tight supply-and-demand technicals mean home prices are unlikely to fall, in our analysis. The government may also step in to prevent steep declines in home prices in the future.
How has the government prevented dramatic drops in home prices in the past?
In some instances, the government has taken measures to prevent home prices from spiraling down. For example, in 2020, when the US economy shut down due to the COVID-19 pandemic, policymakers enacted a forbearance program to assist homeowners—and they acted quickly, thanks to lessons learned from 2008. This program kept home prices from plunging. Borrowers who suffered economic hardship were able to pause their mortgage payments instead of defaulting on the mortgage or being forced to sell, which would have flooded the market with inventory.
What is your current outlook for CRTs?
We don’t see another housing crisis on the horizon. In fact, our outlook for CRTs today is good. There are two main considerations when assessing CRT fundamentals: the credit quality of the underlying pools of loans, and the outlook for the US housing market.
We expect home prices to rise modestly in 2024. With interest rates now relatively high, affordability has declined. But while rising mortgage rates have put a damper on demand, they’ve also contributed to a tight supply of single-family homes. The inventory of existing homes—which comprises 70% of all homes on the market—is extremely low. That’s because homeowners have locked in low interest rates from past years and don’t want to move. In turn, ongoing supply constraints provide support for home prices, even as demand moderates.
We expect housing demand to remain low due to affordability issues caused by higher rates and home prices. But, while renting has become more affordable than home ownership, the percentage of renters has been declining because of a general preference for owning, combined with fear of missing out on potential further home price appreciation.
Though mortgage rates are high, existing borrowers have locked in lower rates. Thus, all else being equal, their ability to service debt has remained unchanged even as mortgage rates have risen. (Of course, this could change if economic conditions were to deteriorate and homeowners were to lose jobs. However, given the massive appreciation in home prices in recent years, most homeowners would be more incentivized to sell their homes than to default on their mortgages.)
As to the credit quality of the CRTs, fundamentals appear to us to be very strong. The ratings agencies seem to agree; every bond issued before 2021 that was originally rated has received a rating upgrade.
Home price appreciation has also led to the natural deleveraging of CRT bonds. More seasoned bonds have especially strong LTV metrics (Display). What’s more, as older-vintage CRTs have de-levered, the GSEs have been making tender offers to buy the bonds back from investors. After all, the GSEs issued these bonds to hedge their credit risk, so there’s no reason to keep them on the market once that risk is no longer there. As a result, investors today can potentially own CRTs with very little credit risk but with attractive spreads; they may also benefit from GSE tender offers, which, alongside limited issuance, have helped push CRT prices up.
