Uncategorized

Putting a price on potential Ginnie Mae buyouts

admin | September 25, 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.

Ginnie Mae MBS have taken a beating throughout much of the summer, with prices falling and spreads widening relative to conventional MBS. The 3.0% and 3.5% coupons are currently trading at lower dollar prices than conventional, which is unusual territory for those coupons. A fear of buyouts is likely the underlying reason for this underperformance, but the bonds appear to be priced to the worst-case scenario—that 100% of delinquent loans are bought out on October 1. This seems unlikely and suggests that Ginnie pools are priced at too steep a discount to conventional pools.

Since the start of July, the Ginnie Mae TBA 3.0% coupon price has fallen nearly 30/32s more than the conventional TBA, and its OAS has widened 9 bp (Exhibit 1). The price of Ginie Mae TBA 3.5%s fell 22/32s more than the conventional TBA and its OAS widened roughly 32.5 bp. Asia was a heavy seller of Ginnie Mae MBS over this time, and one likely reason was to avoid exposure to further buyouts.

Exhibit 1: Ginnie Mae TBA has underperformed relative to conventional MBS

Source: Yield Book, Amherst Pierpont Securities

However, the large drop in the price of the Ginnie-to-Fannie swap and the current discount to conventional MBS both appear to much larger than the potential cost of Ginnie buyouts over conventional buyouts. Each conventional and Ginnie TBA from 3.0%s to 4.5%s was run through two scenarios at that TBA’s OAS (Exhibit 2). The first scenario excludes buyouts, while the second scenario includes buyouts.

Exhibit 2: The Ginnie discount to Fannie is typically larger than the price impact of buyouts

The cost of a buyout is calculated by calculating the OAS for each TBA, and then running each TBA though a scenario with a surge in buyouts. The difference in price is shown in the “buyout cost” column. The “% DQ” is the percentage of loans at least 60 days delinquent as of September 1. The “G2SF−FNCL Price” is the value of the Ginnie to Fannie swap. All valuations run as of 9/23.
Source: Yield Book, Amherst Pierpont Securities

The analysis assumes conventional buyouts occur in nine months and half of eligible loans cure, while the Ginnie buyouts happen immediately for all eligible loans. Loans that are at least 60-days delinquent are considered eligible for buyouts. This is an extreme scenario designed to maximize the cost of Ginnie buyouts and minimize the cost of conventional buyouts.

The difference between the cost of the Ginnie buyouts and conventional buyouts, shown in the column “G2SF Excess Buyout Cost”, is less than or equal to the current discount at which Ginnies trade to conventionals in the 3.0%, 3.5%, and 4.5% coupons. Put differently, if the “fair” price for Ginnies is flat to conventional, the current discount has built in a view that every delinquent Ginnie Mae loan will be bought out immediately. For example, the G2SF 3.0% currently trades 11/32s below FNCL 3.0%s, but buyouts on those pools should only hurt valuations by 6.5/32s.

However, the price difference is not likely to be this extreme. Non-bank Ginnie servicers have not been able to ramp up buyout efforts, which may indicate that it will be difficult to do so. They might find it better to allow loans to cure in a pool and earn income by attempting to refinance those loans later. Exhibit 3 shows the results of a few other buyout scenario. The first scenario replicates the results of Exhibit 2, while the second scenario also assumes 100% buyout rate but delays the buyouts by 6 months. The third and fourth scenarios assume 50% of loans cure.

Exhibit 3: The cost of buyouts drops significantly if half of delinquent loans cure

Source: Yield Book, Amherst Pierpont Securities

Delaying the buyouts by six months would improve the value of Ginnie Mae pools by roughly 2/32s to 5/32s, depending on the coupon. The benefit of lowering the buyout rate is much larger; in fact, cutting the buyout rate in half reduces the cost of buyouts by 4/32s to 16/32s, a larger than 50% improvement. And a 50% buyout rate coupled with a 6-month delay reduces the cost of Ginnie buyouts relative to conventional buyouts to only 2/32s to 5/32s. For most coupons, these buyout scenarios are less costly than the Ginnie discount to conventional pools and far less that the amount the price spread has fallen since the start of July. The 4.0% coupon is a little unusual—the coupon swap only fell 7/32s since July 1 and is only 5/32s below conventional.

This analysis does not consider the risk of a second wave of delinquencies, which would likely affect Ginnie Mae pools more than conventional pools.

Investors also can search for pools that have lower than average delinquency rates, which would have less buyout exposure. Seasoned vintages tend to have lower delinquency rates, as do new production pools. The worst delinquencies are typically in the 2017 through 2019 vintages. Pools issued by banks that have already bought out most of their delinquent loans, like Wells Fargo, can have very low delinquency rates and are likely only exposed to a second wave. Certain servicers’ borrowers have better than typical credit qualities so have lower delinquency rates and less exposure to buyout risk.