The case for Ginnie Mae low WALA and post-buyout bank pools
admin | July 24, 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.
Servicer activity has turned the tables on investors expecting dampened prepayments from borrowers taking Covid-19 forbearance. Ginnie Mae speeds jumped in June caused largely by bank servicer buyouts of loans taking forbearance and consequently considered delinquent. Speeds for these servicers should jump again in July but fall afterwards. Since the pace of new delinquencies has fallen, new issue pools should also face much lower buyout risk. Buy the Ginnie Mae post-buyout bank pools and the low WALA pools.
Buyout incentives for banks despite new Ginnie Mae rules
Ginnie Mae recently announced restrictions on pooling loans that the agency hopes will slow buyouts. Loans brought out of delinquency without modification may only be delivered into custom pools and must meet payment history and seasoning requirements before pooling. This is effective for buyouts beginning in July. Servicers can make a lot of money buying out loans at par and resecuritizing at a premium (a recap of the economics is here). Ginnie Mae’s rules cut into this premium by forcing the servicer to wait at least seven months before repooling, which substantially reduces the economics. But this has a much smaller effect on banks than on non-banks since banks’ funding costs are low, so the rules are unlikely to deter future bank buyouts.
Accounting rules also encourage banks to buy out delinquent loans. Wells Fargo, for instance, has noted (here) that “loans in Ginnie Mae pools must be consolidated on a bank’s balance sheet once the loan is more than 90 days past due, regardless of the circumstances.” Without this rule, banks might have an incentive to keep delinquent loans in pools and off balance sheet.
Bank buyouts were very high in June
Loans bought out in June are not subject to Ginnie Mae’s new pooling rules. However, only bank servicers bought out an appreciable number of eligible loans (Exhibit 1).
Exhibit 1. Banks bought out 69% of eligible loans in June
Servicers can buy out loans once they are 90-days delinquent. But this means a loan that starts the month 60-days delinquent can be bought out during the upcoming month. Investors need to look at a pool’s 60-days or greater delinquent population to assess the potential buyout risk in a pool. The percentage of 60-day or more delinquent loans in bank pools fell from June to July due to the high buyout rate, while the percentage increased for non-bank servicers. It is also useful to compare to pre-Covid average levels, to show how much higher the current delinquency rates are than they were prior to Covid.
The 30-day delinquent numbers, however, have fallen dramatically since May and are very close to the pre-Covid levels. This implies that the pace of new delinquencies has fallen close to pre-Covid levels. Another way to see this is to look at the 30-day and greater delinquency rates for non-banks in June and July—11.9% [4.7%+7.2%] and 12.2% [3.5%+8.7%], respectively. Buyouts and natural curing were both very low, which means there were few new delinquencies in June. The slow pace of new loans entering forbearance has also been reported by the MBA.
Reviewing the largest bank and non-bank servicers shows that most of the buyouts were conducted by five large banks (Exhibit 2). Wells Fargo, U.S. Bank, Truist Bank, and Chase each bought out virtually every loan that became 90 days delinquent during the month of June, and M&T bought out 66% of eligible loans.
Exhibit 2. Four of the largest banks bought out almost all eligible loans
The percentage of loans at least 60-days delinquent is much lower for the servicers with large buyouts, while other servicers’ pools still have a huge population of delinquent loans. Some of these loans will become buyouts—most VA loans are likely to require loan modifications to cure their Covid delinquencies, which mandates a buyout. And some FHA loans will not qualify for a deferred payment, also mandating a buyout.
Non-bank servicers might choose to leave FHA loans that receive partial claims in pools. Ginnie Mae has clearly signaled that the agency prefers this outcome by releasing the APM and its associated commentary. But the risk remains that a servicer might choose to buyout loans even though the economics are significantly worse than before. For example, maybe they pocket a portion of the premium by selling the delinquent MSR to a bank, to take advantage of the bank’s funding advantage.
Hedging the buyout risk in pools with a large percentage of delinquent loans is challenging. The uncertainty regarding the magnitude and timing of buyouts is likely to weigh on prices for these pools. At the same time, the pools with high buyouts will be left with much less uncertainty and should command a pay-up over otherwise similar pools with high delinquency rates.
The most obvious securities to benefit will be the high-buyout lenders’ single issuer pools and securities backed by those pools. This can even include seasoned Ginnie I pools since those banks were active lenders when that program was larger. CMOs backed by this collateral can also benefit from more predictable speeds once the buyouts have passed.
Because the Ginnie Mae market used to be dominated by bank originators seasoned Ginnie Mae multi pools tend to have high shares of these banks’ loans. For example, consider G2 MA0534, the 3.5% multi pool issued in November 2012. Roughly $2.6 billion is outstanding and 37% is serviced by Wells Fargo. The pool’s 60+ delinquency rate fell to 5.3% in July from 6.3% in June. But if Wells had not bought out any loans, the delinquency rate would have increased to 8.2%. In July Wells buyouts are likely to add 10 CPR to the pool’s speeds but the overall delinquency rate is likely to fall close to the pre-COVID level.
For comparison, consider G2 MA5594, the 3.5% multi pool issued in November 2018. Wells and other banks are not a large share of the pool—Wells is only 7%–so the buyout rate was low and the 60+ day delinquency rate increased to 9% in July. Investors in that pool face significantly more prepayment uncertainty than G2 MA0534.
Finally, pools aged one month or less should accumulate far fewer delinquencies than even slightly more seasoned pools since the pace of new delinquencies has slowed. Low WALA pools consequently have a much lower risk of future large buyouts. Low WALA 2.5%s are trading flat to TBA and low WALA 3.0%s are 3/32s to 4/32s over TBA.