Invest1 publisher3 min readPublished
Banks return as second-lien sponsors using loans nonbank lenders originated
Citigroup and JPMorgan Chase are sponsoring home-equity securitizations filled partly with loans from nonbank channels. The capital rule that makes those deals attractive to a bank is not final. Nonbank sponsors are left competing on cost per deal.
The Investor · Invest desk
What happened
- A KBRA outlook report says nonbank originators rebuilt the home equity market banks had withdrawn from, and that bank-sponsored second-lien transactions have now reemerged and broadened issuance.
- JPMorgan Chase pooled mixed and temporarily frozen HELOCs with Shellpoint and LoanDepot as servicers and loans contributed by UWM and Better, according to Fitch Ratings.
- KBRA's HELOC arrears index, which it describes as predominantly nonprime, shows a 3.9% 30-day-plus delinquency rate against 2.6% for the equivalent prime product.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- constraint Fixed securitization costs sit on small second-lien balances, so cost per deal, not origination reach, decides which nonbanks can keep sponsoring their own transactions.
- decision Nonbank originators now choose between running their own shelf and selling production into a bank aggregation, and UWM took the second route in both named bank deals.
- exposure Bank second-lien issuance is exposed to a capital rule still in draft: recognising leverage from seconds held elsewhere would take away the reason to move them off balance sheet.
- contradiction KBRA calls bank re-entry a broadening of issuance while Dallas Capital's chairman says share is moving away from independent mortgage banks, and counts broken out by product cannot settle which.
Add the product counts and the market has already done 49 second-lien securitizations against 75 in all of 2025 [6][1][2], about 65% of last year's total with the year still running [3]. The composition moved too. Closed-end seconds were 43 of 75 deals last year and are 25 of 49 now, a slip from 57% to 51% of the count [4], and the small category mixing closed-end collateral with HELOCs went from 2 deals to 4 [5]. HELOC-only deals held their share, 30 of 75 then and 20 of 49 now [7].
Both bank deals on the record bought nonbank production. Citi aggregated first- and second-lien HELOCs including HomeTrust Bank's loans and United Wholesale Mortgage's broker channel, with Fay as servicer, according to Morningstar DBRS [7]. JPMorgan Chase pooled mixed and temporarily frozen HELOCs sitting in first or subordinate position on single-family properties, with Shellpoint and LoanDepot servicing and UWM and Better among the contributing lenders, according to Fitch Ratings [8]. UWM is in both.
The share in play is sponsorship. American Banker reported the re-entry is happening as rates keep rising, and that nondepositories in the space will need to manage expenses and diversify further to compete [1]. KBRA's analysts said bank interest has opened room for the two types of players to work together while compounding the lack of scale efficiency in a market with small loan sizes [16]. "Banks and brokers are going to take market share from independent mortgage banks. I think they have to lower their costs," said Bill Dallas, chairman of Dallas Capital. "They have to develop more modern platforms" [4][5].
The capital case that brings banks in is unsettled. The current Basel III proposal counts a second lien toward loan-to-value differentiators only while it stays in the institution's portfolio. So a bank holding the first mortgage has a reason to move the second off its books [12]. Robert Kazdin, former director of mortgage credit pricing at Fannie Mae, said rulemakers may revise that, because a second lien coexisting with a first adds to loan performance risk whoever holds it [13]. "By failing to recognize the additional leverage created when the second lien is held by another institution, the banking system becomes undercapitalized relative to actual credit risk," said Kazdin [14], who has suggested a registry to track the coexistence of firsts and seconds [15]. If the revision lands, the capital saving that makes the securitization exit efficient goes with it, and bank sponsorship thins.
I would expect nonbank origination volume to hold while nonbank sponsorship share falls, on the evidence of two bank shelves that both took UWM production [7][8]. The counter-case sits in credit. KBRA's HELOC arrears index has a predominantly nonprime composition and shows a 3.9% 30-day-plus delinquency rate against 2.6% for the prime equivalent [9], one and a half times the rate [6]. KBRA treats a pool as nonprime when weighted average FICO is below 720 or full documentation covers less than 75% [10]. If bank appetite stops at prime, nonbanks keep the collateral that carries the higher arrears and lose only the cheaper end of the shelf. KBRA's tallies in the outlook are broken out by product, so the sponsor-level shift shows up deal by deal and not in the totals. The thesis fails if bank deals add to the count while nonbank sponsorship holds. KBRA's team wrote that reemerging bank-sponsored transactions were "further broadening issuance" [3].
What to watch
- Whether the final capital rule keeps the portfolio-only treatment of second liens that makes selling them attractive to a bank holding the first mortgage.
- Whether KBRA breaks its next second-lien tally out by sponsor type, which would show whether nonbank shelves are losing deal count.
- Whether the nonprime HELOC 30-day-plus rate moves further from prime's 2.6%, and whether bank-sponsored pools reach below a 720 weighted average FICO.