Excess Spread — Whipsaw, progress, the long game

I skip two weeks, and suddenly there’s a ton of news!

Prior to the midterms, deal executions that would previously have sailed through were struggling, credit spreads were widening, and execution windows were narrowing, only for this to rip back the other way after the Blue Wave in the midterms, followed by a sell-off in the AI complex the day after. Markets eh?

There was nothing ABS-specific about the wobble — market jitters often emerge late in the year, with nobody much inclined to be a hero in Q4, and yearly allocations largely spent. The excuses are valid, and available to suit one’s perspective; is AI supply causing indigestion, is CoreWeave a house of cards, is there a pod or pods blowing up, is France on fire / on track for sustainable public finance? Or were credit spreads just too damn tight and due a correction? All these and more have done the rounds.

StrideUp’s second Islamic trade, Meridian Funding 2026-1, seemed to run into the market turmoil, with a drawn-out syndication that started 12 October but didn’t wrap up until 2 November, resulting in below par placement. Before the badness started, Pepper got its sterling seniors away at 89bps, vs 120bps for Meridian, despite the considerably better collateral in the StrideUp deal.

Still, this wasn’t as ugly as Goldman’s Irish exit, GSMBS IRPM 2026-1, which we discussed before going away — the relaunch with a senior anchor got this bond to 1.1x covered, but even a 50% protected order for classes B-D couldn’t get final coverage levels above 1x down the stack when spreads were locked.

Final pricing was 90 / 165 / 185 / 220 / 380 / 540, vs 65 / 95 / 120 / 145 / 270 / 400 for Barclays’ Beckett Mortgages 2026-1, backed by the same Nùa Money collateral (though admittedly preplaced across the stack), and a touch less levered.

Despite this backdrop in the runup to Halloween, NewDay nonetheless braved the market, but this is NewDay all over — it’s a big capital structure, there’s a lot of funding to do, and they are gonna issue their bonds and damn the torpedoes.

This is a firm that printed senior notes at 260bps in 2022, a little 15bps of widening isn’t going to scare it out of the market (and assets yielding in the mid-30s can bear a bit of liability softening).

In the event, it might have started cheap in a tricky market, but the market turn meant a big move in syndication – guidance of 92a / 140-150 / 170-180 / 200a / 330a turned into final guidance at 88 / 120 / 150 / 190 / 310, on monster coverage levels at guidance of 6x / 9.6x / 8.4x / 4.9x / 6.2x (class A coverage on free float). It’s still cheaper than the last NewDay Funding (80bps on senior) but torpedoes largely avoided.

Ellington Management also launched a Spanish RPL transaction, Sierra Residential, which looks to have surfed the same wave — the partly-protected senior tranche (there’s a €75m lead order, and BNPP, likely provider of said order, joins Morgan Stanley in the syndicate) is talked at 115-120bps, vs the last Spanish RPL print, Lugo Funding 2026-1, which landed at 120bps.

It’s not really a fair comparison — all Spanish RPL pools are different, as evidenced by a triple-A advance rate 15 points apart for these two trades — but it’s the most recent comp we have.

As we’ve discussed, it’s not just different pools and different deal structures at play, but sometimes, fundamentally different approaches to making money from these assets and transactions — some trades are for funding, others are for arbitrage.

Banks are working through their asset sale programmes, though, and selling cleaner pools — potentially more suited to a buy-and-hold approach. This is the best quality Spanish pool Ellington has bought; to once again use triple-A advance rate as a rough proxy, Clavel 4 in 2025 had a 71% triple-A, vs 80.5% in the new deal

It’s also interesting that the last RPL deal backed by Santander’s Spanish mortgage collateral was issued by Santander itself, Santander Residential 1. As we remarked at the time, this was relatively unusual in using a Spanish FTA structure, rather than the more usual double-layered package where mortgage certificates secured a bond-issuing Irish DAC.

That deal was a bit of a journey — partly because of the Spanish capital markets authority — so perhaps the portfolio sales team has decided to revert to its more traditional approach, and let MS and Ellington once again get the benefits of securitisation.

Progress!

It’s also been a fortnight that’s seen a genuine regulatory breakthrough, with the European trilogue reaching provisional agreement on the securitisation package. Writing about regulatory progress can feel a little like hiking up a mountain; every time you reach a peak, there’s just more damn hill ahead of you.

But agreement between the European policymaking bodies is a real milestone, and the measures proposed will make a real difference. STS for unfunded SRTs will be helpful, but narrow — few insurers currently qualify (likely between two and four) though it will be a big advantage for those specific firms.

The big deal is the capital cut for senior tranches, which we already discussed.

Markets are busily gearing up for the Solvency II changes — here is a piece from 9fin’s CLO team looking the new mandates and possible allocation changes ahead of the implementation of the changes in January.

But if they’re busy on Solvency II, the senior capital cut will be even bigger. Alas, though, there’s still a bit more hill to climb; there’s a “technical trilogue” to sort out some of the details, and then actual implementation will have to run through the usual round of consultations and technical standards from the European regulators. As we saw with risk retention, these aren’t trivial.

Anyway, here is Celeste’s excellent longer form piece on the agreement.

Playing the long game

Say what you like about the market backdrop, but there are still some highly innovative transactions getting done.

Morgan Stanley printed Ironwood Residential 2026-1 on Tuesday, the first deal put together through the painstaking process of calling and refinancing legacy 1.0 transactions — a much-discussed but difficult to execute trade. We started discussing the possibility back in June 2025, but things have heated up this year, with Bank of America also working to get it done.

Ironwood Residential is backed by a £419m loan pool comprising three sub-portfolio — RMAC, derived from the unwind of several deals in the RMAC 2006 series, Gloucester, from the Mansard 2006-1 deal, and Stilton, bought from Balbec earlier this year.

Originators are a wide spread of largely defunct pre-crisis names — GMAC RFC, High Street Home Loans, Future Mortgages, Amber Home Loans, GE Money Home Lending, Household Mortgage Corp, and Rooftop Mortgages.

The loans are so old that many are nearly final maturity — some 50% of the pool is scheduled to mature in 2031. 87% of the pool is interest-only, so despite the 243 months of seasoning, the balances remain outstanding.

But decades of house price appreciation has taken average LTVs from the high 80s at origination to around 50% for this deal. S&P’s rating analysis recorded a 4.9% principal loss on repossession of the properties — lower than average in legacy non-conform, according to the agency.

Plenty of pre-crisis collateral has appeared in newly securitised format — most recently Hawksmoor Mortgage Funding 2026, a Pimco-sponsored refinancing of loans originally aggregated by Davidson Kempner.

There are multiple potential approaches, depending on deal circumstances and documents — but the RMAC deals, to take one example, required a consent process to raise the threshold to exercise the clean-up call.

This allows the deal to be called not via an option held by the equity, but through an option held by the servicer.

Mansard 2006-1 had been theoretically callable under the clean-up since January 2025, but a portfolio auction carried out then did not result in a sale.

There are multiple ways to profit, depending on the transaction in question. Deals with significant cash trapped inside will flush this out to the resid once called; owners of bonds bought at a discount naturally profit from a call at par. These beneficiaries might be the same entities or different!

The re-rack is also a profit source. While these borrowers have had outstanding mortgages for the better part of 20 years or more, they’re not necessarily ‘mortgage prisoners’ — the pool is nearly all BBR-linked, a product that’s died out.

Pool yield is 6.23% as of the August pool cut, or 2.48% over Bank Rate at the time. While a refi in the ZIRP era might have been more optimal, there are valid reasons a borrower might want to hang onto their IO product as long as possible — and refi only at maturity, which is now approaching. That could mean owning this book isn’t just owning a book of zombie mortgages with a new cap stack; it’s owning a book where refinancing could be poised to pick up.

But fundamentally, we like this trade because it’s cool. It takes a long view to line up this kind of transaction, it takes legal and structural expertise, work (in the form of the consents), and a picking of the optimal moment in the life of the mortgage pool. We dig the ambition and the vision.