Excess Spread is our weekly newsletter, covering trends, deals and more in structured credit and ABS — subscribe to this newsletter here.

Wave it in

The glory days of bank deleveraging are behind us, but there’s a new wave of disposals coming through. Banks are optimising their perimeters, tackling the corners of the balance sheet which are suboptimal, and casting a ruthlessly capital-efficient eye across asset holdings.

If the urgent clean-up operation 2014-2019 was a mix of “chuck it in a skip” and “hide it in a cupboard”, at some point the cupboards must be opened and the contents sorted out. That point is now.

So we have monster disposals of performing assets, like HSBC’s Project Temper. As we reported last week, it’s an unusual one. French mortgage origination is generally done to tight criteria, especially at HSBC, and the products are long-dated and fixed for life, with most of them written below current interest rates. It’s therefore kind of a rates trade; the discount where this will trade (HSBC expects to book at least €1bn of loss on the €7bn portfolio) is pretty much an artefact of duration.

That will be particularly interesting if one of the bidder groups (Rothesay and My Money Bank) triumphs. My Money Bank is the old GE Money Bank France, bought by Cerberus in 2016, and it bought the actual retail operations of HSBC France in 2021, known as CCF.

We don’t know much about the negotiations ahead of that sale, but HSBC almost certainly would have preferred to sell the whole thing assets and all. It can’t have set out on that journey thinking “yes let’s keep a stub of €7bn in mortgages we don’t even service in a country we’re closing retail operations”.

Presumably price for these assets was a sticking point and the simplest exit was to carve out mortgages from bank and move on.

If the mortgages are subsequently bought by My Money Bank at a substantially higher discount that was offered in 2021, that’s probably the most comical outcome, though all it really shows is that HSBC was on the wrong side of a rates trade. Most of the financial industry was as well, that’s what you get for 10 years of ZIRP. Bank accounting is a wonderful thing though, since presumably if these end up with a bank they get dumped in a held-to-maturity book and written up again. HSBC, as it’s doing a real sale, has to mark-to-market, but most banks don’t hold their mortgage books that way.

Other bidders in the process are the usual suspects. Pimco and Apollo aren’t generally far away from any big asset sale process, and benefit respectively from the ability to buy unlevered and access to long-term insurance capital which suits 20yr+ low CPR portfolios much better than securitisation leverage. CarVal is the odd one out, as it does require external leverage, and that can be a struggle to find in bank or securitised format for this sort of collateral. The most suitable capital source would probably be one of the non-captive insurers (something like the Waterfall / Aviva equity release deal we discussed early in the year), but the ticket size is massive.

The Spanish banks are also opening up their cupboards and sorting through asset portfolios, with plenty of activity especially in reperforming sales from some of Spain’s largest institutions.

We’ve discussed the Balbec transaction, PRPM Fundido 2025-1 before (a reader notes that Queso Fundido is an apparently very delicious Mexican fondue type thing) but what’s relevant for our purposes is that this aggregates three different portfolios, and Balbec has already loaded up its Goldman-provided “hunting line” facility with enough assets to come back.

Also in market was a CarVal-owned RPL portfolio, set to be securitised in Miravet 2025-1. CarVal used the Miravet brand for its previous Spanish RPL portfolio, carved out of the giant Project Hercules CaixaCataluyna-Blackstone sale in 2014. But these in turn were sold on to One William Street, which securitised what was left in Lugo Funding last year.

Indeed, 2024’s securitisation supply was closed out with a Spanish RPL deal, Jeronimo Funding, sponsored by Elliott’s captive insurer Medvida, and financing a portfolio bought from Unicaja.

But more primary RPL portfolios are in the pipeline, and likely need funding. BBVA had a portfolio called Project Sayonara out, while Santander is marketing a deal called Project Flamenco, with final bids due in a month or so.

The NPL market has also been active lately, with CaixaBank selling portfolios to Cerberus and Apollo last year, and currently marketing a residential NPL portfolio, Project Malbec, with Apollo, KKR and Balbec in the final round, as first reported by Bloomberg.

Although we’ve lumped them together a bit here, NPLs and RPLs are very different animals, especially with respect to servicing.

RPL portfolios bought in primary (direct from banks) have already gone through some of the trickiest parts of the servicing process — borrowers fell behind or couldn’t pay, they’ve been offered a payment plan, perhaps with maturity extensions or interest capitalisation, they’re paying again.

Banks are generally unwilling to push these borrowers out into the arms of a more NPL-orientated servicing platform, and want to maintain their relationships. So in the Balbec deal, for example, the three originators whose portfolios back the deal (Sabadell, Abanca, and Cajamar) service the loans, while Pepper, a servicing specialist with more experience in credit-intensive servicing, sits behind as a master servicer, and takes over servicing when borrowers fall behind.

NPL portfolios, meanwhile, have tended to trade on the basis that a third party servicer can be brought in, be it a captive like KKR’s Hipoges or the formerly Cerberus-owned Haya Real Estate, or a third party (often historically one of the corporate debt purchasers such as Intrum, or one of the Arrow Global units). These servicers are well paid for their work, and servicing strategy tends to drive deal performance and structure; transaction triggers for NPL deals are usually based on servicer business plans and expected collections, while RPL deals work essentially like an ordinary RMBS, albeit one with a low WAC / low CPR pool (and get better rating agency treatment as well).

NPL securitisation in Spain has a somewhat chequered history.

KKR’s Salduero (ProSil Acquistion) in 2019 was one of the first publicly placed Spanish NPL transactions, though there had been private deals before — Credit Suisse’s financing for Blackstone’s purchase of Hercules, for example, was syndicated out to market.

At the time, banks generally advanced around 70% of purchase price for NPL pools. KKR’s deal docs disclose a purchase price of €225m for the portfolio, and the class A and B notes were for €200m notional at issue, so, uh, this was quite levered!

Nonetheless, an IG rating was achieved on the senior notes (though Moody’s assigned a Ca to the class B), and the deal was done!

Unfortunately, the deal ran headlong into the disruption of the pandemic, paying two interest payments to the class B notes before Covid shut down cashflow below the senior.

KKR, however, had by this time sold the equity in the deal, getting out just before the lockdowns hit. Now the deal is nearly 50% behind business plan, and it doesn’t look to be turning round any time soon.

We couldn’t firm up who currently owns the junior, but a bit of vigorous googling reveals Prytania owns part of the senior, financed on repo with JP Morgan, and MSIM is in it as well.

Regulatory change is also afoot, as Spain incorporates the EU NPL Directive into local law. Right now, NPL and RPL deals alike have a complicated construct where deals are backed by mortgage transfer certificates and mortgage participations, rather than the mortgages themselves, as a route around Spanish laws on land registration and title transfer. The new law proposes to streamline this process, and to loosen registration requirements for a purchaser of NPLs to be registered with the Bank of Spain. Bring on the next round of bank sales!

Shooting from the hip

European policymaking usually takes, oh, about half a decade.

The timeline from the “Capital Markets Union” initiative in 2014 to the start of the Securitisation Regulation in 2019 fits the usual mould, with multiple rounds of consultation, three different proposals from Commission, Parliament, and Council, an extended trialogue process to harmonise all three, a draft document, a final document, final implementation…..and then the real work starts, as the regulatory agencies publish their “Level 2” documents (Implementing Technical Standards and Regulatory Technical Standards) to tell the market what the “Level 1” actually means.

Sometimes, however, the process gets short-circuited, as it did with the report from the Joint Committee of the ESAs (European Banking Authority, European Securities and Markets Authority, European Insurance and Occupational Pensions Authority all working together) at the end of March.

This wasn’t supposed to be making rules at all — it’s a “report on the implementation and functioning of the Securitisation Regulation” — but it laid out the regulators’ thinking on risk retention, specifically the “sole purpose test” (to see if a risk retention entity is a proper thing or just an empty box to swerve an annoying requirement).

The thinking unveiled (that risk retention vehicles should earn no more than 50% of revenues from holding risk retentions) was pretty different from the interpretations that had prevailed in the market, and it briefly caused chaos in the CLO market, the intended target of this warning — partly because it was held to be valid immediately, no carve outs for, say, deals which had priced but not yet closed, no grandfathering, nothing except a bald and surprising announcement.

A variety of rumours are doing the rounds about how this came to pass, who frightened the horses, which odd currents of Europolitics are at play, which we won’t comment on until we can stack something up, but it’s a deeply odd situation.

Shooting from the hip sometimes hits the wrong target. Although the report is very specific about the intention to apply this test to CLOs, and address concerns with the CLO market, a 50% revenue test could inadvertently wound portions of the specialist lending market.

Lenders which use thinly capitalised assetco structures as risk retention entities might readily find themselves generating more than 50% of revenues through risk retention pieces, either of public deals or private warehouses; the other major revenue stream for a lending business is fee income, which would run through the opco.

Pepper Money, which just priced the very successful Castell 2025-1 second charge RMBS, is one such entity, having switched over to an assetco approach last year. The seller and risk retention holder is assetco UK Seconds Lending (UK Residential Mortgage Lending is the first charge equivalent supporting the Polaris programme).

Accordingly, the Castell deal docs flag a risk factor for the Joint ESA proposals, the first example we’ve seen in a public securitisation, though likely not the last. This notes: “Given the recency of the publication of the March 2025 JC ESAs Report and as it is pending further clarification, it is not yet clear to EU securitisation market participants what the full consequences or scope of such proposals discussed within the March 2025 JC ESAs Report are intended to be and/or whether further amendments to the EU Securitisation legislation or EU Risk Retention RTS will follow, and if so, when. No assurance can therefore be given that the EU Securitisation Regulation will not be further changed in the future or that there will not be further views expressed by the EU or national supervisors as to how the ‘sole purpose’ test should be interpreted in practice in certain circumstances, which views may or may not reflect the conclusions reached by EU securitisation market participants.”

These assetcos are not the separately capitalised third party risk retention vehicles common in the CLO market, and the theoretical target of the rule change, so lenders are hopeful that any application is nothing more than accidental oversight which will be amended and clarified down the road.

Which brings us, of course, to the next big round of regulatory change. The European Commission will be releasing its securitisation proposals the week after Barcelona, and surely that, rather than a non-binding report, is the place for tearing up the rulebook? Expectations aren’t running high for a fulsome reworking of the rulebook, with adjustments, tweaks and removal of pain points more likely.

Whether the proposals are ambitious or not, the road will be long. Let’s hope there’s no collateral damage along the way.

Make Mortgages Great Again?

Britain has few great industries left, few sources of economic dynamism, but a deep and abiding love of property speculation.

So it is that the FCA is mulling loosening rules on mortgage origination, though it is framed as removing rules which are redundant with last year’s introduction of the Consumer Duty. More will follow in June, with a fulsome public consultation, but the changes outlined on Wednesday will simplify affordability assessments, streamline rules over offering advice, a regulated activity, and remove some guidance over how firms should deal with legacy interest-only products.

The FCA wants to enhance the availability of “execution only” mortgages (without regulated advice). Before the 2014 Mortgage Market Review, around 25% of mortgages were non-advised, but since then, it’s generally been less than 5%.

The changes to affordability assessments are meant to make external remortgaging easier. The FCA says that 83% of remortgages remain with existing providers, thanks, in part, to the costs associated with lenders looking to re-underwrite. If the FCA changes move the needle, one would expect even fiercer competition, and perhaps reduced takeup of product switches, which became increasingly important as mortgage rates rose in 2022 and stretched borrowers wanted to avoid re-underwrites.

An easier route to execution-only mortgages sounds promising for the various fintechs looking to streamline the home loan process in the UK. These efforts have been only partially successful, because it’s hard to streamline away much of the regulation.

But it could prove mainly to be an enhancement for the High Street, since “execution only” is much more compelling for prime, low LTV employed borrowers. Specialist lenders of owner-occupied rely heavily on brokers for their origination and there will always be much more engagement required for a borrower with a past of credit issues and complex income sources.

Also of interest for securitisation markets is a change in treatment for legacy interest-only products, many of which were originated pre-GFC and are seeing final maturity dates looming up in the 2030s. Lots of these products back the still-outstanding legacy RMBS transactions; being IO, they don’t tend to repay quickly, though the FCA noted that this had run a bit ahead of regulatory expectations.

The new proposals ditch existing guidance and rely instead on the provisions of Consumer Duty to protect customers. Does this increase CPR on loan portfolios which are 20 years old? TBC.

Get in touch below to find out what else we offer for ABF.