The strength is particularly encouraging because it comes in partial defiance of the ECB’s withdrawal from the market. Nearly all of the recent supply has been ECB-eligible STS paper, and the central bank would once have taken a sizeable bite out of the senior total. The naysayers have been proven wrong, and the best month for ABS since the pandemic has come with no central bank support at all.

But look a little closer and you can kind of see the mark left by the ECB withdrawal. We wrote about the unusual Vasco Finance No. 1, a Portuguese credit card ABS, when it was first mandated at the end of August. The fact that it’s coming into land three weeks later with seniors only 1x done thanks to a JLM order suggests that it’s been a complex syndication process.

In the days of the ABSPP, this would be a nailed on ECB trade; the central bank didn’t have an explicit capital key mandate for ABSPP as it did for the government bond buying programmes, but it was permanently over-indexed to Dutch prime and German autos, being the largest primary asset classes in euros, and would surely have swooped on some rare Portuguese bonds.

There remains a certain defensiveness about senior placement in other deals, too. Renault subsidiary DIAC opted to place only €500m of its €700m senior tranche, allowing an upsize to the overall deal by retaining the balance. But it came pretty close to the line, as it was just 1.1x done on a €400m senior placement when final guidance was released on Wednesday morning.

Santander Consumer Spain Auto 2023-1 has protected orders in the top two classes, while Golden Bar 2023-2 was entirely preplaced.

First Citizen Finance, a Magnetar Capital portfolio company, took a more fundamental approach to CIART 2023, an Irish auto ABS. First Citizen hasn’t issued a public deal since 2020, but the latest outing adds STS eligibility, requiring tweaks to the deal such as excluding loans which haven’t yet made a payment. The deal also ditches the revolving period featured in CIART 2018 and 2020, which is not an STS requirement, but should also point the way to tighter pricing.

To the extent that the market now depends on private sector bank treasury demand, regulatory treatment is crucial — and so the divide between triple-A rated northern Europe and double-A southern Europe could well widen in the post-ECB world.

We note that Marzio Finance 12-2023, an Italian CQS (salary sacrifice) ABS from IBL Banca has managed to achieve a triple-A with Scope, if not with Moodys and DBRS, which on my reading of the rules should do the business for LCR-eligibility, but the rating agency rules are punishingly complex; the ECB, if I recall, took a much narrower approach to which agencies it trusted than the wider European regulatory apparatus.

Euros in the shade

My growing team of CLO-focused colleagues have produced an excellent piece lifting the lid further on anti-”snooze drag” provisions in the current crop of CLO new issues. “Snooze drag” is the colloquial name for the process by which CLO managers do not vote in favour of an amend & extend transaction, either because they are restricted by their docs or for fear of annoying triple-A investors, but where they nonetheless want to be rolled automatically into the amended market-coupon facility rather than left in the stub.

This is typically better for equity, but not for triple-A investors hoping to see their old bonds amortise.

We clocked some of this language in the latest Onex deal over the summer, which basically imposed a big haircut on par value when a manager abstains from voting on an A&E, but it’s a very live issue with triple-A accounts at the moment, and a heavily negotiated stip.

In other words, we’re in the creative period for CLO docs, when the turmoil of supply and demand are balancing and market standards are being forged. Ask if you’d like a copy of the full report.

The menu of options include treating assets as defaulted, forcing a sale, altering voting permissions, and enhancing disclosures.

To be fair, it does seem like some “abstaining” managers are not really abstaining. We hear of consent forms for A&Es in which managers can tick a box to confirm they abstain on the understanding that they will be dragged. Don’t tick the box and you stay in the stub, tick the box and you roll. You’re affirmatively consenting to abstaining, which apparently makes sense to lawyers.

Of course, this is a classic case of bolting the stable door when the horse is already across the border. This crop of deals won’t drop out of reinvestment until 2028 (if their reset options go into the money before then it will be even longer). Even now, loan market conditions are such that the bulk of A&Es are probably behind us; and those that remain should be able to lean on new money as much as docs-constrained old CLOs. 

On that new money point — primary is feeling good. The summer break was never really a break, this week brought two new deals, including an upsize from Oaktree, and there’s more to come, with Invesco marketing for execution next week and the Pemberton debut clicking along, set to price by the end of the week. Even aside from the A&E debate, it seems that triple-A investors remain in charge; Pemberton’s deal features another make-whole feature at the senior level, as seen in Onex’s OCPE 2023-7. This basically compensates senior investors in the event that equity rushes to call as soon as the non-call is up; the twist for Pemberton is that the double-A gets a make-whole too.

For decent loan borrowers, we’re now back up to pricing levels where a straight refi is fully on the table. The ELLI is at its highest level since May 2022, and we’ve even just about in repricing territory for some borrowers. Nord Anglia Education, a January A&E candidate, has been trying to shave the margin lower on its main dollar and euro facilities this week. There’s not likely to be a huge wave to follow, but market conditions are clearly supportive.

A rising tide cannot magic up a wave of new LBO supply, and here the news is not all good. 

Euro loan investors have been somewhat stiffed by the hotly anticipated Worldpay LBO. The original deal package contemplated $1bn-equivalent of euro TLB, already a skinny slice of the $8.4bn debt to fund GTCR’s $17bn takeover.

During syndication this was slimmed down to $500m-equivalent of highly oversubscribed facility, with the balance more than taken up by increasing the dollar TLB from $3.4bn to $5.2bn. The dollar bonds were boosted from $2bn to $2.175bn, and the sterling cut back from £700m to £600m. Final demand levels were such that sponsor GTCR decided to reduce its equity contribution as well.

You wait all summer for a decent large cap new money LBO, and it all gets monstered by the dollar market. 

A rising tide, however, does lift all CLO equity NAVs, and that seems likely to shake out some old portfolios. A call notices has been published for the decidedly aged Jubilee 2007-XVIII transaction, for example. Sound Point Euro CLO IX, a static from last year, may also redeem, according to a cleansing notice, rather join the growing crop of 2022-vintage resets being explored.

How the mighty have fallen

Back when securitisation markets were free, bonuses were fat and truly deranged products like CPDOs stalked the earth, there was some pretty egregious behaviour in the field of commercial real estate hedging.

My personal favourite was the Barclays conduit Gemini (Eclipse 2006-3), which had a 20 year senior interest rate swap attached, maturing seven years after note final maturity, and 10 years after expected maturity. The swap break costs were so egregious (Fitch cited £287m in this 2012 downgrade) that value for the deal broke in what was once the triple-A class. Apparently I wrote about it at the time for IFR.

Anyway, in the “CMBS 2.0” era the pendulum has swung right back in the other direction, and that’s also storing up some trouble. 

CRE loan hedging isn’t a profit centre so much as a necessary evil, required as a route to distributing CRE loans in whole loan or CMBS format, and the predominant method is no longer a ultra-long swap with the mark-to-market baked into the deal, but a tiddly little cap which doesn’t even cover the potential lifespan of the loan.

Loans backing European CMBS usually have an expected maturity, with one or two extension options, and some discretion for further extensions. Caps attached to the loans, however, might only cover the initial expected maturity, with sponsors expected to go out into the market and source a new hedge afterwards, or encourage the existing hedge provider to roll.

Let’s arbitrarily take Citypoint, a prime central London office block which is no stranger to the CMBS market — it was securitised pre-crisis in Ulysses (EloC 27), and was subject of a contentious restructuring battle in 2012 and 2013. The same asset, now in the hands of Brookfield, now secures Salus (Eloc 33), another Morgan Stanley deal.

Anyway, the initial deal had a 2.5% cap with Wells Fargo until “Initial Senior Loan Maturity” in January 2022 (with an extension to January 2023). Wells had extended further to match the “Second Senior Loan Extension” to January 2024 — but the sponsor has yet to reveal its plans for January and beyond.

It’s a very prominent trophy asset, so one hopes that it can dodge the broader office sector worries, but the current valuation is £670m and passing rent is £24.1m — so, y’know, just buy Gilts instead? 

Certainly don’t buy this thing levered — the current coupon on the 61% LTV senior class is 6.15%, so you’re better off owning triple-A CMBS bonds than the actual tower itself. The 2.5% cap is taking a lot of the strain (Wells Fargo kicked in nearly £2m last quarter), and when that goes, or is restruck at current rates, it’s going to be very painful indeed financing this asset.

Big picture, CMBS loans post crisis are structurally underhedged, particularly as the loan extensions pile up, and that can only add to the strains in the sector — sponsors either need to cough up for off-market hedges, or find a way to seriously crank up asset yields.

Interest rate caps on office buildings are one thing, but some deals have some seriously weird mechanics built in. 

Sage AR 2021, the Blackstone-sponsor social housing CMBS, could see its hedging fall away in November this year (a 1% Sonia cap). If that happens, the deal contains an unusual provision to cap Sonia itself at 4%, and subordinate any payments that would have been made over the 4% market.

S&P says that “Based on our conversations with the sponsor, we expect that the hedging will be extended in November 2023”, and the deal provides that, if the new hedge maintains interest cover at over 1.5%, then no drama.

But if not, then noteholders that expected their full payments will start receiving a “Sonia excess amount” at the bottom of the waterfall. Read the docs! And if you see any other weirdness, send it my way.

Under the microscope

Quiz question — what’s the smallest securitisation currently outstanding? I’d have probably guessed some kind of structured funding for a startup fintech nobody had ever heard of, but it’s kind of a trick question.

The UK government, via the Prudential Regulatory Authority, actually defines any individual mortgage loan covered by the government’s own Mortgage Guarantee Scheme or by Gallagher Re’s “Deposit Unlock” guarantee as an individual securitisation, so the actual smallest securitisation out there is probably <£100k? Maybe smaller?

Both of these schemes do indeed act like a tiny SRT trade, giving a mezzanine guarantee on the mortgage loan between 80% and 95% LTV. Guarantees for this portion are particularly relevant because bank capital rules impose a bit of a capital cliff once loans go above 80% LTV, giving a powerful disincentive to lend to this segment cheaply. 

That’s why several of the big clearers (Lloyds, Santander UK, Barclays) have executed some form of risk transfer or portfolio sale of high LTV mortgages (Syon, Blizten, and Pavillion respectively).

Just like an SRT, once there’s a guarantee in place, a bank can substitute the risk weight for the guarantee (the UK government at 0%!) for the 75% capital weight which would otherwise apply to this portion of the mortgage.

The PRA, in its wisdom, does exempt these micro-securitisations from the full force of regulatory disclosure obligations. Banks lending Mortgage Guarantee loans do not have to notify the regulator individually for every single mortgage they write.

But it is adamant that each loan does indeed constitute a securitisation — and that’s bad news, because resecuritisation is still banned in the UK, so these guaranteed mortgages can’t form part of a securitisation collateral pool. 

At the time of writing, only Nationwide, Newcastle Building Society, and Yorkshire Building Society (via Accord) have signed up to “Deposit Unlock”, a private sector guarantee backed by the housebuilding firms and applied to new builds. Two of these institutions are active users of securitisation, through the Silverstone and Brass shelves, though these are large enough that they don’t need to throw guaranteed loans into their prime programmes.

But transferring risk on high LTV mortgages through full-sized securitisation is clearly a useful tool (as demonstrated by the clearers), and especially so for the challenger banks. Current UK proposals for overhauling the securitisation regulation inherited from the EU will see a slight softening of the resecuritisation ban (it will be allowed with regulatory approval, tbc how easy this is to obtain).

For now though, this seems like a bizarre regulatory own-goal. The best thing for first time buyers in the UK would probably be some kind of epic house price crash which came with absolutely no other economic spillover, but in the absence of this elusive possibility, anything to ease the mortgage environment would help.