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

Just resting

Europe had exactly one CRE CLO, and now it has none. When Starz Real Estate issued Starz Mortgage Securities 2021-1 in October 2021, it genuinely promised to open a new capital market. Other real estate debt funds, with bigger brands than Starz, were waiting in the wings to issue and hoping for a clean deal to fire the starting gun.

This didn’t quite work out. The complicated dual waterfall and uphill struggle in syndication seen by Starz made follow-on deals more difficult, though it didn’t shut the market. Around a year later, there was a clear signal from arranger Credit Suisse that its £900m warehouse facility for Aeon Investments would lead to three further deals. But those have not materialised.

Now Starz Mortgage Securities 2021-1 has repaid.

CRE CLOs are perceived as risky instruments, and the intervening years haven't been kind. 2022 saw soaring interest rates (and Liz Truss), 2023’s elevated rates backdrop saw extensive handwringing focused on CRE (especially office buildings), 2024 was full of elections and 2025 has any number of worries about the US administration baked in. Blackstone, an institution with somewhat more heft than Starz (and a big US track record doing CRE CLOs) didn't issue a US CRE CLO between 2021 and March this year.

The performance of the Starz deal, though, should help to overturn this perception. All of the loans repaid in full, some with amendments and one with an extension, but there have been no material credit issues. Extensions can be “delay and pray” agreements hoping that something turns up, or they can be carefully calibrated breathing space to get a refi finished up, and the Starz portfolio seems more like the latter.

Bank of America had an excellent piece out in February comparing the Starz performance with that of 2021-vintage large loan CMBS, and it’s clear Starz did better.

Admittedly, there were only nine loans, so random chance could have thrown up a portfolio that's better on average than the broader CMBS market, but there are a few less random reasons why these loans might have outperformed.

First up, the mid-market lending Starz originated might well just be better. Mid-market corporate lenders often make this point when comparing themselves to large cap loan funds.

Debt providers get paid more, get better documentation, better information, and can step in earlier to make adjustments in more of a genuine partnership approach. The average document in a large cap CRE loan gives sponsors a ton of flexibility and precious few constraints, while mid-market borrowers can’t dictate terms to the same extent.

Starz kept the servicing, rather than outsourcing it, and kept the junior risk in the CRE CLO transaction, aligning interests around the performance of the transaction, while in a typical CMBS structure, the third party servicer is acting as an agent of lenders, for a fee; they don’t have much upside or downside risk.

The CRE CLO structure is relatively more complicated, and could be unwieldy in a restructuring situation. But these portfolios also get some benefits from granularity and diversification. Nine loans isn’t granular enough to rely on statistical modelling, but it does mean that any idiosyncratic meltdown in the portfolio would likely have left third party investors whole.

In the period since the Starz deal has launched, banks have been ever-more active in lending against similar kinds of credit risk, financing real estate credit funds through fund level facilities or whole loan repo. Starz itself has been obtaining much of its lending from NatWest, for example.

In theory, NAV facilities and repo alike both have elements of mark-to-market, while the crucial advantage of a CLO structure is locking in non-marked finance for a term. These distinctions are a bit blurrier in practice, since the marks on CRE loan facilities aren’t especially volatile, but the basic question for the possible return of the market is price.

Is the price differential between public takeout and bank facility enough to pay for the structuring and the headache?

Then there’s the incentive structure of the banks to consider. The appetite of banks to distribute risk seems to be heading in one direction. Banks with no SRT shelf are starting one, banks with big SRT programmes are adding more assets, banks are shedding non-core platforms and striking credit fund partnerships to move assets.

CLO tech is a complement to these broader trends. Infrastructure CLOs exist, CRE CLOs have existed, private credit CLOs have been issued. But all of these could be real, vibrant markets matching the scale of their respective underlyings.

Breaking through?

Celeste has been on the road, at Invisso/FT Live’s US SRT conference in Charlotte and in New York. But she brings news that despite the disruption to US regulation, a crucial initiative to smooth SRT issuance for US issuers is still on track.

Her story is here (behind the paywall but DM if you’re interested) but to me the irony stands out.

SRT regulation for US banks gets ever closer to overcoming the string of regulatory obstacles which have bedevilled the market for a decade (the September 2023 reservation of authority was a good start, but only a start), just at the point when the US is seceding from international standards, gutting financial supervision and regulation, and turning the dial on capital up to “YOLO”.

A healthy SRT market, with banks regularly distributing risk of all kinds to the fund management universe, requires a course somewhere in the middle of the regulatory spectrum — enough clearly articulated good quality rule-making to make capital relief certain and structures reliable, enough of a capital constraint to make it worth doing, but no punitive anti-securitisation discrimination and no accidental capturing of SRTs by regulation drafted for totally different purposes.

Different lenses

Different corners of the capital markets have different standards and expectations, and different ways to think about credit.

Leveraged finance bankers would be happy to get a 6x levered (debt to EBITDA) deal away and suck their teeth over an 8x. On a debt-to-equity basis, a typical corporate LBO is more like 2x.

Securitisation structurers pitching such an unambitious structure would be laughed out of the room; a mortgage portfolio wants a healthy 20x (debt to equity) at least. All else being equal, a HY-funded company with 2x interest cover is less than ideal, while an infrastructure asset with 1.5x is healthy and financeable. To an extent these are unfair comparisons, and there are good sound credit reasons behind them all, but at least some of it comes down to what you’re used to.

Assets which sit in the middle of the Venn diagram for several markets, like data centres, will tend to choose the market which gives the best terms. There’s a far more scholarly discussion of portfolio financings, CMBS and ABS for data centers from Clifford Chance here.

So we have the second data centre ABS out in Europe, Vantage Data Centers Germany 2025-1. Vantage was also the issuer of data centre ABS #1 last year, with UK assets, so there’s everything to play for to bring issuer #2.

In the portfolio is some €928m-worth of data centre assets across Berlin and Frankfurt, with the finishing touches still to be added to one of the Frankfurt assets. These are securing up to €720m of debt, though €80m of this is in variable funding note format, to be drawn once the Frankfurt centre is finished. Right now, net operating income is €43.7m, though the contract price escalates over time.

So senior debt of €590m is 13x earnings, which sounds bad (in the LevFin universe). But property LTV of 63% sounds very reasonable (in the CRE world), and exposure to a Microsoft contract (52% of rent) at a massive spread premium to Microsoft’s bond debt sounds amazing (considered as a lease ABS).

Microsoft isn’t actually disclosed anywhere I could see, but there are only so many triple-A corporates in the world, and Johnson & Johnson has yet to establish a cloud/AI business.

If you want to maximise leverage, it’s generally best to talk to securitisation people, who will structure deals in such a way as to make the best assets (the Microsoft contract) the most important part of the credit picture.

Still, an EU deal does present a different set of problems from the UK transaction last year.

The UK investor base has a long history of real money participation in securitisation-like structures, be they watercos, rail lessors, ports, airports, real estate platforms, or housing associations. There’s a comfort with structures which blend hard assets into something like a corporate bond, and these assets sit with regular sterling real money funds.

Both deals are genuine securitisation deals, with risk retention and fully compliant securitisation disclosures (unlike most of the rest of the UK secured corporate and whole business market), but it’s still the case that the asset-class lines are more easily blurred in the UK asset management community.

Europe lacks the same tradition and has few comparables. EU-domiciled WBS never really got off the ground, project bonds fizzled and died, and in general there’s far more separation between the distinct worlds of infrastructure debt, securitisation, and corporate bonds than in the UK.

The new Vantage deal also has a €50m class B, unlike the UK transaction, which placed only a £600m A-2 note at 225bps over Gilts.

Placing a €50m tranche in a hot asset class, in an undersupplied and constructive bond market should be eminently doable, but there’s still some discovery to be done for JLMs Barclays and Deutsche Bank. Fixed rate asset-backed product doesn’t really exist in Europe; perhaps the 144A docs will come in handy here.

It is about the money

Conference panels on specialist lending tend to emphasise the cosy relationships between finance providers and specialist lenders and the importance of a partnership mentality, especially for structures like forward flows.

So it’s reassuring to get some hard numbers which underline that what lenders really want is money, both in the form of low interest rates and a high leverage point.

The above is from a survey done by Interpath, in collaboration with JP Morgan, which runs a large specialist lending conference and either banks or aspires to bank a huge swathe of the sector — aided by its arrangement with Quilam Capital, which we discussed here.

Relationship with funder is a solid number three, though, so the fuzzy conference feelings are certainly worth something — but these figures are encouraging for banks or asset managers looking to build up their business in the sector.

“Funder reputation” matters, but it’s way down the table; show up with a bag of cheap cash, an attractive leverage point, and be nice about it!

It’s also interesting to me that speed of execution is down the table — I presume this reflects the fact that a prudent lender doesn’t let themselves get to a point where they’re in a rush. Refinancings will be done well ahead of time, capacity will be managed carefully, and timelines will be cautious.

Other slides that stood out were a stat that 92% of the surveyed firms were looking at renewing, amending, or increasing debt facilities over the next 12-24 months, and the slide below on potential M&A — which, in short, should be coming down the road.

Admittedly the survey was conducted by a firm specialising in arranging debt financing and advising on M&A, so it may not be a totally pure scientific reflection of the sector, but it’s nice to have the numbers all the same — I’d encourage anyone interested in the sector to take a look.

Industrialising CLOs

My colleague Michelle has an interesting article about the increasing use of “master warehouse” structures in the CLO market. In their simplest form, these are basically common terms agreements that can spawn a bunch of subsidiary CLOs without having to renegotiate warehouses each time. Managers like them because they cut down legal costs, banks like them because they lock in managers, and cookie cutter CLO issuance can more easily result.

This means giving up flexibility on documentation, but this flexibility isn’t really a benefit; managers often strive for consistency and comparability on documentation and approach, with investors preferring the reassurance of a tried and tested recipe.

These are not master trust structures, like you might see in credit cards or sometimes mortgages but….. would that work? It wouldn’t fit with current market practice, culture and rhythms, but there’s no in-principle reason why not.

A leveraged loan master trust wouldn’t be that different from a BDC, and it would smooth out some of the vintage-related performance in CLOs; you’d get paid on credit, not on how cheap loans/liabilities were at the time of issue, plus it would be easier to structure deals with firm maturities. On the other hand, right now CLO equity is long an option that’s probably being underpriced by CLO tranche holders…. so who’d want to give that up?

Much of the work in CLOs is about simplifying and commoditising a process (add leverage to a pool of leveraged loans), while diminishing in no respect whatsoever the handsome fees which can be extracted from the management of these loans. In Europe especially, there’s a ton of issuer overlap across CLOs, and, especially in periods of thin secondary liquidity, many deals are effectively buying the market and swallowing up whatever comes in primary.

For this, a manager can receive 50bps in fees and 20% of anything above 12% levered IRR. Nice work if you can get it, and it’s little wonder that there are virtually no loan fund managers without a CLO platform today.

This is a nice business to have, and there’s no desire to disrupt it, but anything that cuts costs further is welcome.

Our customers receive this content ahead of the crowd — find out more about 9fin’s news and analysis.