Stubborn refusal

Two facts about the European CLO market this year.

1) New issue volume is tracking pretty much in line with a “normal” year (not sure there’s any such thing really, but better than 15/16/17, slightly behind 18).

2) The arbitrage available at new issue has never been worse in the 2.0 era.

Depending how you count it (I looked at a graph of generic new issue excess spread produced by Barclays research), arbitrage even managed to briefly turn negative this year. CLOs are less profitable than at point in recent history, and it has barely dented volumes.

What sort of ridiculous market are we talking about here?

It’s instructive to think about how it differs from CMBS in this respect — European CMBS switches on and off (mainly off) depending on pricing. If there’s a better execution to be had syndicating CRE loans in whole loan format, or better sources of financing through loan-on-loan repo or fund-level back leverage, then CMBS stops dead.

That’s precisely because there are meaningful alternatives. CMBS is one small product in a suite of products designed to provide debt funding to CRE investments. When times are good in securitisation, it’s efficient, and can often deliver looser loan terms to sponsors, but it’s just one source of capital.

CLOs, meanwhile, are pretty much the only game in town for leveraging up leveraged loan portfolios. They’re vastly more efficient than the existing alternatives, which exist basically as precursors to CLO issuance, rather than standalone alternative markets. Sure, you can go get a total return swap on whatever leveraged loan portfolio you want, or maybe repo your book line by ISIN, but when the CLO market is operating, it will deliver better terms and a safer structure within which to manage a loan portfolio. The greater granularity and diversification in a CLO portfolio vs CMBS amps up the advantage of using securitisation technology.

There’s a few other things in the mix as well. Although volumes are tracking surprisingly well, the volume of “CLO classic” deals are minimal. If we’re talking 2021-style deals, with 1.5 year non-call 4.5 year reinvestment, broadly syndicated, with full capital structures and heavily ramped portfolios, there’s been very little since February. The CLOs of 2022 are not the CLOs of 2021.

The unusual deal structures seen this year — static, print and sprint, short WAL, low ramped, delayed draw cap structure, turbo notes, whatever — all embed different assumptions about the future of the market. Where are loan prices today, where will they be next year and the year after, and where can CLO liabilities be financed or refinanced?

Very few of the deals, however, embed the basic assumption of a functional CLO market — that loans can bought at a spread 200+ bps outside the WACC of a CLO structure, so owning this portfolio on a levered basis will make good money.

That ought to imply more variation in management style. If the European leveraged loan index vs CLO spreads delivers an attractive arb, then it’s possible for managers to sotto voce buy the index. No need to be a hero — keep that diversity score high, keep in with the sellside, clip your fees.

In today’s troubled times, that doesn’t work. Monkeying around with the liability side only gets you so far; you need to find something special on the asset side too, be that illiquids, concentrated conviction positions, working second lien and bond buckets hard, active workouts (of which more later). The par loan investor’s creed used to be “don’t lose money”, but when the arb is this skinny, you need much more.

Paper profits

CLO managers have been flirting more actively with bond buckets in Europe this year — first as the rates selloff started in March, and then more latterly as bonds lagged the loans rally through the summer. In some respects it’s a no-brainer — if you’ve got the same credit at the same seniority level for a cheaper cash price….that’s free money?

But the hiking cycle is well underway — we all enjoyed the 75 bps decision — and these bonds are not heading back to par any time soon.

The structure of the CLO vehicle, however, creates a certain amount of weirdness in this respect. Last week we dropped in on S&P’s European Structured Finance conference, where Invesco’s Nuno Caetano described this effect as “cosmetic cheapness”.

Pretty much everything baked into the CLO docs references par values, so buying discounted bonds looks fantastic — they’re marked just the same as higher priced equivalent loans when it comes to, say, interest rate diversion tests — but real world CLO trading often looks to Market Value Overcollateralization (MVOC) or equity NAV as a starting point (for which bonds are decidedly unflattering).

The unrealizable bond gains can flow through to how best to rework the standard issue CLO for the current environment. Is it worth, for example, doing a short WAL deal if you’re planning to buy a ton of bonds? Come refi time, the bonds won’t have traded up — you’re not profiting from a market dislocation, you’re just taking interest rate risk. You do well when the sponsor refinances the business, but that could be in 2025 or 2026. It’s a paper gain that might not be much use in practice.

Also coming out of the S&P event was a pretty interesting note on the current state of the art in CLO documentation.

We already talked a little about the “Uptier Priming Transactions” language; language allowing participation in “Asset Drop Down Transactions” pretty much falls into the same bucket, tooling up CLO managers in anticipation of a round of aggressive sponsor behaviour to come. If restructurings are going to get more complicated, and leveraged loan docs have more loopholes, CLOs need more tools beyond “sell and get away”. Especially if, thanks to the increasing prevalence of whitelists, sale isn’t an option.

“Asset Drop Down Transactions” are what everyone else calls “Getting J.Crew’d”, after the US retailer which pioneered the move, but it kind of comes to the same thing as priming from a lender’s perspective — someone else is getting a bite at collateral you once had. Might as well be you.

The LBO loan backing KKR’s purchase of bike company Accell has commitments due on Thursday, and it’s quite topical in this respect — it added a “J.Crew blocker” to loan docs during the syndication period, but tweaked the standard language to make it much weaker and more limited. See our coverage (shoot an email to team@9fin.com if you’re not a client).

Currency costs

S&P also observed that fewer new deals were including “Form Approved Swaps”, as a cost-saving measure. This piece of CLO esoterica was new to me, so it required a little enquiry, but basically these means pre-agreeing the form of a cross-currency swap with potential counterparties and with the rating agencies, such that CLO managers can purchase non-euro assets without going back to the agencies for approval.

If this saves cost, it seems a little short-sighted — surely one of the ways CLO managers can differentiate themselves is looking outside the euro-denominated mainstream, though dollars is unlikely to help much, as dollar assets often trade tighter than euro equivalents. But raising the bar to buying non-euros risks worsening the already heavy currency bias of the European leveraged finance market.

Sterling or Swissie loans have become an increasingly awkward proposition as institutions have supplanted banks as the main lenders. UK borrowers make up more than 15% of the European leveraged finance universe, but sterling-denominated loans clock in around 3%, in part because lots of UK borrowers have extensive international operations they want to naturally hedge……but mostly because the sterling lender base is small, tightly knit, and expensive. In a currency-neutral market, you might expect borrowers based outside the UK with substantial UK operations to also bring sterling supply, but in practice they rarely bother.

CLOs need perfect asset swaps to hold non-euro assets — these are expensive, and offered by few counterparties, and so sterling loans need to offer some extra juice to pay for the hedging costs. There was one bold and beautiful sterling-denominated CLO (PGIM, arranged by NatWest Markets, or possibly RBS at the time), but this rare creature has now redeemed.

But if fewer deals are pre-agreeing swaps with their counterparties and the rating agencies, this raises an administrative, as well as an economic barrier, to doing non-euro currencies. It may no longer be worth the brain damage, even if the spread is compelling.

The Carolean Age in ABS

If enough people believe in something, perhaps it becomes true? I wasn’t sure I believed in a proper September reopening, and yet we’ve seen a fat new issue calendar featuring auto deals from the UK, Spain, France, Finland and Germany, plus three UK RMBS deals spanning the full credit spectrum, and a Dutch RMBS announced Thursday.

ABS doesn’t live in a vacuum — this comes on the back of IG corporates and FIG firing for a couple of weeks — but this looks like the real thing.

Several transactions came out with some element of derisking, be that full preplacement and loan notes in the case of Oodle, an option for partial retention in Friary No. 7, or announcement a week prior to general syndication (Autonoria 2022).

Given the parade of terrifying energy headlines and the thin liquidity all summer, it makes sense that arrangers would be cautious about their approach to market but prior to Tuesday afternoon, book updates were suggesting a pretty normal market was back on — healthy 1.5-3x type coverage at the senior level, and some wild early January-style mezz books (9.9x on Autonoria class D, admittedly only an €11m tranche, is the week’s record).

All was well, therefore, until the US CPI print was released at 8am Eastern Time, prompting all hell to break loose in US equity markets — the worst single day performance since June 2020, S&P down 4.32% and Nasdaq down 5.16%. European risk markets took less of a knock, but it was certainly not the ideal backdrop to tie off new issues.

ABS has a certain natural resilience to these situations though. Spreads move more slowly, primary books generally feature less fluff and momentum-chasing than other markets, orders placed tend to stay firm. An ominous delay between BNP Paribas’s announcement that Autonoria was going subject down the stack on Tuesday afternoon and final terms on Weds betokened not trouble with the deal, but a €100m upsize in the works.

We should, in fact, talk about BNP Paribas, since a remarkably large portion of the recent supply seems to have emerged from the Marylebone vicinity. Last week BNPP arranged Tommi 3, this week it was co-arranger on Friary 7, joint lead for Dowson 2022, and of course led the deals for BNPP’s auto subsidiaries in Spain and France (Autonoria 2022 and FCT Pulse France 2022). The only deals BNPP hasn’t been on were from an Societe Generale subsidiary (quel surprise) and a Goldman-sponsored RPL deal (also a difficult mandate to win). It’s also leading Hops Hills and Dutch MBS, both announced Thursday after the other deals had cleared the market.

The French bank does a lot of deals, but this rush seems like a signal of strong conviction — the market is there, but it might not be there for long, so hit this window as hard as possible.

Putting Autonoria and FCT Pulse out together seems particularly curious — two BNPP-backed auto deals, with senior spreads less than 10 bps apart, pricing on the same day (I got the priced message three whole minutes apart). Autonoria was a full capital structure, while Pulse was senior-only, so this won’t have been a worry for the Autonoria mezz, but again this looks like a very clear market read….go now and get it done, don’t fritter the autumn away looking for a competition-free window.

The levels for this flurry of deals are what they are — 90 bps for Finnish autos would have caused much wailing and gnashing of teeth in 2021 — but despite the predominance of autos, it’s not like all the trades here are super vanilla and straightforward.

With so many jurisdictions of auto deal in market at once, we can pause to ponder the fact that Finland (90 bps) is now pricing outside Spain (84 bps) and France (75 bps) and now miles outside Germany (56 bps). There are differences between the shelves and the structures for sure, but Finland used to trade just back of Germany. Tommi 2 in February came at 21 bps, Volkswagen’s VCL 35 (Germany) at 15 bps.

One could argue that Finland is at greater risk of being invaded by Russia, but perhaps more worrying is the rising rates backdrop, and the fact that the fortunate Finns pay less for their car loans than most countries do for their mortgages — the Tommi 3 portfolio pays a weighted average interest rate of just 3.1%.

The mismatch between this level and the market has forced hard credit enhancement up (11.2% vs 8.7% in the last trade), but excess spread in an auto deal can smooth out a lot of problems, and Tommi is running short. It’s also a smaller sponsor, and perhaps a more niche audience at the senior level — big ticket bank treasuries may feel better in one of the BNPP or SG shelves out there.

Oodle, as we’ve discussed, is seen as one of the originators at the sharp end of potential UK consumer stress. That’s reflected in the 1200 bps spread at the bottom of the stack, but, no matter. Even paying away that it can support an excess spread note with a 900 bps coupon. In a time of rising rates, the originator writing 17% loans can stand tall.

The class ‘G’ note at the bottom of the Autonoria stack also landed at 1200bps, some way back of the 5.25% achieved at the last outing in 2021, but that seems pretty reasonable against the broader market — inside corporate risk transfer, inside CLO single B, way inside CLO equity.

So it’s fine for investors, but what’s in it for BNP Paribas? The bank has plenty of capital, is unlikely to have a concentration issue in Spanish auto lending, so why’s it paying away 1200 bps for cash risk transfer? Maybe there are relationship reasons to keep the shelf active (but then why the upsize?), but more important is probably the question of how the group manages its various subsidiaries and entities.

Autonoria and Pulse both come from standalone entities (Banco Cetelem and Arval Leasing respectively) and BNPP may have sound group-level reasons for its distribution of capital and liquidity resources among its different arms. Arval Leasing (sponsor of FCT Pulse) followed up its ABS debut by heading straight out into the senior market, launching an four year unsecured deal at MS+175 bps on Thursday, so there’s clearly some desire to fund and raise capital independently.

But the bigger picture is that banks might not feel like they’re as well-capitalised as they’d like. The larger European institutions are very active in synthetic risk transfer at the moment — most of the big banks have programmatic SRT shelves anyway, but they’re definitely eager to tap this capital resource. The recession is on its way, it’s going to hurt, and it’s better to source capital when and where it’s available now rather than wait until you’re forced to later on.

Vive la Revolution

We have discussed the difficulties facing specialist lenders a few times around here, primarily looking at the UK — the mortgage market is more diverse and fragmented here, and rates rose earlier than in the eurozone.

But the situation in France has got the mortgage brokers out on the streets protesting. It’s not just a greater cultural propensity to man the barricades….France is facing a particularly absurd usury law which essentially forces banks to run behind a rising rates cycle. We talked about it a bit but you still can’t do better than follow Johannes Borgen on Twitter for all things banking.

Perhaps the UK lender community should be taking more drastic action?

What do we want — ORIGINATION LEVELS TO KEEP PACE WITH RISING SWAP RATES, when do we want it — PRIOR TO COMPLETION.

How expensive is Rizwan?

The conspiracy theory about Rizwan Hussain is that he’s a rogue agent working on behalf on the legal profession, trying to ensure the maximum possible diversion of proceeds from securitisation transactions into the pockets of lawyers. He’s certainly boosted billable hours by an impressive amount — every nonsense claim or counterclaim requires a proper response by a decent class of law firm, and there have been more than I can possibly track (still working on hiring that Rizwan correspondent).

Fitch even downgraded the Business Mortgage Finance deals on the back of the excess legal costs incurred by those transactions, though the issuer refused to ping me the latest investor reports for the latest figures (if anyone wants to scrape them off Bloomberg for me, I’m on owen@9fin.com).

Davidson Kempner, as discussed, has beaten off more Rizwan attacks than most, including during the marketing period of its Stratton Hawksmoor 2022 transaction. The announcement of this refi prompted a flurry of lawsuits filed by Rizwan (we counted five when we last looked), as it refinanced three of his favoured target transactions.

It even prompted some hefty anti-Riz provisions baked into the docs — apparently added during the marketing period.

First off, DK commits to covering the legal costs of fighting any Rizwan court cases (so it won’t come out of the waterfall as an issuer fee). Second, it’s given a hefty £45m indemnity to the issuer, in the event of any successful judgement against it. Rizwan is holding steady at a track record of zero successful court cases, so this is unlikely to be called, but underlines the scale of the potential damage.

Even here there’s another legal hedge….even if the issuer receives a judgement against it, the Retention Holder (DK), still has the option to purchase the loans. It might cost money, but DK is not going to lose control of this portfolio.

Use of Proceeds gets weird

Kensington’s “Green” labelled Finsbury Square deal has come up a couple of times in my buyside conversations of late, as it illustrates some of the oddities of applying these labels to securitisations.

At issue in May 2021, Kensington didn’t have sufficient “green” collateral to max out the green-labelled funding it had raised, but no matter — the green label refers to “use of proceeds”, not the deal’s collateral, and as long as you reinvest the money in funding more green origination, no problem.

The deal explicitly contemplated this approach….from the investor book: “Given that the total amount of Eligible Green Projects on the issue date is expected to be less than that of [Green Bond] proceeds, Kensington intends to originate loans (secured by Eligible Green Projects) having an aggregate nominal amount equivalent to the amount of unallocated proceeds within a maximum of 5 years from the issue date”.

The difference between the green origination at issue and the future plan was pretty big — at issue, there were £68.1m of “Eligible Green Projects in the Provisional Mortgage Pool”, giving £571.2m of unallocated proceeds at closing. The deal also comes with a controlled amortisation feature and revolving period, allowing Kensington to add other loans into the pool for the following 4.5 years.

These do not, however, have to be green loans.

Indeed, Kensington said that these “will in most cases, not be transferred to the Initial Loan Pool and may be used in other future securitisations”, though it commits to avoiding double-counting. If green origination is funded by the first green bond, these mortgages cannot be counted as ‘eligible green projects’ for a future green deal.”

But Kensington’s sale process, with its front book / back book split has added a twist to this arrangement. Barclays has bought the business, subject to regulatory approval. That includes the flow of new origination and the unsecuritised warehouses, while Pimco is in process for the back book, including the loans within the Finsbury Square transactions.

That means the Finsbury Square green deal isn’t being topped up with new origination as was originally intended — and the capacity/obligation to do new green origination will pass to Barclays, rather than the eventual sponsor of the deal when the back book sale closes.

Quite a lot has already been done to close the gap though. Per the last reporting date at the end of May, the original £68m green allocation had climbed to £398m, with £242m of proceeds still to allocate. Perhaps by the time the deal clears regulatory approval, we’ll be at 100%, so even if Barclays pulls the green product line (why would it?), it will still count as a green UoP bond fully ticked off.

Does any of this actually matter? Investors in the deal that bought to a five year controlled amortisation are seeing faster CPRs than otherwise expected, so they’re probably happy with the situation. It just feels a little strange. Securitisation investors are used to mostly thinking about asset pools – the assets they finance are the ones right there in the SPV, disclosed in the loan tape. “Use of proceeds” language for green bonds disrupts this close connection, even as it makes it far easier and more flexible for issuers to print green bonds.

To put this another way — from a credit perspective, the FSQ Green investors are exposed only to the loans in FSQ Green. But invisible ethical tendrils of ESG connect them to a different asset pool, made up of new hypothetical mortgages, written by an institution which will one day no longer be connected to FSQ.