It’s interesting stuff! Nothing like the smell of napalm in the morning to make markets exciting again. Should you desire more SVB coverage, I did one piece on Friday and a follow-up on Tuesday.

Anyway, at Excess Spread we’re ruthlessly focused on the business of securitised products, and so we’re going to cover it from that angle, and try to consider what’s in the loan book of SVB UK, acquired by HSBC on Monday morning for a solid English pound. 

I should stress that there’s absolutely no information it. The consolidated group accounts are singularly unhelpful in assessing the UK sub, and HSBC hasn’t opened the kimono since acquisition. But we can make some reasonable assumptions.

The overall SVB loan book is dominated by capital call facilities, predominantly to VC and growth equity firms. We’d expect that to be mostly done out of the much larger US entity. The US venture capital industry is much larger than that in Europe; even the local investment operations of the global funds probably did much of their fund banking through the US. Other lending includes venture debt, private banking, lending to early stage companies, and lending to vineyards, all relatively small industries in EMEA (well, not the private banking).

What we’re really interested in, though, is whether there’s much warehouse financing. SVB’s EMEA business certainly had bankers out there covering startups and offering warehouse lines — Folake Shasanaya went from RBS/NatWest Markets’ securitised products group to become “head of EMEA warehouse lending” in 2019; she then became head of “new and strategic channels”. The team also included Conor Sheehy, once in Macquarie’s SPG business in London, and Matthew Nyong, who’d worked in the treasury at securitisation issuer Prodigy Finance and rated securitisations at DBRS. It was actively hiring, prior to the weekend’s events!

In short: these are securitisation people and they were offering securitised funding. The total SVB UK loan book wasn’t large but it wasn’t small either; £5.5bn-equivalent, based on HSBC’s takeover announcement. 

But perhaps SVB UK’s warehousing operation hadn’t really hit its stride. Bankers at more traditional securitised products outfits told us that they rarely, if ever ran up against SVB in the competition for a financing — every big shop, though, spent the weekend running through their possible exposures. 

Startups in general love press-releasing their funding rounds, whether debt or equity, so you can get a decent feel for the presence or absence of a bank in fintech warehousing simply through motivated Googling. 

That throws up a few SVB debt facilities — Canada’s Properly, embedded finance specialist Liberis, raising €30m, global trade finance platform MODIFI, electric vehicle lender Tenet, Mexican SME export finance platform Mundi, US immigrant financial services platform Stilt. Of these, only Liberis looks like a nailed on European facility which might have been funded from SVB UK, and that’s a very small ticket.

Still, maybe small tickets were the SVB sweet spot (despite the lack of press release trail) — the full size investment banks can’t make the numbers work on structuring deals in the $30m and below range. Certain other specialists — Hamburg’s Varengold Bank, for example, might be more regularly found fishing in this pond, or perhaps the competition comes from the non-bank side, from asset-backed private credit operators like Fasanara Capital.

Anyway, it’s certainly possible that HSBC is acquiring a rather nice client list in various forms of asset-based lending and high growth fintech, be it consumer, SME, revenue-based financing, asset-based lending or other products.

That’s not a natural fit at first with the HSBC securitised products group. It’s enormous in corporate securitisation; this is, after all, basically corporate banking and it’s one of the world’s largest corporate banks, using its Regency conduit as well as its own balance sheet to support the business. The financial institution end of securitised products skews more towards banks than most of its competitors, with securitisation offered alongside covered bonds and other funding. But the sponsor and hedge fund community tend to look elsewhere for their securitisation banks — CarValbeing a notable exception, giving HSBC a string of mandates for the DPF deals from its RNHB portfolio company in the Netherlands, and lately a sole arranger mandate for Spanish reperforming loan transaction Miravet 2023-1.

But if HSBC can connect the dots and help to bring a bunch of small cap FinTechs to the big time, with some of the lowest capital costs in the business, that could revolutionise the look of HSBC in securitisation. HSBC, predictably, declined to comment.

Losing side

CLOs always seem to be on the wrong side of the elegant dance between leveraged loan supply, leveraged loan prices, and CLO liability spreads. Perhaps in the long run these factors self-correct, but ever since the Russian invasion, it’s seemed out of balance, to the detriment of economics in the CLO market. Liabilities are faster to sell off than loans when there’s a shock, and loans are faster to tighten than CLO tranches when the market improves. This has also proved to be the pattern in the shockwaves from the SVB collapse, which blew out CLO triple A 20 bps-30 bps, leaving senior bonds back the wrong side of 200 bps, while loans barely moved.

Neither market has much direct exposure. You can construct various causation chains which might change things — SVB collapse means less tightening, lower terminal rate in the US, new bank funding facility is de facto QE revival, so stimulus for US consumer products, lower base rates, better interest coverage on cross-border LBO capital structures, less exported tightening by the Fed? So that’s good? Maybe? Unless it ruins the fight against inflation, in which case higher rates longer term, and reverse all of the above.

There’s a lot of macro going on, in short, but you’ve got to squint pretty hard to identify definite ways it flows through to leveraged loan credit quality and hence to CLOs. The flow through to triple-A CLOs is even less obvious.

So it’s market technicals in charge. The CLO liability investor base in Europe is fairly narrow and highly attuned to relative value across fixed income products. 

The leveraged loan investor base is, well, mostly CLOs, but there are plenty of managers with slightly different motivations, and enough deals are newly priced and ramping up to keep a solid back bid in place when there’s a widespread credit freakout. CLO tranche investors don’t have to buy; CLO vehicles themselves do. 

Every day post-pricing that they aren’t fully invested is a drag on returns. My loan reporter colleagues have been tasked with seeing the fallout in European loans, and are coming up empty-handed — witness the syndication for CD&R’s Motor Fuel Group, a UK petrol station chain which was pushing out maturities on £765m and €1.087bn of 2025s. Admittedly, being an A&E, there’s a very limited new money component, but it does fund a small dividend for a sponsor which has already taken a ton off the table. Despite this, the execution looks pretty textbook, with OID set to land at the tight end of a revised range and the deal upsized a little to cover this OID.

To the extent that there is a knock-on impact via loans, it’s probably also in a way that’s bad for the CLO market. Several loan deals are in early-bird, and there’s at least one chunky underwrite (the €2.9bn-equivalent for Advent’s purchase of DSM Engineering Materials) that was originally slated for Q1. With the standard two week syndication on an LBO for a new credit, we’re bang out of time already.

The other recent source of loan supply has been from BWICs, which have been running at a high rate lately. A couple of these have been old CLOs getting called (2015-vintage deals from CQS and Alcentra), but most have been driven by third party equity liquidating old warehouses. Where the takeout’s no longer attractive but the loans were bought cheap, better to lock in the price appreciation and move on with life.

Loan prices might not have moved much, but febrile market buffeted by rumours and headlines might still not give best execution. There’s a difference between stuck prices on zero flow, and a comfortable cushion of competitive demand to take down BWIC supply. So the smart money funds, the likes of Elliott and Anchorage, might decide to hold off on bringing any further warehouses out.

Silver lining

Over the medium term, you can squint and see a route to a stronger CLO market off the back of SVB’s collapse. One trend that’s being discussed is a major reallocation of deposits from the smaller regional banks to the money center giants. This might be short-term irrational…..SVB and Signature Bank deposits are now the least risky banks in the world as counterparties, thanks to the Federal Government...but forcing every small and medium company in the US to wake up to counterparty risk is surely going to leave JP Morgan, Citi, Bank of America and so forth absolutely awash with liquidity.

So one might expect that the treasury/liquidity books of those institutions will very soon be on the hunt for a place to park the cash. Most of the answer is likely to be rates products, presumably disgorged by the regionals as they fund the deposit outflows, but all the big banks have been sizeable CLO players at various points and on both sides of the Atlantic.

And I do mean sizeable. A quick look at JP Morgan’s latest numbers gives $629bn total in the CIO book. JPM had $5.9bn of CLOs marked as “available for sale” (i.e. M2M, trading book) and $61bn as “held to maturity” (non-M2M). On the HTM side, that’s more than 3x the muni book, and 14% of the total.

Life and deadlines are too short for me to pore over the regional bank balance sheets, but I’d assume the money center institutions have a greater propensity to buy CLO tranches than regionals – it’s kind of a specialised skillset, though that said, there are plenty of medium-sized European institutions who play in triple A.

So if JPM gets a $50bn deposit inflow and keeps the same ratios, it’s buying $7bn or so of predominantly triple-A — that could mean anchoring 28 deals additional to the deals it would have already played.

The numbers are cocktail-napkin sketchy, but I think the point stands — a meaningful reallocation of deposits to big banks potentially translates to a meaningful increase in demand at the top of the CLO capital structure. For the short term, market fear, volatility and headlines rule ok, but medium term there could be a wall of money on the way.

Gorging themselves

Despite the ropey secondary picture, we did see some interesting primary, with the year’s first debut manager in Europe, Canyon Capital Advisors, finally printing via Jefferies on Monday. The deal is named Canyon Euro CLO 2022-1, though this refers not to a long and painful process in coming to market, but to the fact that the warehouse was opened in July 2022. 

At the launch of marketing, the deal was around 40% ramped, suggesting Canyon was nibbling rather than binging on the cheap loans available in the autumn. The execution levels are not at all bad considering the backdrop, with last week’s 180 bps market standard on the triple-A (though, in fairness, syndication was pretty well locked up last week, with all tranches but the double B subject)

The double-B, indeed, is the notable outlier, with most top tier managers pricing this tranche in the mid-800s over the past month, vs 940 bps in Canyon. But fair enough! There was a major banking crisis running at the time!

Canyon is a well known US CLO manager, with $6bn or so under management across 13 deals. It’s got a epic pedigree as a distressed credit hedge fund (here’s a 9questions interview with Chaney Sheffield from the distressed side) founded by (inevitably) Drexel alumni, but now invests all over the capital structure, and in real estate, equities, securitisations and much else. So there’s not much case for a huge debut premium on the grounds of experience. 

But a new CLO manager in Europe still has to answer questions about platform longevity, sustainability, resourcing, sponsor and sellside relationships, and business plan, and syndication is usually longer and more involved that that of established managers. 

Investors don’t want to see a setup that leaves Europe as a semi-neglected branch office dependent on the US for credit work and the big decisions. Business plans matter because CLO M&A matters; this can upend carefully drawn assumptions about manager style. It’d be hard to argue that, say, Harbourmaster and Avoca have suffered under Blackstone and KKR ownership, but it’s also a qualitative risk that investors struggle to plan for. 

As CLO 2.0 opened up in Europe a decade ago a few people were sniffy about the managers coming out of distressed debt shops, but at this point, nobody would argue that Angelo Gordon, Anchorage Capital, Oaktree et al weren’t delivering on their par investing promise.

Structurally there are a couple of wrinkles but pretty vanilla overall; a split triple-B tranche a la Redding Ridge, delayed draw single-B (some managers are now placing this again)and fees collected through a €10m subordinated management fee note (also common in Redding Ridge deals, but essentially a function of internal fund structuring and allocation).

To split or not split the triple-B is something of an “angels dancing on the head of a pin” exercise — it depends very much on individual investor demand and motivations. Eyeballing the Canyon capital structure it seems like a no brainer to prefer the D-2 — an extra 100 bps on the discount margin (675 vs 575) and more convexity (97.5 vs 99 cash price) for giving up a rating notch. But the IG cliff risk is real (at least for some investors). The D-2 is rated BBB-/BBB, so it’s only a minor market wobble away from junk status, and that’s a risk some account don’t want to take.

Triple Bs quite frequently go to a single fund, so it’s quite likely that in this case, the two D notes have been tailored to whoever’s in the bonds. ICG’s triple-B landed at 615 bps, so the blended cost of <600 bps shows that it’s worth doing.

For a clue about what kinds of investors often anchor the mezz, turn to the always-excellent TwentyFour Asset Management blog. Last week’s edition lifts the lid on ICG’s recent deal ICG Euro CLO 2023-1, discussing the enhanced protections this transaction offered for mezz accounts, particularly around WAL extensions, consents, and post-RP tranactions. 

It doesn’t quite say “we pushed for this protection and now we own these bonds”, but TwentyFour has a long track record as a fairly active mezz account in CLOs, with strong opinions on docs and structures. It shows up in sufficient size to get its views heard and stips in place, so I doubt it’s a total coincidence that ICG got this glowing review.

“We think this is great governance and a really positive development and speaks for ICG’s longer term commitment to all investors. And we encourage other managers to follow their example and hope that all debt investors maintain a firm stance on documentation.”

Whether TwentyFour is talking its book or not, I’m inclined to agree. Mezz has historically got a raw deal, coming in late to the CLO syndication process once triple-A has got its stips in, and one consequence of that has been that extension risk protections for mezz are riddled with holes. Might be worthwhile fixing that.

So Canyon shows that debut managers were possible, in the benign markets of last week, but it seems likely that the next round of debuts will be on pause until the dust settles. M&G, we understand, would have liked a Q1 deal but that seems unlikely; the visible pipeline was already dwindling before the SVB collapse, and it’s not likely to fill up again in the short term.

Over the parapet

West One was supposed to be the market reopener for UK specialty finance, when it announced Elstree Funding No. 3. 

February had seen a ton of euro supply, with a bias to euros and STS (simple, transparent and standardised) structures. But UK specialist resi, a crucial securitisation client base, had been pretty well out of the market, aside from Belmont Green at the beginning of the year.

However, Enra/West One had the misfortune to announce in the middle of the day on Thursday, just as the SVB run was getting under way (the fateful capital raise / bond sale combo was released to the market at the end of Wednesday, and the shares were already plunging).

This was not the ideal backdrop to execute a deal! 

The immediate read-across from a west coast banking collapse to UK RMBS performance is….obscure, but markets were in bad shape and whipsawing all over during the execution period. Credit puked on Monday, recovered on Tuesday and puked again harder on Wednesday, as doom-landed headlines about Credit Suisse ripped through markets. Then the Swiss National Bank showed up with a CHF50bn bazooka and we’re bouncing back again.

Chaos stalked the land, in other words. At the first book update on Wednesday morning, the deal was nicely covered at mid-200s/mid-300s/mid-400s down the stack but the final pricing came 50 back for class B/C/D (300/400/500). This is not especially hard to parse; Crossover was 50 wider, asset managers wanted more, and they got it.

But announcing the deal with a big anchor in the triple-A proved to be an absolute masterstroke. It’s likely bank-type money, but it stayed in place, and it stayed at 125 bps — relatively tight pre-announcement, miles through the market by the time of pricing the following Thursday afternoon.

When issuers are broadsided by events like this, they can opt to press ahead or postpone, and Enra has received plaudits for pushing on — the tight senior keeps costs in the overall structure manageable, the warehouse is emptied out, and the treasury team can get on with their lives.

It’s not an execution that will encourage other specialist lender to rush to market (there may not be many more reverse enquiries at 125 bps), but getting a deal away in the teeth of a literal banking crisis is no mean feat. The next few trades could revert to preplacement, and well-funded issuers might press pause, but as one of Enra’s peers commented… “Big respect”.