At the time, we were in awe at the resilience of the market, in the face of an arbitrage that was at historic lows. Managers kept printing deals, gathering AUM, and getting on with it, despite the poor economics this seemed to lock in.

There were manager-level incentives for this behaviour — terming out warehouses, proving a business plan, gathering assets, deploying committed capital — and there were market incentives too. The show must go on.

But there are more nuanced considerations as well. 

A CLO should be considered over its lifetime, not just the equity return at the outset. A better liability environment which allows the 2022 deals to be reset could transform this vintage, and there are already some moves in this direction (though for Q4 2022 deals)

Apollo’s Redding Ridge unit said in July that it would reset Redding Ridge Europe 14 in October, while Barings Euro CLO 2022-1 published an “exploring optional redemption” notice this week (presumably for November when the non-call is up).

The Barings deal had a very unusual 1NC/1RP maturity structure, so was presumably always intended to be refinanced this year. It was slipped out in October, in the wake of the LDI sell-off, with no pricing / DM info, just a list of tranches and coupons. Coupons were 199/432/502/679/803, so it’s already way in the money vs today’s market (Bridgepoint landed this week with coupons of 180/260/355/500/722). 

Subsequent inspection of the docs shows that every Barings tranche was placed below par (99.5/97.5/96/95/88), so resetting this deal will prove an expensive exercise, but probably worth it if the final ramping efforts were done in the bargain basement aftermath of LDI.

Fixed rate exposures are the other big caveat to throw in to the picture of poor performance for Q2 and Q3 deals last year. Lots of managers saw value in loan-bond switches as the rate hikes took hold, selling high priced loans to buy cheap bonds from the same issuers, and loading up new issues with as much fixed rate as they were allowed to buy, in some cases enlarging their bond buckets vs their usual limits.

This is effectively trying to benefit from interest rate moves using a vehicle that’s designed to be rates-neutral, so monetising it is difficult. The trade relies on issuers refinancing out their old bonds, so is inevitably back-loaded. Issuers with loan and bond maturities have generally preferred to address loans first through A&E processes, since there’s a bit of game theory about how much capacity CLOs have to extend.

Anyway, the point is, if fixed rate was your trade, it hasn’t yet played out. The interest might be disappointing for now, but the principal will be juicy when it comes.

My former colleague Kat Hildago wrote a nice piece for Bloomberg on the “snooze-drag” element in loan A&Es — basically, allowing CLO managers which aren’t able to formally consent to a maturity extension to abstain and nonetheless get rolled into a longer loan with a new market coupon.

We wrote a bit about this in January, as part of our deep dive into CLOs and A&E requests, but the tricky part, as Kat notes, is whether managers want to annoy their triple-A investors by agreeing to be snoozed. 

Much depends on the situation of each deal; vintage, performance so far, equity placement, triple-A placement. But equally important is the current market backdrop. If you’re a manager with big plans issuance plans which rely on the goodwill of the senior investor community, keep them sweet.

The business that free beer built

9fin is a current WeWork tenant, and I can thoroughly recommend the combination of free beers and free barista coffees. I imagine they’re motivated to close deals at the moment if any of my readers are looking for space; the going concern warning has made headlines around the world.

WeWork isn’t huge in European CMBS, which itself is not a huge market, but we do wonder how exactly this will play out. Of particular interest is Devonshire Square (a favoured 9fin lunch spot just round the corner), which is securitised in Taurus 2018-2 UK.

This is a deal which is both anchored by and partly sponsored (10%) by WeWork. It’s also in what are presumably the final stages of refinancing. The original loan, funding the purchase of the property by WeWork, Nuveen and Danish pension fund PFA, had a 2021 maturity with two extension options taking it to May 2023. This date, as you will observe, is in the past, but servicer Mount Street added another three months, saying in May “The Servicer has been discussing an exit strategy with the Borrower for some time and has been made aware that the Borrower is in the course of refinancing the property. The refinance will not be complete before the Final Loan Repayment Date, so the Borrower requested an extension to the date for repayment of the Loans to 22 August 2023. The Servicer has confirmed with the incoming lender that the refinance will complete ahead of the proposed new maturity date.”

The extra extension came with some prepayments and other waivers, but there’s been no further announcement, and August 22 is coming up fast. Part of the reason for waving through the extension request was the eminently refinanceable nature of the deal; at an LTV in the mid 20s, this is handily covered even at the worst possible assumptions.

Whoever is being lined up for the refi has also, presumably, thought long and hard about their WeWork exposure. It’s been rated Triple C for a while and the equity story has been much-remarked; if credit committee is happy with one of the anchor tenants being in this condition then a little going concern warning doesn’t change the picture much.

If WeWork is basically a building manager for an office full of happy tenants, then it shouldn’t matter too much if someone else steps in and leases / manages the building. If, however, WeWork is taking the risk of how many people show up to co-work and whether that justifies the overall lease cost, any successor might strike a different balance. 

Lots of UK restructurings in the past decade (mostly in the retail sector) have ended up burning the landlords, basically because the CVA process made it easier to deal with leases than grapple with bank debt or bonds. But the UK restructuring regime has changed since then, with the advent of the new UK Restructuring Plan (my colleague Chris Haffenden is the expert on this stuff, here’s a primer from him on the Plan), while WeWork might be more reluctant than most to attack the leases, given that access to prime real estate is very much the core of the business. Any process is going to be US-led; it remains to be seen how it would break out by jurisdiction.

Anyway, one nugget of positive news from servicer Mount Street’s latest report — Karaoke chain Lucky Voice has committed for a 15 year lease in Devonshire Square. We will do our best to patronise this establishment and ensure refinanceability for many years to come.

Cerberus and the weird interest rates

Before I went away, a curious notice caught my eye — an announcement from cash manager US Bank / Elavon that it had miscalculated coupon payments on Towd Point Mortgage Funding 2021 - Hastings 1, a Cerberus-sponsored equity release RMBS. It’s a weird and interesting deal; equity release collateral is hardly ever seen in the market, and it’s structured with a fixed rate, matching-adjustment tranche at the senior level. 

Screwing up an interest payment isn’t a great look at the best of times, but in this case I have some sympathy — instead of using a simple standard Sonia figure, there’s a cap and floor baked into the documentation, limiting Sonia to 3%, a level which it has exceeded over the last two interest payment dates.

It’s not an interest rate cap provided by an investment bank, transforming one set of cashflows into another for a price — it literally just tells the paying agent to assume Sonia is 3% if it prints above that level.

It’s not a secret. Clearly it’s in the docs for rating agency purposes, to take off some interest rate stress and make it easier to carve out a bigger MA tranche — but it is sotto voce. Open the first page of the prospectus and the reference rate described as “Compounded Daily SONIA”, with a footnote saying “subject to certain floors and caps”. These are disclosed in full on page 241.

It’s fairly common to have some kind of cap structure on interest payments, especially with low yielding legacy collateral, it’s common to have a derivative contract which caps a reference rate; some of these have been restructured over the past year, and floors are ubiquitous in certain asset classes. But I don’t recall many deals with a contractual cap of this sort — do write in if you have any data on the subject.

I think it’s unlikely this snuck past many investors at the time. Equity release mortgages are not an asset class for occasional securitisation tourists to play in; the mezz tranches here were not huge and were probably placed tightly.

The deal came to market when I was between jobs, so I can’t recall the execution, but to what extent did Cerberus have to pay for this? It’s clearly a cap with value now, with Sonia in the 5% postcode, but the forward rates curve was pretty flat in 2021, and so these were probably a long way out of the money.

Nonetheless, even if there isn’t a placement premium for the value of the cap, perhaps there should be a complexity premium. Markets work best when Sonia=Sonia; liquidity is best served by other investors and trading desks having a clear comparison across transactions.

This complexity and pricing question is relevant for other issuers. The basic question of how to finance low-yielding assets at current floating rate spreads remains highly relevant; “just pretend that Sonia is lower that it actually is” is unlikely to fit many fact patterns, but could it be useful for some deals?