The “draw a line” approach is basically bad news for CLOs, and by extension, for the leveraged finance markets in Europe. Since the bank failure volatility calmed down, credit spreads in CLO tranches have gone nowhere particularly interesting. A touch tighter some weeks, a touch wider in others, a little more manager dispersion one day and a little less the following week. Big BWIC flows or heavy primary weigh on markets, but technical tightening comes in when supply eases off.

If this stasis continues, more and more CLOs will drop out of reinvestment and some that aren’t will start butting up against WAL Test constraints, restricting the ability of borrowers to refinance or A&E their facilities. New issue loans might have to review their standard seven-year term structure. Liquidity will tighten further in the loan market, with less of the sector AUM able to be actively managed. 

This might seem a little alarmist, but the walls are closing in quicker than you might think. There are very few loan maturities this year in widely held CLO collateral (Keter is the biggie, and sponsors BC Partners and PSP are working on it), and not that many next year (French supermarket Casino is the largest, but the dynamics there are quite unique; it’s just entered the French conciliation process).

The maturity spike starts to look chunky in 2025, though, which means sponsors will want to address these in 2024 before they go current. And if CLO triple-A spreads are still bopping along in 180-200 bps context that’s going to be much harder, because that kind of level means no resets will have been possible. 

S&P Ratings had some good stuff on this in a webinar last week — European CLOs: Where do we go from here? — available at this link. There’s a chart below, but the rest is worth a look too, with some heavy duty structural stuff on interest smoothing, workout language, and how maturity extensions can be handled under the documentation.

If spreads do continue to go nowhere, then perhaps it’d be possible to split the difference with a full reset and try for some kind of bespoke A&E? Post-reinvestment deals tend to perform worse, all else being equal, which mainly hits the bottom of the stack. 

The sweetener (for the top of the stack) is supposed to be amortisation, but this is probably going to be very drawn out, as obligors themselves try to roll funding through A&Es, and leave stubs outstanding where CLOs can’t agree to extend.

So is there a deal to be done pushing out reinvestment periods and WAL tests for a margin bump that’s more modest than a full reset to market spreads, plus some cash up front for the consent? In general, triple A investors would rather just have their money back so they can go and buy secondary or new issue 100 bps wider, but some might have their own reasons for staying invested and taking a fee and a bump instead.

Though reinvestment periods get all the attention, WAL tests do the heavy lifting in terms of constraining manager behaviour (we went quite deep on this topic in January).

Most deals have the ability to extend WAL tests by some amount (12, 15 or 18 months), with consent of at least the controlling class (sometimes others too). That doesn’t change anything else about the deal, and it’s a one-time only thing, but we’re likely to see this more — so far it’s only been Tikehau IV this year.

This is a more plausible option for the better-performing transactions out there. For all that investors are supposed to be ruthlessly economic and self-interested, pushing out maturities on a transaction that’s in trouble could create a lot of bad blood with the junior noteholders, who may want out and want equity encouraged to get on and call.

Another possibility is turning old deals into new. Most CLO calls remain uneconomic, as we discussed last week, but managers may be able to manage arms-length novation of selected assets into new warehouses. A truly arms-length price shouldn’t come out very different to sticking the whole portfolio out on a BWIC, but there’s still an advantage to be had there getting first dibs on the right collateral in the right size.

But maybe the line doesn’t keep on going, and something actually does turn up — CLO credit spreads normalise, whatever normal might mean (triple A sub-130 maybe?) and resets become more possible.

What’s the catalyst? Ideally for managers and equity, Norinchukin would splash back in and start buying bonds miles through the market, igniting a rush tighter and a scramble for paper, but that’s more a triumph of hope than experience. When the Japanese whale has broken the surface in the past year, it’s been in a limited and selective fashion, and it pulled out of a CVC trade when the LDI stuff broke.

More prosaic relative value changes could potentially help — CLOs don’t exist in a vacuum, and a big rally in IG corporate bonds could drag spreads along. European CLO triple-A is cheap to US, after hedge costs, according to Barclays…but CLO spreads are slightly rich vs IG, so make of that what you will. It doesn’t suggests a powerful rally any time soon.

Italian GLAMour

We understand there’s a new kind of NPL securitisation in town, in the shape of a scheme called GLAM. This is a scheme being rolled out by AMCO, the state-backed “asset management company”, and is intended to eventually reference around €11bn of loans. This is a chunk of change, but only represents 4% of Italian guaranteed SME lending.

The actual structure is pretty neat — it’s basically a way for banks to shed their government-guaranteed Covid-era loans and convert them to securitised notes. I’ve not managed to dig up much on the performance of these loans, but I’m gonna go ahead and suggest it’s not great. 

The whole point of the various government-backed loan schemes during the Covid period was to shovel money out of the door to SMEs facing total or partial closure of their businesses, staving off mass failures and preserving the economic capacity of countries. It was no time for sober-minded assessment of credit quality. In Starling Bank’s results (of which more later), some 65% of borrowers under a UK scheme are up to date with payments, this being “in line with expectations”.

Banks were willing to cooperate because of the government guarantees, covering 80%-100% of the loan notional depending on the scheme, but it’s not ideal for them to have to claim on these guarantees, and many of the loans would benefit from some sort of workout — maybe a can kick, maybe new money, maybe a payment reprofiling. Could be nasty, could be nice but it’s probably better to have these decisions in the hands of a specialist debt manager. Also, if possible, governments don’t want to pay the guarantees.

There’s a deadline on its way — the EU’s “State Aid Covid Temporary Framework”, which included a carve-out to allow member states to provide solvency support, fully expires at the end of the year, so SMEs in Italy will need different kinds of support.

State aid rules have shaped the structuring. The European Commission had to comment on the scheme last year, with boss lady Margrethe Vestager considering it “an important step towards recovery for the Italian economy, while ensuring competition is not distorted”.

The state aid rules mean that markets have to be involved, with private investors helping to set prices. The price that AMCO pays for the loans will be based on a private sector auction, while the notes transferred back to the banks will either have a price agreed among all banks “ensuring no beneficial pricing for any portfolio of loans”, with a third party evaluator checking the figures.

But the notes can also be partly sold to investors at a market price, removing the need for this mechanism, and we understand that marketing efforts may be beginning via UBS, longtime advisor to AMCO on the scheme.

Other state-aid friendly features include restrictions on the new financing that AMCO can offer (must be “alongside financing from private operators under the same terms and conditions”. It may also “provide short-term liquidity assistance to the platform to cover mismatches between the inflows from the loans and the required payments for the notes”. These loans will be “remunerated with an interest rate that is in line with market benchmarks”.

The state aid stuff is a bit of a fudge, as it always is — if there was no market distortion, what would be the point? The GACS scheme had the same kind of fudge baked in; the senior tranche guarantee was priced off a basket of Italian CDS spreads to inject a whiff of “market” pricing into a vast state subsidy.

It’s a multi-level fudge in this case — state-guaranteed loans being shifted to a state-owned loan servicing group in return for state-backed notes, so the state-controlled banking regulators can upgrade their assessments of bank capital and liquidity. All done on “market” terms!

That sounds like I’m against the whole thing, which is absolutely not the case. Most interesting financial structuring is trying to solve for several sets of constraints, optimising price, tax, capital treatment, accounting and so forth. Throw in EU state aid and yet more elegance is required.

This seems like a good solution to a problem which lots of other European countries will surely experience — most geographies had some kind of SME support / loan guarantee scheme, and in all cases, there’s going to be pressure from finance ministries to minimise the pain of paying these out and preserve value in the SME sector where possible. 

Italy’s expertise in managing NPLs pre-Covid, and the extensive infrastructure at state and private level, probably gives it a big leg up when it comes to grappling with the Covid hangover, so hopefully these AMCO notes will sell!

Fit bird

We wrote about challenger banks buying LDI-loosened bonds back in March, highlighting the massive expansion of Metro Bank’s RMBS book — more than a billion in the last year. But information was harder to come by for Starling Bank, which is privately held and only reports annually. 

But this week we have the considerable joy of the Starling annual report to pore over. On first glance, it doesn’t look like there’s been much bond-buying action — RMBS holdings increased from a paltry £11m to a slightly less paltry £48m between March 2022 and 2023. Now, this doesn’t mean that Starling wasn’t active; buying a ton of bonds in October 2022 and selling them in February 2023 would have been a nice trade. You might expect to show up in the income statement if it happened, but it’s hard to disentangle from price movements on the much larger pile of Gilts that Starling holds for liquidity purposes.

More interesting than chasing shadows on bond buying is Starling’s portfolio buying activity — a further £984m of mortgages in the year, including £503m in April 2022 (that’s Masthaven) and £482m in buy-to-let in September 2022. This followed an acquisition the previous year from Kensington alongside Goldman Sachs (with the investment bank taking the sketchier end of the portfolio and securitising it in Parkmore Point 2022-1).

Big picture, Starling is reorientating from a portfolio heavily focused on the UK government-guaranteed Covid loans (BBLS and CBILS) towards a more conventional challenger bank model led by higher interest mortgage lending. Fleet Mortgages, bought in 2021 for a £50m, an earnings multiple which inspired the whole sector, is now up to £1.5bn in lending.

Staying with UK challenger banks, Monzo also reported this week, and frankly whoever’s covering the Monzo treasury team needs to step it up a bit — Monzo increased its treasury investments from £1.67bn to £2.73bn, but the bulk of this increase came from supranational debt, rounded out with commercial paper and term deposits. Very tedious stuff. Monzo reports that treasury interest income increased by a whopping 1125% in the year…..from £2.4m to £29.4m. On a portfolio of £2.7bn. Surely someone can introduce them to the delights of floating rate, yield-enhancing senior RMBS paper?

I also tried to figure out what Chase UK (JP Morgan’s UK retail deposit-taking arm) was up to. Full disclosure, I’m a customer (Jamie 2024), but it’s also one of the most significant challengers to emerge in the last couple of years, capitalised with a very healthy $2.1bn and with money poured into advertising, cashback offers and a shiny backend system from Antony Jenkins’ 10x Banking. It has been buying deposit share, and doing so very aggressively.

I’d originally assumed this would all form a deposit base for JPM’s investment banking, trading, underwriting activities in the UK, or at the very least for its push into principal finance for UK mortgages (e.g. the Pierpont BTLshelf). But I am dead wrong — this kind of activity, including the Credit Portfolio Group and parts of DCM, were transferred out of what’s now the Chase entity as it ramped up, leaving it a relatively pure retail bank. Unfortunately the most recent numbers are from 2021, so no sense of what plans to do with the vast inflows of deposits. Will it manage treasury and liquidity separate from the group? Will it buy some bonds?

Light of the world

To finish with a good news story, Citi announced a solar securitisation in Kenya for ‘Sun King’ this week. The $130m-equivalent deal is backed by Absa Bank Kenya, British International Investment, Citi, FMO, Norfund, Stanbic Kenya and the Southern African Trade and Development Bank — quite the coalition of great and good.

Genuinely, though, this is a cool and worthwhile deal, securitisation markets actually making the world a better place. Solar installation in sunny but poor places with many off-grid households is just a slam dunk bit of development and a great use of the power of finance — there were households without electricity that now have electricity, of the clean and green kind to boot.

Per the press release: These payments can be made via mobile money or cash for as little as $0.15 a day. Approximately half of Sun King's registered pay-as-you-go customers in Kenya are women, the majority of whom access formal financing products for the first time.

Under the securitisation structure, investors are financing the pooled expected future payments from over a million Sun King customers. The structure connects unbanked or underbanked customers to the finance they require to purchase solar assets and provides investors with access to a steady yet underserved market that offers risk-diversified returns.

This is great, but it does show the challenges of doing this kind of deal. $130m-equivalent is not a large transaction, but it still needed a million customers to get to this level. Kenya apparently has 12 million households (presumably solar installation is a household thing, rather than a population thing?), so Sun King has had to get to an absolutely stonking market share to make it viable.

It’s also local currency, which likely limits the audience for placing this kind of deal on a fully private sector basis. Connect up the dollar-denominated market to African off-grid solar and you have a truly awesome firehouse of capital, but we don’t seem to be quite there yet — cross-currency swaps in securitisations are difficult at the best of time, but I imagine there aren’t that many counterparties doing bespoke amortising swaps on Kenyan shilling currency pairs. Happy to be proven wrong though.

But still, the space seems to be growing. From the outside, the deal looks similar to Credit Agricole and Societe Generale’s solar securitisation deal in Cote D’Ivoire in 2020, which also had pay-as-you-go features, some support from impact/philanthropic capital, and aided the roll out of off-grid solar. This was an €18m-equivalent tiddler, but no doubt very meaningful to the borrowers in question, and hopefully we’ll see more of these grow up in other geographies.