Servicing-as-a-service

Much of the modern world is moving towards XYZ-as-a-service — you can get your music, films, chocolate, coffee, laundry liquid or even primary residence on a subscription basis. It’s an open question whether this is because consumers really want the world to go this way, or because capital providers like it — startups might not need to display much EBITDA, but their VCs do like a nicely growing ARR.

Anyway, you know which market really really likes streams of predictable leveragable cashflow? Securitisation! The more 'as-a-service' subscriptions there are, the greater the opportunities for securitisation to diversify into increasingly esoteric streams of cashflow. Securitise the world! Dream big!

There’s probably not much call for securitisation to fund the working capital of like, chocolate subscription companies, but when you get to pricier durables, that’s when you need cheap capital.

Hence the excitement and growth in the car subscription business. We talked about Finn’s big deal back in 2021, Octopus’s EV salary sacrifice deal last year, and, of course, the unfortunate events at Onto, whose specialisation in electric vehicles and short term contracts left it calling in the administrators.

Switzerland’s Carvolution announced a chunky CHF200m facility this week with Barclays and Waterfall Asset Management, providing senior and mezz respectively. Waterfall was also involved in the Finn deal, but happily the jurisdictions don’t overlap.

Car subscriptions, fundamentally, are little baby leases, giving consumers more flexibility than a traditional long term lease, and potentially coming with more bells and whistles (maintenance, insurance, tolls covered). So it’s not a huge stretch to build a securitisation facility around these contracts. But thin-slicing the contract length does increase the great challenge of leasing contracts — calculating the residual value of the car at the end of the term — and the sponsors themselves are more likely to be thinly-capitalised startups like Onto rather than the big fleet lessors or OEMs.

Carvolution, though, is profitable, and the biggest car subscription business in Switzerland. Scale matters in this business — buying cars at a discount takes off some of the residual value risk day one — and scale also gives stability. It also offers subscriptions across different car types and different fuel types, removing some of the risk of an Onto. EV residual values are less certain than those for ICE cars, as the technology is developing fast, and the, uh, “Technoking” of the largest manufacturer is known for a certain whimsical caprice when it comes to corporate decision-making.

Also announced this week was a very substantial deal for Optio Investment Partners, which has a fundamentally different business model — effectively a white-labelled capital solution for big OEMs looking to establish, or rollout car subscriptions. The manufacturer handles all customer-facing aspects, including maintenance, with Optio committing to buy the cars, handle the tech, provide the funding and act as a master servicer.

This is channelled through a Luxembourg fund structure, which should give the flexibility for expansion across Europe. 

The initial commitment is $750m-equivalent (the UK funding is in sterling) to cover Optio’s “Auto Evolution” leasing and subscription service, which is rolling out across the UK, though the facility also covers other jurisdictions. Optio is definitely dreaming big; it’s a “scalable multi-billion USD solution” that’s in mind. It has a partnership in place with Volvo Cars, and is looking to sign with other major manufacturers. 

Scale is important when you’re pitching Europe’s biggest car makers — this OEM-driven strategy requires being able to speak for size.

Also noteworthy are the counterparties. Kennedy Lewis Investment Management, which is providing the equity, is a big US credit shop, most regularly featuring in 9fin’s pages through its Generate Advisors CLO platform and CLO equity partnership with TCW. But it’s been a little more under the radar in Europe as an asset-backed sponsor — this was the first European deal one banker had heard of.

KKR is a much more familiar beast in European asset-backed markets, and markets itself partly on its ability to be a strategic partner to corporates, plugging in its asset-based funds for years in major size. One excellent example is Project Danube, the big PayPal transaction covered here; KKR, plus its mezz and senior lenders, could plug in capital to the PayPal platform to fund its BNPL origination efforts for years to come.

Optio’s proposition, unlike the other car subscription platforms, has a solution to handle RV risk — we’re not sure what it is, but it derisks the whole effort, allowing it to pass on credit risk, not asset risk, to its investment partners.

KKR and Kennedy Lewis, one might suppose, has a senior lender somewhere in the mix — Irish filings suggest that this is BNP Paribas.

Optio might have hedged out its RV risk, but another aaS facility announced this week is Citi’s deal with Everphone. This is a €250m combined securitisation and inventory financing, featuring Citi and KfW as senior investors, and Israel’s Phoenix Insurance Company in the mezz.

Residual value risk on mobile phones is more complicated than for cars — cars have well-established RV curves, pricing services which specialise in providing this granular data. Old phones, though, aren’t worth much, and can easily be damaged. 

In a sense, the traditional models of device funding through networks and phone bills already fulfil some of this 'as a service' role; you already subscribe to a phone network, which also bills you until you’ve paid for the device.

This can also be securitised — like so much, it’s a big market in the US and a private market in Europe — while some companies, such as Virgin, use receivables to back corporate bonds.

Eastern Promises

As issuers know very well, there aren’t a ton of players at the spicy end of the capital structure in public European ABS. It’s not a large market, and the biggest funds don’t want to spend all of their time fighting over £5m mezz tranches; they’d prefer to launch “asset-backed private credit” strategies instead.

For a long time, though, East Lodge Capital Partners was very much in the club. Started by Ali Lumesden, the former head of ABS at CQS, the fund ran from 2014 to its eventual winddown in 2023. There wasn’t a big fanfare or a flameout; it just fairly quietly shut its doors and sold its assets.

PM Richard Skeet joined crypto fund Hivemind Capital (founded by Citi structured products head Matt Zhang) as head of research in 2022. CMBS PM Rob Riley switched to the sellside last year, joining Standard Chartered’s strategic investor group sales effort.

CTO Vishal Shah joined Alcentra in a similar role, head of risk management Nirvan Sunderam went to Pemberton Asset Management as head of risk.

Lumsden himself has a new role as chief investment officer for debt strategies at Temporis Capital, an impact investment shop focusing on renewables and cleantech. One of the East Lodge’s Lux entities has called in the liquidators; the main London LLP said in its recent accounts that “as the assets under management of the firm declined further in 2022-23 the decision was made to liquidate the remaining assets in an orderly manner with a view to closing the firm”.

LPs are also packing up and going home — San Francisco Employees Retirement Scheme, which had $200m in the main fund and another $25m in B shares, terminated its investment in July last year, and got its money back at the end of the year. LPs including the UK’s Local Government Pension Scheme, Texas Municipal Retirement Schemeand Massachusetts State Pension fund also invested with East Lodge, though we couldn’t dig up the status of these investments.

The firm had a rough Covid, but who didn’t? Bloomberg ran a story on April 21 2020 noting that the fund was down 26% in March, but I mean, so what? Senior secured TLBs from basically good companies were down 20%+ in March 2020. Everything was down! If you’re a fund buying deep mezz and resid notes with a bit of leverage to boot, 26% really doesn’t seem terrible.

Nonetheless, it wasn’t great for the fund. The previous year’s accounts said “a reduction in company income is attributable to fund performance in 2020 as a result of the Covid pandemic leading to much of the capital under management remaining below its respective high water market. This negatively impacted performance fee income…. capital under management has also declined over the last two years with a corresponding impact on management fee income. The firm has taken steps to reduce its cost base commensurate with its lower AUM and is working on several new mandates that are expected to increase AUM and income towards the end of 2023”.

Presumably most of the assets have also found new homes. Just before IMN/Afme’s Global ABS in Barcelona last year there was a bid list which was widely reckoned to be East Lodge — deep mezz from pre-crisis, non-conforming deals which essentially never trade, a treasure house of rarely-seen mortgage bonds.

Curiously, though, at around the same time, East Lodge appeared to be still signing new investments. Here’s a forward flow with Blue Motor Finance dated to 12 June 2023. Another private investment was a mezz piece with auto lender 247 Finance. A process last year sought to replace this mezz, but morphed into a full sale of the portfolio (we think M&G won it).

Anyway, this is all quite disappointing. We want more adventurous hedge funds, whatever they call themselves. Pour one out for East Lodge!

When the loans are too good

I haven’t been writing much about CLOs lately, partly because I have a crew of excellent CLO-focused colleagues running a fully fledged CLO news operation (drop me a line if you’re interested in what we’re doing behind the paywall, there’s some excellent stuff on the CLO pipeline, fundraising activity and more).

But it’s striking that, over the last couple of years, CLOs have been consistently lagging behind the loan market that brings them their material. Loans widen and tighten faster than CLOs; when there’s a selloff, there’s a miniscule opportunity for a print-and-sprint, but when markets are rallying, loans have tended to come in faster than CLO liabilities, squeezing CLO arbitrage.

In 2024’s screaming rally, we’re seeing the same sort of thing. Loans are readily prepayable and repriceable, and the idea is that CLOs past their non-calls can match this spread movement by refinancing or resetting their liabilities tighter.

CLO triple-As have indeed screamed tighter, with some of the more than nine deals marketing at the moment showing seniors at 150bps. But the only reset flow so far has been from the dislocated 2022 vintage. 2021-era deals are sitting pretty with low capital costs, but with triple-A prints in the 90-110bps range the reset options are still a long way out of the money.

Meanwhile, loan borrowers are piling in to cut their own borrowing costs — repricings from Tricor-Vistra, Sebia, Rubix, Idemia, TMF, IQ-EQ, TOI TOI & Dixi, Refresco, Restaurant Brands Iberia, Group of Butchers have all been in market lately. In some cases straight refis have seen such strong demand that this has spilled into repricing — theme park operator Merlin placed a €200m refi tranche but came straight back for a €1bn drive-by repricing of its 2029s.

Some of this is pure technicals — nine or more new CLOs gathering assets, tiny scraps of new money — but it also reflects general optimism in credit and the possible turn in the rates cycle. 

For CLO managers, though, they’re being punished, in part, for picking the good credits. Some of these deals came with handsome OIDs, so they’ve seen some capital appreciation. In a sense, that’s always the game; managers with low-risk, high quality portfolios will tend to come in at the bottom end of equity distribution. Yieldier assets make better arb.

The other interesting dynamic is the extent to which repricings and refinancings are now taking out deals out of reinvestment period. Rather than last year’s A&E dominated dealflow, which, combined with flexible CLO docs, allowed post-RP deals to be dragged along into new facilities, the latest crop of repricing and par refis should light a fire under triple-A amortisation…

…which will, in turn, give triple-A investors more cash to invest, tighten triple-A spreads, and then, maybe, bring more resets back into the money, restoring cosmic balance. For now though, the loan market is running ahead.

How much do NPLs cost?

Traditionally HY-financed debt purchasing firms are remixing their capital structures, as we’ve discussed around here. AFE’s equity-free HY-financed portfolio was probably the apex of the LevFin-does-NPLs trend, a deal that’s a punchline in certain quarters (and now safely tucked up inside Arrow Global’s third party fund).

Intrum Justitia is the biggest debt purchaser in Europe, and no stranger to a little securitisation — a co-investment with CarVal in Intesa’s Project Savoy, €10.8bn GBV, is probably the biggest, and also something of a debacle.

But now, belatedly, the firm is following Arrow and Lowell in trying to sweat its capital a little more, pivoting towards capital-lite servicing and bringing in third party money. This, however, can come in different forms. Since its capital markets day in September, Intrum has been marketing part of its back book, and on Tuesday announced that Cerberus was the winning bidder for a larger-than-expected portfolio sale of seasoned unsecured NPLs — with the disposal raising enough cash to cover Intrum for its 2024 and 2025 maturities and buy time to pivot the business.

So given this is a publicly traded business with extensive disclosures, what did Cerberus pay for the loans? Should be a nice simple mark for the current value of NPL portfolios, right?

Actually, this is an incredibly painful and difficult exercise. 

Intrum’s headline figure was a sale at 98% of book value. Book value was SEK11.5bn (€1.01bn), from SEK382bn nominal, and Intrum took a writedown of SEK219m (1.9%) related to the discount.

This, however, is the only figure that almost nobody in the market accepts as the “real” discount. Cerberus is not noted for its generosity to highly motivated sellers like Intrum; just knowing the two counterparties, who would you think got the better end of the trade?

Book value for the debt purchasers is also based on IRR at the time of portfolio purchase, and there’s a very limited requirement to remark these to market levels. These portfolios are 2017-2018 vintage, and the environment is very different today, with IRRs up in the high teens or above — base rates have gone up, debt purchaser capital costs have gone up, and so have required returns.

Particularly noteworthy is the date of economic transfer, which has been backdated to 30 September 2023, with Intrum recording a further SEK626m in foregone investing and servicing profits between the economic transfer date and the expected closing of the deal.

So you have a sale of SEK11.5bn in assets, with SEK845m of total writedowns associated with it. Perhaps this suggests a seven point discount to book?

The complexity starts to multiply when you consider that Intrum will sell the loans into a vehicle, funded by non-recourse leverage (an NPL securitisation basically) from Goldman Sachs, in which it will retain a 35% stake. 

Intrum also says it will receive SEK8.2bn of cash proceeds from the sale — a reasonably trustworthy number, as this will be applied to reducing debt. Can’t pay down your RCF with non-cash adjustments!

Selling 65% of SEK11.5bn at 98% should get you proceeds of SEK7.33bn, so there’s clearly something going on here. But what?

If we assume that the cash proceeds number includes the servicing cashflows until closing, it makes a little more sense. Let’s say the SEK626m non-cash adjustment relates only to Cerberus portion of the vehicle (65%), and consider that Intrum is receiving, in cash, the full 100% up to closing (SEK963m). Add that back to the SEK7.33bn and you get to SEK8.29bn….which still doesn’t quite tie to the SEK8.22bn, but at least it’s in the right ballpark.

But that leads to an even lower effective purchase price — SEK626m on the SEK7.45bn 65% is 8.3 points, plus the two points headline book value discount, for a purchase price sub-90.

If you thought this was bad, consider that the servicing fees are also off-market, by some meaningful amount. Intrum would disclose only that they're below its 25% target, but not how far below. These fees will be in line with existing arrangements… but the existing arrangements are an internal transfer price between Intrum entities, not an arms' length market level. 

The debt purchasing segment is a controversial one, and one which attracts motivated financial engineering. For Intrum, it’s hard to see what other options were available — it needs time and space, and presumably ran a competitive marketing process to achieve the best price. It took more size than it originally intended to (which actually relevers the company a little, since the EBITDA decline from the sale hits before any collections flow back from the securitisation) but that’s probably the right move if market access is not assured.

It would, however, be a little easier to give Intrum credit for this strategic move if the numbers weren’t such a mess — there’s no shame in selling assets at a discount; everyone knows the market backdrop has changed.