Excess Spread is our weekly newsletter, covering trends, deals and more in structured credit and ABS — subscribe to this newsletter here.

‘Deed of Indemnity’ is an indemnity, rules judge

The long-running High Court case pitting NatWest Markets against Dutch servicer CMIS wrapped up in a hearing last July, but the judgment was published this Wednesday (15 January), and represents a clear victory for NatWest for the claimed €155m in payments — a sum which is more than sufficient to bankrupt CMIS.

The case concerns old Dutch and German RMBS from the E-MAC series, which had an unusual swap structure — instead of a balance-guaranteed swap, any mismatches between the deal notional and swap balance were trued up with “Notional Adjustment Payments”. Given that these bonds are getting on for 20 years old, and haven’t had an active sponsor since Fortress bought G-MAC Europe in 2010, a fair bit of tracking error has built up, with these payments piling up in the “Subordinated Swap Amounts”.

After the put date, when the deals were supposed to be redeemed, there’s another deal mechanic allowing the hedge counterparty to redo the swap contract on market terms (any additional payments also in the “Subordinated Swap Amounts”).

These, as the name implies, are junior in the waterfall, and there hasn’t been a huge amount flowing through the deals to cover them, so the question in the case was essentially whether CMIS, the servicer and original sponsor entity, should cover them or not.

The judgment underlines the difficulties of picking over a deal this old.

Witnesses for NatWest included Sachin Arvind Zodgekar, managing counsel for the structured derivatives business, and Jean-Paul Westgate in the structured derivatives team, while Sean Joseph Daly, chairman and managing director, gave evidence for CMIS.

But as Nigel Cooper KC, sitting as High Court judge, notes: “None of them were employed by the party on which behalf they gave evidence at the time when the relevant agreements were made. While their evidence was helpful in explaining the structure and background of the transactions which are at the heart of the present dispute, none of the witnesses could give first-hand evidence as to the reasons for particular features of the transaction structure, such as, for example, the purpose of the Deeds.”

Simplifying a lot of heavy duty legalese, the essence of the CMIS argument was that, if it were on the hook for these payments, the structures wouldn’t really be delinked from the sponsor, and that’s not really what securitisation is all about.

From the judgment:

“CMIS submits that the Deeds are curious documents, which were executed behind the scenes between NWM and CMIS. They contend that it is unusual for an originator of a securitisation to provide a guarantee or indemnity in favour of the swap counterparty because the whole point of a securitisation is to remove assets and liabilities from the originator's balance sheet. They submit that if the Deeds are construed as entitling NWM to the indemnity they seek, this would run contrary to the purpose of a securitisation by restoring a liability for the EMAC Indemnifiable Amounts to CMIS balance sheets.

CMIS also points to the fact that the drafting of the Deeds is bespoke and is not a standard form document promulgated by the Loan Markets Association or similar organisation or in the standard form of guarantee as required by a bank. They highlight that the Deed is not a one-sided document but confers benefits on both parties. They also suggest that the origins of the Deeds are shrouded in obscurity to a far greater extent than a normal bank guarantee or indemnity.”

It’s a fair point, but on the other hand, the parties are literally arguing about a document titled “Deed of Indemnity”, which Cooper finds persuasive.

“If the intention of those parties and their advisers was to draft a deed of indemnity which was in reality a contract of guarantee or to ensure that CMIS were to have no greater liability in respect of the EMAC Indemnifiable Amounts than the EMAC Issuers, then I consider that the contract would have been drafted in very different terms and would not have been titled ‘Deed of Indemnity’”.

While the judgment is fairly clear, there’s still a pretty hefty “what now” question hanging over the shelf. These are not the most liquid transactions out there (we covered a BWIC in the shelf last July) but there are still some active holders (not legacy accounts or run-off vehicles), and occasional activity.

Further court activity is probably required to clear up final payments and possible appeals — Cooper says: “I would further ask the parties to liaise as to an appropriate final order dealing with any other matters arising from this judgment. To the extent that the terms of such an order cannot be agreed, then I will deal with any outstanding matters at a hearing to deal with consequential matters including costs and any application for permission to appeal, if any”.

…and then there’s the question of CMIS. Fortress has already more or less walked away (the HoldCo vehicle for CMIS has already written it down to zero) and is unlikely to fund a donation to the NatWest shareholders.

The deals, as you’d expect, have language contemplating the replacement of a servicer on servicer insolvency, but they also confer hefty obligations on the servicer. Not only are they supposed to fund the Swap Subordinated Amounts, but they’re supposed to actually redeem the deals through a servicer advance allowing the investor put options to be exercised. A replacement can probably be found to collect borrower payments and manage delinquencies, but any such entity may not want to assume all of CMIS’s roles.

On the other hand, this is a potential opportunity to do something decisive with the shelf. Legal uncertainty isn’t quite banished, but it’s certainly reduced, and a period of noteholder engagement will surely have to begin to address risks from the probable disappearance of CMIS. Perhaps this a chance for a long-overdue refinancing?

Superpower

We wrote a fair bit about The First Data Centre deal in Europe last year — a Barclays-arranged transaction for Vantage Data Centers, backed by two assets near Cardiff. The deal was a sort of ABS-CMBS hybrid, structured as a fixed rate note and marketed with 144A docs (US investors took 36%). A fixed rate asset-backed structure is squarely in the wheelhouse for UK real money. Even those funds which don’t have dedicated ABS functions probably have exposures to asset-heavy property-secured structures; think Broadgate Financing, Westfield Stratford City, Center Parcs…..or, I guess, Thames Water!

2025 will hopefully bring the various data centre transactions that have yet to surface into the market, but there is so much more to do! To sync up with UK prime minister Keir Starmer’s announcement on Monday (13 January) that the UK wanted to be an “AI Superpower”, Vantage said it would invest £12bn in UK data centre infrastructure.

Last year, it bought a huge site in Wales formerly occupied by a Ford factory (also near Cardiff), and is consulting on constructing 10 buildings and three substations, over a 10-15 year period. That doesn’t spell immediate data centre ABS deals (the very nascent market is probably some way from assuming construction risk on a project of this kind), but it does mean there’s everything to play for, and every incentive to ensure there’s a capital markets route to finance this kind of investment, rather than relying on the bank market.

The government release also highlighted CoreWeave’s commitment to the UK — the firm commitment £1.75bn in UK investment last year.

This financing could end up looking a little different though.

CoreWeave raised a truly enormous private asset-backed facility ($7.5bn) led by Blackstone and Magnetar, the latter of which was the first institutional investor to back the firm. Other investors included Coatue, Carlyle, CDPQ, DigitalBridge Credit, BlackRock, Eldridge Industries, and Great Elm Capital Corporation.

This was a GPU (chip)-based financing, backed by the compute capacity directly, rather than the buildings, land and operational structure that goes around a data center. Jim Prusko, partner and PM at Magnetar, runs through the structure here.

Prusko says: “Depending what stage you get involved, you have the breadth of all those different risks. If you’re investing in high performance compute, but it’s a greenfield data center, you have to think about delivering of the power, timing on all the components, but if you’re making a GPU-based loan, that loan is based on a running GPU in an existing high performance compute data center. You don’t have to think about some of the early stage issues, you have to think about how long is my contract, how good is my contract, what do I think the value of renting out the chip will be at the end of the contract, how much rent could I get if I had to re-rent in the middle of the contract.”

This is a somewhat different model though — CoreWeave (and AI generally) requires more power, and different infrastructure compared with a more straightforward data centre rented for ordinary cloud usage. The clients are different as well; although most large tech firms have substantial AI operations, that’s different from simply running the Microsoft Azure cloud or Amazon Web Services.

Still, whether we get GPU-financings, more real-estate-like deals, secured bank-led platforms or a bit of everything, there’s a ton of deals coming. Line is going up and to the right!

Fashionably late

After a pause last week to draw breath, the market is finally open and feeling fine! Santander UK wrapped up a rapid fire two day Holmes syndication, with a nice upsize to £750m, coverage of 1.3x at final size, and 53bps spread (from mid-50s). That looks pretty textbook, as syndications go; not the kind of bunfight that can result from starting cheap, nice coverage, taking a decent deal off the table, 35 investors allocated. Beaut.

A particularly encouraging sign is the strong showing from Europe — 18% continental Europe on the distribution stats.

Holmes 2024-1 went 99% to the UK, Lanark 2024-1 was 93% UK, Permanent 2024-1 was 96.5%, to take the three trad master trusts out in January last year, while Yorkshire Building Society’s White Rose Master Issuer 2024-1 was 100% UK.

This year’s outing also saw more asset managers allocated — 48% vs 47% banks. This compares to 29% asset managers and 66% banks the previous year, 40:60 in Lanark, 36.5:63 in Perma. The “type” column is rounded out with 5% official institutions… possibly EBRD or the Central Bank of Latvia.

Signs are encouraging elsewhere, too, with VW taking a €1.192bn tranche out of the market for German lease deal VCL 44 on a €2bn book, just the right side of 50bps. Delpinius 2025-1 with a 1.6x covered class A and Koromo Italy II also at 1.6x also look notably healthy; this is not euphoria, it’s a balanced market ready and willing to do business.

It’s just a coincidence that Dutch prime RMBS, UK prime RMBS and German autos, the big three vanilla asset classes are all pricing with a credit spread of 50bps or thereabouts (for a five-year, three-year and 1.2-year WAL respectively) but it’s relatively unusual as a historical phenomenon, and raises the question of which asset classes have most room to tighten and why, and what’s going on with the term premium.

I haven’t crunched the numbers, but for most of my career, I’d expect VW tightest by some way, followed by Dutch prime and UK a bit further back. Sonia and Euribor are not strictly comparable as benchmarks, so it isn’t apples to apples, and this also coincides with a period where the ECB had been fairly active.

Still, even if you assume one euro triple-A STS bond is much like another, a term premium of 0-1bps for going from 1.2 year VW to 4.9 ASR seems excessively small.

There’s various noise around the German economy and VW in particular, but not to an extent that ought to affect VW ABS triple-A much. VW clearly went for size in this deal, but in a clear mid-Jan market where size was available; with a €2bn book it wasn’t stretching.

So if I was going to make a spread prediction, I’d say captive autos have most room to compress as we go into the year ahead! Don’t trust a journalist, but it’s now game on.

No flow

Securitisation issuers don’t get much more established that NewDay, which is banked by more or less every active securitisation bank in Europe (counterparties in the double digits), and a regular visitor to the market with its NewDay Funding Master Issuer (£2.46bn) shelf (three outings last year, now an STS shelf) and, somewhat less frequently, the smaller NewDay Partnership Funding (£646m) programme which funds its store cards.

These facilities, and masses of bank capacity, generally serve the firm well — the latest numbers show £1.6bn of headroom, which covers the trusts as well as various private warehouses and other facilities, such as the funding for the Argos acquisition — but that doesn’t mean it won’t mull other options.

NewDay has been looking for a long time at a credit card forward flow arrangement, with Alantra running the process.

That wouldn’t be a total departure for the firm, as it already has a personal loan forward flow in place for a few years (it said in its 2022 numbers that it wanted to “create forward flow revenues from personal loans”), and it has other flow arrangements for dealing with bad debt (including with current restructuring candidate Lowell).

But it is somewhat different. What better owner of credit card receivables is there than NewDay? What kind of institution would get better leverage terms for credit cards than NewDay? Who would be willing to buy credit card receivables at a high enough price that NewDay would want to sell?

The structure of credit cards also makes it a little trickier. Credit card arrangements have an indefinite lifespan, but a revolving balance.

Drawn debt of defined term (like personal loans) is a more straightforward proposition for a fund, and much of the structuring around credit card portfolios goes into managing these characteristics — that’s the reason NewDay, and most credit card ABS issuers, use master trusts (we look forward to welcoming Capital on Tap’s public master trust debut; the vehicle is up and running in private).

It’s perfectly possible to get over the hurdle, but at the margin it makes a flow deal harder to do — and so we understand the process is now on the back burner.

More challenging still is the process from the issuer side, since the structure of credit card contracts makes it difficult to achieve accounting derecognition. One could still do a forward flow, but keeping the loans on balance sheet while selling their entire economics is not a trade most institutions are keen to put on.

The process could also be on the back burner because the NewDay sale is supposed to be on (per Sky News last year).

Sky says this could mean an equity listing, but there must be strategic interest as well — I guess one could call NewDay the Kensington of credit cards, and just as with Barclays and Kensington, there’s the chance for a clearer to make a genuinely transformational move in UK retail finance. Lloyds already has a merchant point of sale JV with NewDay, in which Lloyds funds prime and NewDay near-prime customers, but a wholesale acquisition by a bank large enough to swallow it could make a lot of sense.

Directing Trafig

At certain kinds of healthy real economy companies, arranging working capital financing can be a sort of afterthought — if a firm has adequate debt available from banks or bonds, getting involved with factoring, inventory finance, invoice discounting, receivables securitisation or similar is just a nice-to-have, a piece of cap stack optimisation an ambitious treasurer might complete.

Commodity traders are the absolute opposite. Financing in general, and inventory or working capital financing in particular, is an essential part of their DNA, a mission-critical enabler of their activities, a function on which the business relies absolutely. Stripping back everything, the basic business is moving high value cargoes from place to place, and buying and selling commodities at different times, locking prices on one side and taking price vol on the other. This naturally creates massive working capital needs!

So it is that Trafigura, one of the largest of the bunch, has 150 banks providing credit lines, with capacity of $77bn (larger than the $76bn in total assets). It has 38 banks in its Asian loans alone. According to its annual report, Traf has access to $14.6bn of same day liquidity available.

Anyway, it said on Monday that it had signed a first-of-a-kind facility, a $1bn inaugural uncommitted discounted facility of credit-insured receivables and prepayments, allowing off-balance sheet treatment, and structured to optimise banks’ capital weighting under the Capital Requirements Regulation.

Per a Trafigura spokesperson: “CRR regulation is the basis of the structure of the facility”. One assumes this would allow the banks to give better terms, though Trafigura is giving nothing away, with a spokesperson saying: “There is no other similar structure in terms of size and covered scope. The Facility was competitively priced.”

I don’t know enough about commodity trader receivables financing to know how new or interesting this actually is (do write in if you have more colour) but off-balance sheet credit optimised facilities that thread regulatory needles are very much the kind of thing this column’s readers do for a living, and if it works for Traf, surely there’s more to come?

Enjoyed this weekly wrap? Our customers receive this content ahead of the crowd — find out more about 9fin’s news and analysis.