The deal is material for PayPal — it’s cranking up share repurchases from $4bn to $5bn this year, for one thing — but it’s also one of the most dramatic illustrations of the sheer power and scale of the private asset-backed business today.

It also underlines the problem with calculating the scale and importance of European securitisation. As Bank of America’s research team put it this week: “We could not help wondering how a market of €900bn outstanding securitised bonds (about half of them retained by the originator) can generate such a large number of participants; by comparison, the Las Vegas conference attracts about 7,000+ participants representing an $8trn outstanding securitised product market. This is perhaps illustrating the resilience of a securitisation industry that continues to live in hope.”

There are no details available on the deal structure, but this is clearly a deal done by securitisation people on every side of the table. Even the press release makes this clear, quoting Dan Pietrzak, head of private credit (and ex head of securitisation at Deutsche Bank) and Vaibhav Piplapure, who runs the European asset-backed private credit business (ex head of EMEA securitisation at Credit Suisse).

Morgan Stanley was sellside (almost certainly the loan solutions group), and I would be amazed if it wasn’t the securitisation teams at the various legal advisers as well. This is the kind of transaction that explains how 5000 people can stay employed in a market where the public outstandings look unimpressive.

The BNPL product is an obvious fit for PayPal’s existing business — it’s already embedded in the checkout workflow for almost every e-commerce business. Getting this kind of presence (and the associated marketing cost) is the major obstacle for startup BNPL shops. Per the press releases, it processed more than $20bn of BNPL volume globally last year (up 160% from 2021). 

The credit strength of PayPal itself is likely important. Whatever structural features you add to a forward flow to delink it from the originator, you’re still exposed on some level; originator disruption will certainly change the business plan, weigh on servicing, cut volumes, and otherwise cause trouble. BNPL is rapidly revolving credit, and servicing isn’t super intense, so the main risk for a startup BNPL forward flow is really origination drying up — but A-rated market leader PayPal isn’t going anywhere.

Equally, a transaction like this depends on the purchaser being able to deliver size as well. Most structured credit funds have sleeves for asset-backed private credit or private securitisations; the more adventurous funds can buy entire origination platforms if they so desire. KKR’s private asset backed funds are particularly huge, the broader strategy is a major focus for KKR credit (here’s a paper from May enthusing about the opportunity).

So how huge are we talking exactly? That’s where things get murky. KKR’s press release doesn’t mention any external leverage, but we have it on good authority that there’s plenty — major securitisation banks in the senior (we’re just nailing down who but haven’t quite firmed it up for the column) and a debt fund doing the mezz.

One attractive thing about BNPL, from a securitisation perspective, is just how much leverage you can stick on it. The Barclays deal had an 89% triple-A advance rate, Affirm’s US deal this year had 83%. Other forms of unsecured consumer lending might be 70% or below (yes yes not really apples to apples but ).

So it’s very possible that KKR has cranked this up to 90% or so by the time you layer in mezz — a €300m ticket is still a big capital deployment, but that’s the magic of securitisation right there.

We hear rumours, too, that other trades are afoot. There aren’t many BNPL shops with an origination footprint as large as PayPal, but there’s the PayPal US book presumably up for grabs. 

The biggest remaining pool in Europe is Klarna, which has a loan book of around €6bn across jurisdictions. Klarna raised $800m in equity last year at a $6.7bn valuation — not bad, but a very very down round (85% drop from the peak), and it’s in transition. Funding consumer receivables with VC money is inefficient, so it’s been sucking in deposits through regulated banks in Germany and Sweden. Even more efficient, though, would be letting the likes of KKR step in and stick 90% leverage on this portfolio (at a price) and Klarna is in need of efficiencies.

Red Square

Two hobbies of this column are banks’ principal finance programmes, and the rise and rise of Jefferies in European securitisation. These interests, we hope, will shortly collide.

Jefferies, as has been much discussed, has a healthy sprinkling of Citi alumni, and Citi has been one of the most regular issuers from its principal finance shelves. 

There are seven deals under the UK Canada Square brand, and five under the Dutch Jubilee Place. Not all of these have been Citi as such — residuals have been sold, and last year’s Canada Square 7, for example, was sponsored by M&G. 

Then you’ve got the JP Morgan Pierpont shelf, the string of Morgan Stanley RPL deals, the Goldman Sachs RMBS deals (Parkmore Point being the most recent). These aren’t exactly equivalent; the MS and Goldman deals are opportunistic portfolio purchases and exits, while the JP and Citi shelves are based on forward flow arrangements.

But times are tough for the mortgage-based shelves in the UK. Citi’s Canada Square had loan purchasing arrangements with Fleet Mortgages (now owned by Starling), Landbay (now selling into deposit funders), Habito(now back to broking)….as discussed last week, it’s not easy for securitisation-funding mortgage originators to write loans. More terrible UK data, more rates vol, and absolute levels are still just too damn high to make it profitable.

Happily I have been on the right side of a trade for once and managed to refi before the latest round of terrible UK data, so I can write about this with greater calm than usual. If you’re writing loans at low enough spreads to get some business in the door, it will be a struggle to securitise them economically. There’s not enough excess spread in the simple business of buying mortgages and securitising them.

However, if you can find some other assets that are a little juicier (and you have a flexible enough balance sheet to do so) there are still principal trades to be done. Maybe ultra-specialist high returning mortgages, maybe consumer loans, maybe auto HP; there are definitely assets out there where the maths still works. 

You just need an ambitious team with big plans in securitisation, supportive management, supportive risk teams, and skilled distribution. Time’s ripe for a new entrant.

Feeling the heat in Valencia

One of my little projects at Barcelona was a deal called Gedesco Trade Receivables 2020-1, which is, as one correspondent put it “a real dumpster fire of a transaction”. 

I teased Morgan Stanley last week for their involvement in underwriting it, but in general, responses last week were polarised between “what on earth is Gedesco” and “ahhhhh Gedesco” (raised eyebrows, arch expression).

For those in the former camp, it was a trade receivables deal done in 2020 for specialist non-bank originator Gedesco and a few subsidiary brands, including Toro Finance (no connection to Chenavari). “Trade receivables” is something of an expandable category, covering everything from “here’s a load of Volkswagen invoices” to “here’s a load of commodity cargos” to the more SME-inflected origination that Gedesco specialised in — invoice discounting, small scale factoring and the like. SMEs often sell to larger companies, so using their invoices for financing can make a lot of sense……but a bad smell has hovered over parts of the market since the collapse of Greensill Capital, who financed invoices which didn’t actually exist yet, clearly a financial innovation too far.

Actually, calling it “trade receivables” was a bit of a misnomer — even at origination, nearly 60% of the portfolio was actually “direct lending”, though with recourse to receivables. It was Gedesco’s second securitisation financing, following a deal called Castilla Finance in 2016, 100% purchased by lead manager Nomura at closing (though maybe syndicated since). The company also raised financing from CVC Credit on a couple of occasions.

Deal performance bopped along fairly handily for most of the deal’s life — a surveillance report from KBRA said at the end of November that “Although the portfolio has experienced certain spikes in delinquencies since May 2022, the securitised portfolio has not had meaningful performance degradation to date with only €2.7mln (1.0%) of receivables in the 30-60 day past due bucket as of 30 September 2022. Amounts in the 1-30 day buckets are larger at €37.7mln (14.2%); however, this includes receivables that were paid but not cleared at month end. To date, no accounts have been 60 days delinquent and there have been no deferments requested or granted to date.”

This year, however, the deal has fallen off a cliff. 

KBRA again: “reported performance since the transaction commenced its amortisation period has dramatically diverged from the transaction’s reported performance prior to the amortisation period. This deteriorating performance is inconsistent with the observed historical performance based on data provided to KBRA prior to rating the transaction and reported to KBRA during the revolving period, which ended on 28 December 2022.”

Or in graphical form:

So what on earth is going on? Again, according to KBRA, the servicer: “attributed the performance deterioration to a general unwillingness of the debtor pool to pay on time given the lack of refinancing options that the Servicer now has available to offer the debtors in the subject transaction. It further noted that the levels of engagement with underlying debtors has generally declined since the revolving period end date. According to the Company, Gedesco’s sales force had previously maintained a high level of contact and dialogue with debtors to support origination of new business which helped reinforce the servicing activities as sales teams could enquire about upcoming facilities falling due and potentially offer refinancing options or new financings.”

My former colleague Richard Metcalf has done some good work on this over at IFR, and traced the issues back to a vicious legal dispute — JZ International, the investment firm which owns Gedesco, has sued two of its partners, Miguel Rueda Hernando and Ole Groth, over alleged self-dealing. 

The complaint is long and (obviously) legally contested, but the allegations include fraud, conspiracy to defraud, conversion, breach of fiduciary duties. If you’re interested I can send over the document, or you can dig it out from a variety of US court portals.

The allegations draw in Gedesco as follows: “[Rueda and Groth] caused a JZI portfolio company, Gedesco Finance S.L. (“Gedesco”), and a Fund-B portfolio company, Toro Finance S.L. (“Toro”), to “loan” over €80m to companies and projects in which Rueda, Groth, and/or their co-conspirators had substantial financial interests. The typical pattern was for Gedesco and/or Toro, at the Funds’ expense, to fund investments by Defendants’ companies, exposing the Funds to the downside risk of these investments while appropriating for themselves the upside opportunity of the investments in violation of their fiduciary duties.”

The complaint says that Stator Management, allegedly a group controlled by Rueda and Groth, was “operated as a shadow private investment firm”, such that “based in significant part on the unlawful transfer of Gedesco/Toro assets to Stator Management, as of February 2022, Stator Management is valued at €90m”.

Now, most of the NY case so far focuses on jurisdiction issues — should a case involving Spanish citizens and Spanish companies be tried in New York — but there are Spanish legal proceeding as well.

Per IFR: “As a result of the allegations, "Gedesco was unable to renew the financing in the public markets", Rueda said. "If you start telling the banks that someone has committed fraud, the lenders will not continue lending to that company.”

We’re obviously not in a position to discuss the legal case while it is going on, but this is a valid point — investors will run a mile from the potential rep risk associated with this situation. Your basic originator due diligence is not going to go beyond the discovery process for multiple lawsuits across two continents; much easier not to touch it.

The allegations in the court complaint aren’t of the form “Gedesco made a lot of loans to related parties who stole the money”; they’re much more like “Gedesco loans (often with structured elements or warrants attached) facilitated the transfer of companies to a vehicle controlled by Gedesco management”. The companies existed and had value, but the motivations for lending were not necessarily a clear-eyed balance of credit risk and reward.

Whether or not the legal case has any basis, we can discuss the separate issue of what happened to the Gedesco portfolio once it became clear that no refinancing was coming, and the deal was entering its amortisation period. 

Why should these loans have deteriorated so quickly?

Economically, it looks like there’s been a mismatch between the ways the borrowers viewed their Gedesco facilities — revolving, perpetually rolling working capital funding backed by their regular course of business receivables — and the way the securitisation structure and investors view them, which seems to have been more akin to self-liquidating loans with definite maturities. As soon as the Gedesco salesforce stop rolling the loans, the defaults and delinquencies pile up quickly. Perhaps it will be possible to make the deal whole through collection on the underlying receivables, but this, too, may depend on the court case and the continued motivation of the originator and its staff.

There’s nothing intrinsically wrong with refi risk being baked into a securitisation — this is absolutely the case for CMBS and leveraged loan CLOs, which rely entirely on the underlying assets being refi’d. Even in mortgage deals, nobody really expects to hold the bonds while the 25 year mortgages in the pool amortise away.

But this can and should be priced — and refi risk is likely very different for, say, a prime central London office block vs a Spanish midcap which has already turned to the non-bank originators for its primary financing. Presumably if a Santander or BBVA was lined up ready to dole out the cash, these companies would not be setting up receivables programmes with Gedesco. There are other non-bank lenders in Spain, but it’s not an easy thing for these to simply step in.

Gedesco itself has a very well aligned interest in the deal performance — risk retention is horizontal, with the firm holding the first loss — but this, too, circles back to the lawsuit. If Gedesco’s management had divergent interests from Gedesco as an entity, as alleged in the court filings, it doesn’t really help to have Gedesco as a first line of defence in the securitisation deal.

The deal serves as an example of how not to delink a transaction. Securitisation, from first principles, should work as a financing independent of the originator or deal sponsor — true sale, non-recourse, replacement triggers, all that stuff. Everyone knows this isn’t actually the case, and a failed originator (or an originator with legal troubles) is bad news. 

But it seems to have been particularly bad news in this case. Gedesco’s borrowers were dependent on Gedesco to manage their refi risk; in fact, they seem to have been dependent on Gedesco’s saleforce actually getting out there and proactively rolling their loans.

That’s enough Gedesco for one week, but there are several further questions — whose fault is it, and who’s been left holding the bag?

No good options?

It’s been long enough since the global finance crisis that everyone had calmed down a little on the tricky issue of counterparty risk, at least until this year. But then it roared back with a vengeance as SVB and then Credit Suissecollapsed. 

Under UBS’s ownership, Credit Suisse will surely honour all of its obligations, but that didn’t stop several leveraged finance companies drawing down their revolvers in full just in case (coverage for 9fin subscribers here).

The trouble with securitisation counterparties is that you basically want them to be boring and information insensitive. Account banks, paying agents, cash managers should simply be cogs in the machine, functional and unproblematic, no thought required and certainly no credit risk assessment.

In most asset classes, the disruption risks are relatively minimal (the cash mostly runs through the structure) but in SRT and CLOs there’s more reason than most to consider the counterparty. 

In SRT the basic issue is that the bonds in question are often issued by banks. Credit-linked notes are the cheapest SRT format in terms of collateral, legal and structuring, so are generally preferred by the larger banks issuing the product. 

Investors don’t love the format, but have generally preferred to take a better spread and suck up the counterparty risk, rather than see their returns trimmed. It is difficult, though, to consider exactly how to price this up.

Six banks with outstanding SRT transactions have failed in various ways; BES, Banco Popular Espanol, Co-Op Bank, SNS Reaal, Getin Noble and now Credit Suisse. In no case have the SRT transactions been impaired or otherwise affected. 

So does that make SRT transactions like bank senior debt? Being the junior credit risk in a bank portfolio, it feels like bank sub CDS might be a better fit, but it’s definitely more likely that a regulator will impair tier two in a failing bank than choose to impair the SRT. Would an SRT deal be a deliverable obligation in a CDS auction? Not sure.

In terms of actual, effective hedges for bank risk, equity price seems the best proxy, given recent events. The equity death spiral for SVB and Credit Suisse was the most sensitive risk gauge, particularly since it seems like Credit Suisse CDS won’t even trigger.

But equity-based hedging for SRT seems waaaaay too sensitive for the actual bank risk that’s baked in to the product. If you’re running an SRT fund you don’t want the NAV lurching around as bank stocks move. It’s a signal but there’s too much signal

The actual mechanism that’s baked into SRT deals, as in most securitisation, is a ratings-based trigger. That’s widely acknowledged to be completely useless, as you can see from the collapse of Credit Suisse; well capitalised on paper, well rated, basically solid on any reasonable paper-based metric, and toast in less than a week.

You can make them more sensitive by setting the level higher — A+ or above, for example — but this is potentially counter-productive. During the European sovereign crisis, for example, banks were hit by rating caps in certain jurisdictions, forcing some institutions to drop out of the account bank business. This is more of an issue for deal hedging, where providers of bespoke securitisation swaps are few and expensive, but it can still prove counterproductive. 

CLOs are a different animal. The reason I flag them here is because prefunding is far more prevalent — they’re the only securitisation asset class where it’s plausible that you have multiple hundreds of millions raised from investors and sitting in an account ready for deployment. 

Actual market practice tends to mitigate this risk, in a way that other asset classes could learn from; the one month time lag between pricing CLOs and settlement gives managers a window for furious ramping in which they can do their best to minimise the cash still sat around at the issue date.

But it’s still common enough for deals to be 70% or 80% ramped at closing, leaving, say, €80m in an account somewhere ready for action. This is usually placed with a money market fund, rather than purely in a bank account, solving the counterparty problem by spreading the exposure across the usual money market collection of bills, CP, repo and so forth. There’s still a counterparty managing the money fund, but credit risk to the underlying is low and distributed.

I wonder, also, about some of the legacy pre-GFC consumer transactions. These deals often have large reserve funds trapped inside, a tasty nut to be cracked and consumed when and if the deals get called. But where, exactly, is the money? Deals old and new have an “Authorised Investments” concept, allowing excess cash to be deployed into Gilts or other risk-free assets….but I’d put money on the pre-GFC drafting being more flexible than today’s standard, and there may be wrinkles in practice, especially as many of the pre-crisis deal counterparties will have fallen away and been merged or replaced over time.