Holmes has gone straight from announcement to bookbuilding, for a three-day execution — presumably most of the relevant investor base have some idea whether they want to buy some HMI and don’t really need walking through the deck in detail. One doesn’t want to tempt fate (remember Northern Rock etc) but Santander has been repurchasing loans with more than one missed payment, it’s low LTV, never missed a call, this thing is clean clean clean and highly credit-worthy. 

The excitement will be to what extent it prices back of levels for the recent five-year Perma bonds, what the book reveals about total sterling RMBS demand levels, and the current basis between master trust RMBS and standalone prime in the shape of the slightly shorter Friary deal next week. At the second book update we had £850m of orders for Holmes for a £500m tranche, rising to £1.2bn for a £750m tranche later that day, 1.6x coverage, which seems healthy and not a total bunfight — presumably investors are keeping their powder dry for the issuance to come.

The most interesting transaction entering the pipeline is a new Portuguese credit card ABS for WiZink Bank, Vasco Finance No. 1. 

There are precious few Portuguese credit card deals of any kind — WiZink itself generated the only transaction so far, in the shape of Victoria Finance No. 1 in 2020, which was retained (giving it €392m of ECB collateral from the class A notes).

WiZink is certainly a frequent issuer — as well as Victoria Finance it also securitised personal loans in Viriato Finance No. 1, and has printed six Spanish credit card deals through WiZink Master Credit Cards and two transactions through aZul Master Credit Cards, the latest a retained deal from July this year.

But it’s a name with a certain amount of hair associated. The bank is part-owned by Varde Partners, which funded the acquisition through PIK toggle notes from holding company Mulhacen, but following a restructuring last year, it’s now partly owned by former bondholders as well. 9fin’s Restructuring QuickTake had the following to say about what happened.

“In 2018, Mulhacen issued €515m of 6.5%/7.5% PIK toggle notes due August 2023 to finance the purchase of the 49% stake in WiZink Bank. The notes were totally reliant on dividends being upstreamed by WiZink to Mulhacen. In 2019, the sponsor Värde committed to the Bank of Spain to retain a cash reserve from the dividends over a five-year period, up to a total of €250m by the end 2023. It agreed to build a reserve amounting to 20% of the loan principal (€50m) by the end of 2023.

Cash interest was being paid until March 2020, with the PIK toggle activated in September 2020 and interest deferred ever since, with the principal increasing by €56.4m to €548.6m as at end-September 2021. In late 2020, management had declined to give future guidance and confirmed that no dividend would be paid during 2021 to Mulhacen, the Holdco.

WiZink was severely affected by changes in regulation following a ruling from the Spanish Supreme Court on March 4, 2020, which claimed that WiZink’s interest rates (26.82%) on its revolving credit cards were excessive. In January 2021, new rules regarding the commercialization of the revolving credit cards came into force, introducing new transparency requirements and affordability assessments before cards contracts are signed. In addition, it forces payments to amortize a minimum of 25% of the credit used. The Ministerial Order is intended to protect customers from abusive terms, and to prevent them from falling into unsustainable indebtedness. The maximum interest rate that can be charged was reduced to 20% - causing a drop in interest income.

This has resulted in a significant number of usury claims brought by WiZink customers. It increased total provisions by a further €85.2m in September to €191.4m. It is anticipated that further provisions will be announced in the fourth quarter, hence the need for further capital, noted the source close, amid concerns that this would mean a breach in the 13.8% total capital ratio requirement.”

This unhappy situation was followed by a debt-for-equity swap at the holding company, plus an extension of existing bonds from 2023 to 2026, and a €250m capital injection into the bank.

The investor deck for Vasco Finance draws a veil over this ugly period, with the only notable events for 2022 apparently being the launch of a market to digitally sell third-party products, and a partnership with RACC Mobility Club.

In a way this is perfectly valid — nothing especially bad happened to WiZink the regulated bank during this period (unless you worry a lot about regulatory capital ratios) and certainly not to the asset-backed financings. Still, one might think it merits a small mention.

The other wrinkle is that the usury regulations which tripped up the WiZink holdco are still very much in force, in both the Iberian markets where it operates. Market interest rates have moved a bit in the last year, but the Bank of Portugal’s Usury Cap has hardly budged. From the Vasco deck:

Maybe this doesn’t matter in practice. We’re talking about credit cards here, and there should be >9% of excess spread through the transaction. If you compare this to say, NewDay, operating in the UK, a market without the same usury regulations, I get 8.22% excess spread for NewDay Funding 2022-3 in the latest investor report, and generally figures in the 10%-12% range. But usury laws did, y’know, cause a whole debt restructuring at the holdco, so probably still worth looking at.

The drags don’t work

We’ve discussed the “Snooze Drag” provisions in CLOs a few times here — first in our Deep Dive into CLOs and A&E requests, published in January, and more recently in relation to Bloomberg’s nice piece coining the snappy Snooze Drag terminology.

This refers to the idea that, even if CLO managers are constrained by their WAL test from directly voting in favour of a maturity amendment to a loan, they would generally quite like to remain invested in a recouponed facility reflecting current market credit spreads, and would therefore happily abstain from any vote and yet be rolled over into an amended loan. This benefits sponsors as well, and investment banks running A&E processes would also generally prefer to see high participation levels and small stubs remaining.

But it’s not very popular with CLO triple-A investors. They would like to see their legacy bonds in post-reinvestment period deals get on and amortise. Put simply, they want old bonds at 100bps paid down so they can reinvest the money at 170bps. Staying in a longer loan with a higher margin is equity-friendly, while getting paid back early is debt-friendly

There’s some informal pressure senior investors can apply to managers, which can be especially persuasive if said managers are also marketing new issues and need anchor orders.

But in these new issues, there are also documentation changes which can be baked in — as demonstrated by Onex’s latest new European issue OCPE 2023-7. This was priced 18 August, before I went away, but in my defence I’m only going through the docs now.

Anyway, it includes a couple of tweaked terms. Here’s one of them:

“notwithstanding any other paragraphs of this definition, in relation to a Collateral Obligation subject to a Maturity Amendment which would contravene the requirements of the applicable provisions set out in the Portfolio Management and Administration Agreement and which the Issuer or the Portfolio Manager on its behalf had the option to vote against or abstain and they abstained from such voting, and with respect to which by way of a scheme of arrangement or otherwise the Collateral Obligation Stated Maturity has been extended, and which the Issuer or the Portfolio Manager on its behalf have not disposed of prior to the Maturity Date, the lower of its Market Value and [75.0] per cent. of such Collateral Obligation’s Principal Balance as of the relevant date of determination;”

To translate the legalese — if the portfolio manager abstains from voting, the loan in question gets a fat haircut, hurting the par value metrics for the deal, and bringing interest diversion triggers closer. Clearly this deal is a brand new issue, so it’s a long way from bumping up against its WAL test (and we’re a long way from seeing how effective this language is). But marking at 75 or lower is extremely punishing, and will effectively close off the “Snooze Drag” if it’s tried in 2028 or 2029.

The debt-equity arms race continues!

CLO managers are cheap

The fight over the fate of Sculptor Capital is a fantastic popcorn munching occasion involving some of the biggest and most colourful names in the hedge fund industry. But it also throws up a rather unappealing datapoint on the current valuation for a CLO management business.

I’ve written a more detailed piece for 9fin subscribers breaking down what we know about the Sculptor business, but here’s the skinny.

Over the past 18 months, two CLO managers with roughly $15bn in CLO assets under management have been sold — CBAM to Carlyle, and Assured Investment Management to Sound Point. The Sculptor fight is continuing, but per the proxy statement, “Bidder H” has offered an (uncommitted) purchase price of $260m for the CLO platform… which also has about $15bn in CLO assets under management. 

There’s only a limited ability for CLO managers to differentiate themselves on pure earnings prospects — whichever manager you buy, the hard revenues are essentially going to be between 0.4% and 0.5% in management fees on the CLO assets, plus some amount of uncertain incentive fee income. An acquirer might keep the people and the credit platform in place, or simply transfer the management contracts and use existing analyst resources to manage more AUM, but either way, there’s limited latitude for a given CLO manager to earn dramatically more in management fees from the same AUM.

How, then, to explain the fact that Carlyle paid $794m for CBAM in April 2022, Sound Point paid $428m for Assured Investment Management in April 2023, and someone (Bidder H is an asset manager, we suspect a shop with an existing credit franchise) is offering $260m for Sculptor?

No doubt performance varies and reasonable people can disagree about the quality of platform, performance potential and so forth. But all three price points concern transatlantic players with established and successful operations. We’re not talking about three men and a Bloomberg here, but the valuation gap is massive.

Tougher macro plays a part — default expectations have risen since the CBAM deal was negotiated, which hits the prospects for incentive income, and could even, in remote cases, switch off subordinated management fees. Slower LBO supply limits CLO arbitrage and, I guess, CLO formation (not that this seems to be happening in practice) so maybe the growth path for a given platform is slower? 

Still-wide tranche spreads also make equity less valuable. Selected 2022-vintage reset options are now in the money, but most of the reset options for 2018-2019 vintage aren’t, with the well-documented effect that CLOs are leaving their reinvestment periods. The reset options themselves have a value, and there’s also a growth element here too. If a given CLO manager is going to see active AUM dwindle, that’s bad news and will hit valuations.

Higher rates have a duration effect as well — a CLO management business is basically just a low coupon fee stream stretching 5-10 years out into the future, so NPVing this with higher rates gets you a lower number.

The maths here is well beyond my capabilities — and some well-placed sources have also suggested the Sculptor number looks cheap — but if the datapoint is valid, it surely shows its a poor time for potential sellers of CLO platforms to explore their options.

Same stuff different wrapper

I had the benefit of an education at a very ancient university richly garlanded with obscure rules and traditions. This was an excellent preparation for trying to understand financial regulation as a full adult, as it is a similar accumulation of nonsensical precedents from long ago. 

So let’s discuss the recent Second Circuit decision that loans are not securities. For reasons basically of tradition, two instruments which are now identical in economic purpose fall into totally different regulatory buckets, with different rules governing underwriting, issuance, distribution, trading, disclosure, and even how they are packaged (or not) into CLOs.

The decision has been widely celebrated by the industry; the Loan Syndications and Trading Association called this a “great — and critical — result for the leveraged loan market”. 

One might describe this as a great result for the LSTA, which would presumably have to dissolve itself if it had gone the other way, but it also seems to have been celebrated by the industry more broadly, which is keen to avoid the massive amount of work that recharacterising leveraged loans as securities would involve. The path of least resistance is the status quo.

Here are some of the key disasters that might have come to pass, per the LSTA:

“This has a number of implications of course but specifically highlighted here that i) the borrower would need to be an eligible issuer and, if a private company, it must provide certain financial information upon request by lenders; ii) buyers of 144A notes must be QIBs; iii) the borrower and arranger would need to make clear to potential syndicate members that they may be relying on 144A and ; and iv) to avoid general solicitation, the borrower, arrangers and agents should avoid talking publicly about the loans prior to completion of the syndication.”

“Second, the bifurcated private-public model of loan trading likely would end and the loan market likely would become a public-only market. Third, were loans subject to FINRA rules, secondary trading (and settlement) could become more complex and potentially slower, not faster.”

Basically these all sound like good things? 

How terrible for the poor old borrowers, having to provide certain financial information upon request! How terrible that syndication would have to avoid general solicitation! It’s also baffling to me that ending the public-private model of loan trading could possibly be considered a bad thing. It’s not like a market where no participants have MNPI is some weird hypothetical we have to imagine. Pretty much every other financial market except commodities operates on this basis, and most of them work better than leveraged loan trading. The idea that settlement would get worse is also… well, how many initiatives over the past couple of decades have promised to improve loan settlement times? How many have worked? How easy is securities settlement by comparison?

Presumably the implication is also, given that HY bonds and leveraged loans are more or less economically identical, that bond regulation regimes are too harsh? What is the purpose of the restrictions and limitations provided by Rule 144A and the QIB regime? If loans aren’t securities, why are securities securities? Looking forward to prompt regulatory action to amend these rules.

It makes about as much sense as holding an old guy’s fingers while he mumbles in Latin so you can get a degree.

It’s a US judgement and US regulation, but one could look to Italy for an interesting test case (there’s probably a PhD in this for someone).

Italian tax law basically prevents the use of the distributed Term Loan B to finance LBOs, and requires Italian LBOs to use FRN structures (documented as New York law 144A high yield) instead. These really are designed to be as close as possible to a TLB, so Italy vs everywhere else would be a good test case for figuring out if the different regulatory regime affects access to financing in a meaningful way.

In a similar vein, the boundaries between “private credit” and syndicated loan markets continue to dissolve. My colleague Josie, who runs European private credit at 9fin, had an excellent piece looking at this arbitrary dividing line, and citing a CLO manager buying a piece of a recent jumbo direct lending loan.

What if you did private credit, but focused on large cap companies? Maybe investment banks could make markets in the loans, and sell them on to CLOs. Obviously you’d want to make sure it was broadly distributed, and you’d probably want to avoid maintenance covenants, otherwise it wouldn’t appeal to sponsors. Sounds like a pretty cool and innovative debt product someone should invent.