Prime and punishment

When something controversial happens in markets (e.g. Credit Suisse’s AT1 writedown, Europcar’s duff CDS payout, Deutsche’s economic T2 non-call), people tend to fight about it, and they broadly split into two camps. One is the “read the docs” crowd. Code is law; contracts are what matter never mind how screwy the real-world consequences. The other camp is basically “it wasn’t supposed to be like this”, which tend to lean on softer arguments of market practice, custom, the intended effects of a particular instrument.

Failure to call a performing RMBS isn’t quite as controversial (that’s why deals have step-ups and sometimes turbos) and bondholders spend a lot of time thinking about the possibility. But there’s a layer of market practice and custom on top of the specific transaction documentation and incentives.

Issuers that communicate clearly about their call strategy, even if this includes non-calls, are rewarded, provided investors trust them. So no lingering distrust for TwentyFour Asset Management or NewDay for Covid-era missed calls (both caught up at the next interest payment date). No bad blood for Paragon for the clear and simple policy of calling post-crisis front book deals and leaving pre-crisis ones outstanding. Punishment TBC for Cerberus’s recent missed Towd Point calls; communication so far has been long on aspirational optimism and short on specifics.

Does Charter Court Financial Services (now part of the OneSavings Bank group) get punished? This is a complicated one, but not really?

The residual notes in Precise Mortgage Funding 2018-2B were sold to a third party in 2019; we don’t know who but let’s assume for the sake of argument that it’s a financial investor, hedge fund or sponsor, who’s not terribly invested in the brand name of the PMF shelf, fully reliant on high functioning securitisation markets, and declined to call the deal just a couple of months back.

Even though the call rights were sold, it is still Charter Court’s name on the tin. The decision to call or not wasn’t Charter Court’s, but that just pushes the question back one level — should bond investors punish (in the form of higher spread) issuers who sell their residuals to financial investors vs those who retain them in-house? Surely this increases extension risk?

We’ve mused before how bizarre and backwards securitisation transparency can be; investors can download reams of data about every loan in a portfolio and model payment profiles to their heart’s content but cannot readily find out who their real counterparty is and who owns the call. Which part matters more to the market value of an RMBS note?

Anyway, OSB has been in the market this week with a prime RMBS, CMF 2023-1, the second prime STS transaction from the OSB stable, giving us the chance to see if the market is bothered or not.

I guess the first thing to notice is that the deal includes a 2x step-up (i..e. the note margin doubles if the first call is missed). That’s off market; Aldermore’s Oak No. 4, another challenger bank prime deal priced last week, had a 1.5x step-up. Even the non-bank deals lately in market, such as Pepper Money’s Polaris 2023-1 or Belmont Green’s Tower Bridge Funding 2023-2 had a standard 1.5x (max 100 bps) structure.

But arguably nobody cares. If there’s very little spread in a deal, a big step-up tends to whack the tranching and reduce deal efficiency, because rating agencies stress interest cover to the final maturity of the bonds. But there’s plenty of spread in this transaction, and the triple-A advance rate is within a whisker of Aldermore’s Oak deal, even accounting for the carved-out class X which Aldermore did not include. So, assuming OSB is going to call, the higher step-up is kind of a freebie.

OSB also commits, through the risk retention structure, to keep the call rights in place. The class Z is 8.75% of the deal, so if OSB has to hold a minimum 5% in first loss / horizontal form, it’s always going to have a majority of the class and hence the call option. No ruthlessly economic hedge fund will get a look.

Again, this is kind of a freebie. This is clearly a pure funding trade, focused on bringing cheapest to deliver cash in the door. It’s not about optimisation of capital; even if there was a vertical risk retention in place, OSB wouldn’t be knocking on Nearwater Capital’s door for a 100% repo. It’s no more interested in selling the class Z than Lloyds would be in selling the Permanent sub notes.

The last and most important spot for punishment is in price. Here there’s arguably a spot of weakness — the deal priced at 67 bps from talk in the mid 60s, with a £300m book for the £300m tranche. Oak No. 4 came at 62 bps last week, for a similar three year trade, so there’s clearly some concession.

But this pricing impact may just reflect the tail end of heavy supply in UK mortgages and prime specifically. CMF comes after chunky master trust deals from Lloyds and Clydesdale, as well as Oak itself. If you like BTL or non-conforming, there’s also plenty of supply out there, and a reasonably heavy pipeline heading in to the conference. Some of the big bank treasury accounts can be conservative in the issuers they buy from, and sterling securitisation just isn’t that deep. Lloyds took a billion off the table, but at what cost? 

Maybe the market has just got a little heavy lately, something you can readily see in the mezz spreads for the deals in market this week. Tower Bridge is talked at mh200s/mh300s/m400s for class B/C/D, Polaris came at 275/365/435. Compare that to Lendco’s Atlas 2023-1 in April, at 190/275/380 (admittedly a clean pure buy-to-let pool rather than mixed BTL/non-conforming).

So where does all this come out? We’ve got two freebie giveaways, in the form of step-up and horizontal retention. Three if you count the “turbo”, but this is in Oak as well. Crucially, the freebie giveaways work because this is a fundamentally different trade to the deal that wasn’t called, Charter Court is in a very different institutional place, where bank-style funding trades can take priority. There’s not much punishment here!

Straight to voicemail

We’re going to keep talking about call options but flip it to the CLO market, where the dynamics are pretty different. Both asset classes reflect their underlying product; mortgage borrowers have strong incentives to call at the first available date, leveraged loan borrowers have tons of optionality but keep it all for themselves. There’s no expectation that they will call, but should they want to, any time is good!

But it’s a live issue for the CLO market because that’s basically the only thing that could shake up European leveraged finance in the short to medium term.

Despite the recovery of financing markets this year, M&A remains lacklustre, meaning new issue LBO supply is scarce-to-non-existent in euros. There are flickers of life in the visible forward pipeline, but it’s pretty thin stuff. 

Anecdotally, leveraged finance banks are talking up their pipeline of public-to-private (P2P) deals, which a cynic might say is the perfect supply category for these troubled times. P2Ps target listed companies so must be prepared in the utmost secrecy, unlike corporate carve-outs, where sale processes are semi-public and every levfin bank will be crowding around. Valuations might move and deals might never launch, so who’s to say if these fat P2P underwrites really existed or not? 

Meanwhile more and more CLOs are slipping out of reinvestment periods and starting to deleverage. We wrote in our CLO Outlook for the year “Things Can Only Get Better” that some 36% of the market was due to be out of reinvest by year-end (sourced to BNP Paribas’s research desk) which is quite a striking fact.

Let’s assume CLOs are 70%-ish of the European leveraged loan market (maybe more, giving outflows from SMAs during 2022). If 36% of CLOs are out of reinvest, then about a quarter of leveraged loan AUM is going to be semi-passive by the time 2024 dawns.

(this doesn’t account for different volumes, deal counts, repayment of loans etc; as with all numbers in Excess Spread, the calculations are of a “back of a cigarette packet” variety).

The obvious way to bring this situation back into balance and harmony is for lots of deals to get called. Make all the passive money go away, shake loose some loan supply, make markets liquid again and reset the balance between CLO liabilities and assets.

However beneficial this might be to the overall market, nobody is going to be calling deals out of pure charitable inclination.

Per Bank of America: “We think it is reasonable for the equity investors of deals out of the reinvestment period that are starting to de-leverage but have decent NAVs to call the deals, however the universe of deals where this option might be exercised is currently small and it would require loan pricing to hold or indeed improve from here for it to grow.”

It seems plausible that loan pricing could improve, as CLO issuance continues to outstrip net leveraged loan supply, so the incentives could improve in time.

But perhaps a more important dynamic is the extent to which distributions are still sloshing down to equity in the 2017-2018 deals that might otherwise seem like plausible call candidates. Deleveraging is proceeding in very sedate fashion, so capital structures remain efficient and there’s still plenty of spread coming through. 

Equity distributions have been knocked about a bit by base rate mismatches (Euribor has moved so rapidly that there’s a massive gulf between the interest coming off three month and six month pay collateral obligations), but this effect is likely to roll off at the next equity distribution date, normalising the payments coming through.

If you’re in 2017-18-19 vintage equity, what’s the rush? Sure, it might be useful for the market if everyone calls and liquidates older deals, but there’s no incentive for most third party equity investors to act, even if they do have control positions. They’re better off just clipping coupons until the deals have amortised much further.

Deleveraging in post-reinvestment CLOs is very sensitive to conditions in loan and bond markets. How many points separate issuers doing A&Es from issuers doing full refinancings?

Refinancing means deleveraging for the CLOs — principal proceeds go to pay down the triple-A after reinvestment. But an A&E will either drag along CLOs which can’t approve it, or leave them stuck behind in a stub tranche. Either way, no deleveraging until the stub gets knocked out.

Nouryon and Merlin are both big capital structures in European leveraged finance, and both credits which are showing decent performance. But they sit on opposite sides of the A&E / refi divide, with Nouryon opting to push its 2025 TLB maturities to 2028 in an A&E process, and Merlin issuing a new bond to pay down 2025 SSNs and “other secured debt” (potentially some of the dollar TLB). So it’s clearly balanced on a knife edge — a little more improvement and the deleveraging might kick in in earnest.

Equity trading has been fairly active (and fairly public, with more positions on BWIC than usual), so evidently some positions are changing hands. But this still doesn’t mean a raft of calls incoming. New holders might have different incentives, but there’s still no reason to call. A bigger driver for the supply is likely to have been the April 15 payment date, which saw distributions paid out for a large chunk of the market.

In fact, CLO trading in general has apparently been very active this week, with nearly €500m of bonds out on BWIC, compared with €230m last week, and a good supply of bonds up and down the capital structure. This hasn’t smacked spreads too much; primary supply is balanced rather than frantic, market conditions are good and perhaps it’s just a bit of profit-taking. Sell in May and go away (to Barcelona).

The time is now

If you’re a bank without a CLO operation, what are you waiting for? 

It’s the most resilient asset class in European securitisation. Against regulatory hostility, difficult arbitrage, an illiquid and undersupplied loan market, it’s still going from strength to strength, with more new managers joining and issuance, even last year, tracking a long term trend upwards.

So there’s money to be made, if you want it — and the extraordinary rise of Jefferies from non-existent to top of the league tables shows that it can be done. My old shop gave securitisation boss Laura Coady an “Outstanding Contribution Award” to reflect this achievement. Jefferies admittedly went hard on hiring, and lifted the entire top team from the market leader Citi, but it doesn’t really have a balance sheet, and it is genuinely rare that a bank enters a new market quite so successfully.

Now, perhaps, there’s some market share going spare — Credit Suisse was formerly a high ranked CLO bank, but the team there either jumped or were pushed. Dan Bates and James Gray in trading and sales went to Baird, while Dimitris Papadopoulos who headed CLO primary eventually landed at Deutsche Bank. And of course, Credit Suisse itself then collapsed.

Speaking of which, we should congratulations Jason O’Brien and his team, who formally transitioned over from Credit Suisse’s UK securitised products business to Atlas SP Partners this week. This was on the cards since the spinout of SPG to Apollo was agreed last year, but we’re hoping for great things — according to Atlas itself, in the first 100 days since the new platform was set up in the US, they’ve executed 26 securitisations and signed six new warehouses, so no doubt Jason et al will be hitting the round running. Have a look here if you’re interested.

Anyway, for banks of a more commercial, balance-sheet-led disposition, breaking into CLOs requires a different playbook from the Jefferies approach — a combination not just of CLO sales and trading and structuring, but also, ideally, buying some bonds. If you’re number 12 in the market that’s not a great pitch, but it gets better if you can speak for €200m of triple-A investment as well.

Some of the top tier arrangers also ended up buying bonds from their deals last year, but in general you’d expect this to be particularly compelling for banks with cheap deposit-funded balance sheets. That doesn’t necessarily mean buying through treasury or CIO type books, either. Societe Generale’s efforts to broaden its CLO primary business from the US to include Europe came with some capital committed through the asset-backed business to buy senior bonds, for example.

So who’s missing from the European CLO market? Let’s pore over the list of securitisation banks in Europe…. taking Finsight’s list of top banks in EMEA securitsation, we can tick off all the US institutions, BNP Paribas, Barclays, Deutsche Bank and Natixis, all of whom have CLO franchises. SG we already discussed. 

So that leaves Lloyds, Santander and HSBC missing out on the action. The sterling-led franchise at Lloyds might not be the most natural fit for CLOs, Santander is building out in consumer products since Matt Cooke went over in 2021 (Dmitrij Levitski is heading over shortly to get the ex-Lloyds band back together)….but is there an opportunity for HSBC?

It’s been actively trading CLOs in secondary markets for years, and has definitely looked at establishing a primary franchise before. My memory gets hazy but there was an active project to get into the market in the early days of CLO 2.0 — and even an arranger mandate for Investcorp’s Harvest X, priced in November 2014. 

Nearly a decade later, time to get back in the water?