The typical Pimco trade is to buy a pool of assets, which can be legacy or new issue, beaten-up or clean, and structure them into a securitisation, which is taken down entirely within Pimco. So we've had huge slices of UKAR mortgages, originated by Northern Rock and B&B (Slate, Chester, Jupiter), Irish mortgages (Roundstone, Fingal, Glenbeigh), mixed third party BTL and owner-occupied (RMAC No. 3), incredibly ropey back books (Bridgegate Funding), mixed near-prime and non-conforming (Kensington back book), legacy UK non-conforming (Warwick). It's been in the process for the Barclays Peninsula portfolio as well (legacy Italian mortgage).

Some of these deals come in ready-packed securitisation format (RMAC for example, or Kensington back book, which was already held in distributed securitisations), and some of it requires Pimco to hire an investment bank to provide risk retention, reps and structuring services.

The crucial thing is to get from whole loans to 144A/Reg S bonds with ISINs, suitable for the retail bond funds which make up Pimco’s Income Fund complex. 

One might ask how Pimco manages to price a book of assets that it has absolutely no possibility of publicly distributing; the sizes are just too large. It gets marks from the investment bank, who stands ready, in theory, to buy these bonds in reasonable market size.

But the securitisation structure is irrelevant to the overall return on assets. If Pimco structures itself a very debt-friendly structure, it gets a return through the senior notes it owns instead of the residuals it also owns. It does matter for investors in the individual funds; the Income Fund complex comes in different flavours, some riskier than others, some short term, some longer. The allocation of Pimco portfolios to these funds is not perfectly pro-rata.

The presence of Pimco, though, has implications for how to structure and execute a portfolio sale. Pimco's ideal situation is to buy a full stack, including seniors. Since rates rose though, the senior slices of asset disposals have started to look like pretty sweet pieces of paper for banks to hold. 

So Lloyds holds the senior notes in Bridgegate and now the new "Performer Funding No. 1" (we love the optimism in the name; we will also enjoy writing headlines about it if things go south in a couple of years). Barclays holds the senior in Gemgarto 2023-1 (a rare non-Pimco deal) and will, we understand, keep the senior in the Italian transaction.

“There were definitely bids out there for the triple-A, but our main aim was not to raise funding,” said Monica de Vries, head of capital optimisation and securitisation at Lloyds. “For funding, we’d consider other routes, for example Permanent — if you look at the price, you’d see on unsecured personal loan seniors vs a mortgage deal then it’s clearly more attractive to do an RMBS.”

Even if Pimco can’t take the lot, though, it will take the whole of the mezz, which can allow it to outcompete its leverage-requiring rivals in certain circumstances. Performer, for example, a £3bn portfolio of unsecured consumer lending, which Lloyds was marketing this autumn (see here for a little taste) was particularly sensitive to the assumptions baked into it. 

If you bought only the bottom of the stack, you might be buying “the greatest trade of your career, or a widow-maker”

We’ll admit to not modelling it out, but you can get a little sense of it from the ratings. Moody’s and DBRS are pretty much in line down to the triple-B, but class F is split-rated Caa1/B and the X note is Caa1/BB (low), reflecting quite subtle differences in stress approaches between the agencies.

Pimco can take the whole stack, so this sensitivity is cushioned, while the securitisation funds that might want residuals and junior mezz don’t necessarily want the extreme volatility baked into the structure. There are ways around it, including partnering up with different investors, but none of these are ideal. 

Even if you can get comfortable with the sensitivity at the bottom of the stack, syndicating the mezz Kensington-style offers a bit of a hostage to fortune; if the junior is already sensitive to macro and portfolio assumptions, then adding a layer of uncertainty over market conditions on the mezz makes it more challenging still. 

Let’s be clear, the Perfomer deal is very large, and very cool. The class B alone, at £227.5m, is a plausible size for an entire mortgage deal. Fat mezz tranches of £159.3m and £106.2m for class C and D are so large as to be unheard of in publicly distributed securitisation.

It’s also, hopefully, precedent-setting. Not in the sense that it would be nice if Pimco kept buying everything, but seeing cash securitisation used for risk transfer in this way is a good way to get bonds into the market. Deconsolidation and risk transfer is the secret weapon of securitisation, the one treasury tool that can deliver these outcomes.

In a legacy portfolio like Bridgegate, it’s easy to see why you’d sell. 

But Performer is different. It’s a front book deal for a business Lloyds continues to be keen on, a deal of the kind that’s increasingly the bedrock of the market. Much of the euro supply in 2023 has been cash risk transfer deals from the likes of Santander, BNP Paribas, Societe Generale and Credit Agricole; add in the UK clearers doing these deals and there should be more bonds for everyone (if Pimco doesn’t buy them).

“The key focus for this is reduction of risk weighted assets, but the crucial difference between this and Typhoon/Bridgegate is that the strategy for the unsecured personal loan book is still unchanged,” said de Vries. “The unsecured business is crucial to the wider group strategy, and Lloyds will continue to be a very active participant. But it was very important to us to achieve derecognition of the loans from an accounting perspective, and for that we had to see the equity and the majority of the mezz notes.”

Most intriguing of all is the possibility that Pimco starts to show some of its bonds out to market.

The Jupiter portfolio is up for refinancing in the new year, and we hear there’s a chance investors will be offered bonds. It’s probably on a reserve price basis — Pimco clearly has no problem holding the lot, so best feet forward — but it would help market liquidity if some of the Pimco bonds started to trickle into other hands.

It’s an interesting fact about the world that Pimco has no competitors taking the same approach. You need a lot of scale to do it; the Income Funds have assets under management that look like medium-sized country GDP figures, which helps if you want to do multi-billion portfolios without leverage.

But though Pimco is enormous, there are other giant multi-asset bond funds out there, often with securitisation teams on hand. European consumer assets are good value vs the US, where non-recourse mortgages, weaker employment protections and a more volatile economy make for a more complex backdrop. Pimco stands alone for now, but why?

House of straw

Debt purchaser Lowell has issued the third deal in its Wolf Receivables programme of RPL transactions. We’ve written about this initiative before, because it’s potentially a neat bit of securitisation problem-solving — all of the high yield-funded debt purchasers are facing capital constraints, and searching for ways to adapt their business model to a high rates backdrop. 

It’s a case of rapidly repriced liabilities against slow-to-reprice assets — financing portfolio purchases at 12% IRRs when your senior debt is yielding 15% isn’t much of a long term business plan. 

One way to thread the needle is to just do less purchasing (hence Intrum’s pivot to ‘capital-lite’ fee business, back book sale and attendant CFO defenestration), and another is to sweat the assets you do have. That’s the Lowell route, which has encouraged it to print three securitisations of reperforming loans under the Wolf brand, raising funds and deconsolidating portfolios where the heavy servicing lift from NPL to RPL has already been done.

Lowell needs the money. Actually, it needs a lot more money, with £2bn of senior secured bond maturities in 2025 and 2026, all trading at yields north of 20%.

The trouble is, these burdensome debt maturities are starting to impact the deals themselves, with investors who’ve looked at the deal casting a critical eye over the degree to which the securitisation is insulated from any parental problems and Lowell’s interests align with theirs.

Securitisation funding is non-recourse, a legal term of art meaning securitisation investors can only access the assets in the securitisation, and can’t cross default or go after the assets of the rest of the company. But that doesn’t mean it’s not credit-linked, and servicing is the connection that forces securitisation investors to contemplate the credit quality of the parent company.

The Wolf collateral is quite special. It’s the portfolio Lowell bought when it acquired Hoist’s UK operations (originally financed by Deutsche Bank, hence the switch in mandate from Alantra, which arranged the first two deals, to DB for the third outing).

The loans have seasoning of nearly 18 years, average balances around 2.5k. These are credit cards debt that have been hanging over people since before the iPhone was invented, with the obligors probably bounced through multiple layers of debt collection bureaucracy. The current balance is £630m, purchase price into the securitisation was £136m.

It’s certainly financeable (the loans have all made at least four payments in the last six month) but the servicing has probably been….intense.

So how can you delink a structure like Wolf from the performance of Lowell itself? 

One thing you could think is: Lowell is basically an ok business with too much debt. If Lowell can’t address its 2025s, it will go into a restructuring, but this will be a sensible-ish restructuring with grown-ups on both sides of the table. It’s not a fraud or vanity project, it’s a financial investment by an experienced sponsor, Permira, and if Permira ends up losing the keys or the HY bonds take a monster haircut, why should a securitisation investor care? The liability structure gets remixed but the business keeps functioning. 

In Wolf 3, there’s a fairly broadly drafted Insolvency Event for Lowell, covering a range of restructuring scenarios, and giving senior noteholders, if they desire, the ability to replace the servicer. 

This is a bit of a fudge, especially with an intense portfolio like Wolf. If Lowell has a secret sauce, it’s somewhere in the servicing capabilities and data management. This is supposed to ensure good collection outcomes, and enhance Lowell’s ability to price new portfolios. The idea that another firm could just show up and seamlessly take over collecting unsecured debt payments from people who, let’s remember, haven’t cleared this debt for 18 years, seems optimistic.

The gold standard for servicing backups is the “hot backup”, which is basically a shadow servicing system with all the same data and capability in place and ready to go. Unfortunately, it’s so expensive as to be completely impractical. It requires a massive amount of redundancy, especially in an RPL transaction with heavy-duty servicing to begin with. 

Senior servicing fees in the latest Wolf are 5% of collections, with a further 5% of collections to be paid after the senior notes, but before any principal is paid on the junior.

Senior notes are paying 395bps over Sonia — so north of 9%, not a bad yield for a single-A bond, and materially inside Lowell’s other secured funding. 

The issue price for the £126.56m junior is 18.71 with a 3% coupon, so a 16% running yield. Lowell says it is targeting net IRR of 20% on its portfolio purchases for 2023, but this, to be fair, is an older vintage portfolio (Lowell acquired Hoist UK last year, before portfolio pricing had really adjusted to the new environment).

Lowell would probably like servicing cashflows stretching out to the horizon. Investors in the junior notes, however, want Lowell to call the deal on time — the step up date is in 2029

In previous deals, Lowell has kept a 49% minority in the junior notes — enough to achieve accounting deconsolidation, but sufficient to ensure alignment of interest and participation in any upside. In this deal, documents suggest it’s planning to sell the full 95% of the class B it’s permitted to.

So we now have two routes to the “capital-lite” debt servicing model — Intrum’s back book sale, and Lowell’s sale-through-securitisation. So far, Lowell’s ahead of the game with its three transactions; will the Wolf pack grow some more?

Conceal don’t feel

….lines from a classic song about CRE valuations in 2023. 

It seems that Blackstone has finally struck a deal over the troubled Frosn 2018-1 CMBS, a mixed (but mostly office) deal which part-funded the acquisition of portfolio company Sponda. 

This was the first European CMBS to pass loan default earlier this year, though by no means the worst deal out there; the loan default seemed to be more the result of process screw-ups rather than egregious underperformance. Plenty of other deals managed to buy off noteholders and pass extensions and waivers to keep the show on the road, but in Frosn there was too little offered too late to bring senior noteholders onside — a 125bps margin increase, commitments to prepay noteholders with sale proceeds, and a six-month sales target.

Last week though, noteholders passed an extension of the senior loan by two years, with an extra year option, and a legal final extension to 2031. The package includes a 150bps margin bump, a €5m guarantee from Sponda’s holdco, and contingent value rights for noteholders. 

These basically entitle noteholders to a proportion of any money paid out to Blackstone — 20% if the payments come out before the first extension deadline, 25% thereafter.

This was a consensual restructuring, which will bring the loan out of special servicing and restore the deal to normal functioning (though the class X, usually structured to compensate the underwriting banks will switch off).

The latest portfolio valuation is €341m, producing net rental income of €26.9m — 7.8%, by no means the worst thing in commercial real estate right now, especially if you see the 48% vacancy rate through a “glass half full” approach and assume that at least some of it can be let.

Unfortunately, the valuation was considerably higher at issue, and these properties are supporting €295.2m of net debt, for a totally unfinanceable 86.5% LTV. So the current business plan, which is basically to sell everything, is probably the only way out.

To what extent is this a template for the likely way of CRE stress to come? It’s always property-specific. Secured CRE debt is fairly robust and difficult to compromise, but is it more important to keep flexibility, or keep costs down? Is it better, as a sponsor, to delay and pray, or to walk away? How much capex is needed to get assets up to current environmental standards? What’s the market backdrop like?

This deal looks more like a 'walk away' transaction — noteholders accept a margin bump only slightly above the rejected February offer, in return for participating in any exit returns. Delay and pray deals probably accept higher interest payments in the hope these are temporary, and other portfolio restrictions can be dodged.

It’s possible there was a better deal on the table in February — coming in with a bigger margin bump could have brought the senior notes onside immediately and waved through an easy extension. Blackstone might, however, have been more focused on dealing with the €103m mezz loan, repaid at the back end of 2022, and perhaps rightly so; this could have been much nastier with a mezz lender in the mix too, and there’s more risk of losing the keys.

In other office news, Bloomberg has a good gallop through the struggling London office scene, touching on the Citypoint building, among other things. Citypoint is securitised in Salus (ELoC 33), and according to a notice on Wednesday, agreed a package of changes to allow sponsor Brookfield to get on with a sale or, more optimistically, a refinancing.

We talked about this transaction in September, highlighting the difficulties with hedging — the 2.5% cap with Wells Fargo had been extended twice with an expiry in January 2024 — and the fact that senior bonds were paying more than the asset yield, at current valuations. The package of measures includes extending the 2.5% cap for another year, “purchased using funds outside of the senior loan structure”, a cash injection by any other name.

If there’s one trend that’s nailed on for 2024, it’s stressed office financings — we’re looking forward to getting stuck in.

Heavily fortified

Stresa Securitisation, an Italian RMBS for Fortress priced a couple of weeks back, was a throwback to the glory days of securitisation, when packaging assets into ISIN bonds somehow made portfolios worth more, allowing sponsors to securitise and stride off into the sunset. 

This is an unfashionable concept these days, not least because of the risk retention rules, which are supposed to stop exactly that sort of behaviour. 

Still, if you can pull it off, it’s pleasing. Outsize returns are available if you can persuade someone to pay more for something than its worth. Front book securitisation can sometimes be quite collegiate and cooperative, but legacy assets are spicier. 

The portfolio for Stresa was originated by Meliorbanca in 2006 and 2007, securitised in Borromeo Finance, and then sold to Fortress in 2017 and 2018, financed through a private securitisation, also called Stresa. The new deal sought to refinance that original Stresa deal — but the question, for investors, was whether this was a 'walk away' deal that would cash Fortress straight out.

In the event, the senior notes ended up with a spread of 250bps, as compared to a senior coupon of 120bps with a 100bps step-up, so the call is out of the money day one.

With a book that was 1x done, the €123m tranche was presumably something of a niche interest — investors may have lacked lines for Italian RMBS (nearly all broadly syndicated Italian deals are autos, consumer loans or other assets like salary sacrifice), or may have lacked lines specifically for beaten up legacy portfolios like this.

The headline of the DBRS rating calls it “an existing portfolio of mostly prime and performing Italian owner-occupied mortgage loans”, but there’s 13% in three month arrears, 5.6% defaulted, and, since 2017, only 49% of the pool have never been in arrears. The agency notes the pool is “now negatively concentrated in borrowers that have not been able to prepay their mortgage”.

To be honest, 250bps doesn’t seem terrible for collateral like this — if you’ve got liquid triple-A CLOs at say 160-170bps, add some premium for the lower rating (double-A with S&P and DBRS), add some illiquidity premium, add some premium for structural complexity / low coupon legacy, add some premium for sponsor vs a front book bank deal…..fine?

Down the stack, the investor base gets thinner and, unfortunately for Fortress, very focused on questions like “how much is this portfolio worth exactly?” and “will this deal ever be called?”. Arranger Natixis ran an auction process for the class C and D notes. But the bids unearthed did not meet Fortress’s reserve levels, and Fortress retained these tranches. We wouldn’t be surprised to see them again some time.

Big picture, what’s going on here? Fortress has been deep in the legacy asset business for a long time, and given that it has owned this portfolio since 2017, the relevant fund might be getting towards the end of its lifespan. Market conditions are certainly strong, and Stresa might have got less focus in January….but equally, cashing out by year-end might have been a good look for the LPs. 

Secondary sales of portfolios like this are certainly possible, even if there hasn’t been much (the performing part of Barclays Italy is quite a different animal). But selling as a portfolio rather than a securitisation doesn’t necessarily add much to the potential audience. 

Hedge funds that would give up the time to look at this sort of portfolio would also look at a class D note with an ISIN. But if you structure a deal into bonds, there’s always a chance that some dumber money will show up and ease the exit.

If Fortress is indeed coming to the end of its fund life, how is it going to exit the rest of its portfolios? If there’s nobody showing up to hold the bag, maybe securitisation is not the way.

See you in 2024!