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

There will be no Excess Spread next week. Next edition February 5.

Rock Out

Northern Rock has cast a shadow over roughly the whole of my career. The “run on the Rock” and nationalisation of the UK mortgage lender was one of the totemic moments of the financial crisis in the UK, and felled what was at the time the largest issuer of RMBS in Europe.

It gave rise to some black comedy, such as when the Guardian discovered master trusts. One of the pivotal moments for UK securitisation markets was the breach of the Rock’s Granite master trust’s non-asset trigger in November 2008, after two consecutive payment dates in which the seller’s share had fallen below its minimum level (because Northern Rock was no longer transferring new loans into the trust).

This was the first test of master trust technology, and broadly, it passed. The non-asset trigger sent the Granite deals into early amortisation, and payments passed through, with all bondholders of a given seniority receiving the same cashflows.

This made Granite bonds dramatically more liquid than anything else in European structured finance during the post-crisis period, since it no longer mattered which series of Granite was changing hands — one triple-B bond was much like another.

The marketing and sale of the portfolio in 2015 was another milestone. It wasn’t the first bad bank asset disposal in Europe, but it is arguably the largest (£13.3bn at the time).

Project Quasar, the sale of Banco Popular’s real estate and non-performing assets in 2017, and UniCredit’s Project FINO in 2018 were €30bn and €17.7bn respectively in gross book value, but were NPL sales that traded far below par, and in the case of Quasar, was structured as a 51% JV.

The sale of the Granite loans, though, attracted by far the most public scrutiny, with a National Audit Office report and public enquiry, and remain controversial, as some of the borrowers are so-called “mortgage prisoners”, who are stuck on high standard variable rates (the reversion rate) but unable to refinance away.

Cerberus won the bidding for the portfolio, known as Project Neptune, and has owned the Granite loans ever since, through multiple rounds of refinancing. There have been seven Towd Point Mortgage Funding Granite RMBS deals, plus Feldspar 2016-1, sponsored by Bawag (then Cerberus-owned).

But last week, Cerberus tried to sell, undertaking a major BWIC process which sought to clear out residual notes in Granite, Vantage (part of the old GE Money portfolio) and Auburn, the Capital Home Loans BTL back book. Altogether, control rights for around £2.7bn of legacy mortgage collateral by our reckoning.

Cerberus seems to have been moving away from these exposures for some time. It sold the origination platform of Capital Home Loans to Chetwood Financial in May 2024, and cleared out its front book —loans originated between relaunching the product in 2021 and closing it down in 2023 — in Edenbrook Mortgage Funding.

Bank of America bought the Auburn portfolios, which had languished in three Towd Point transactions which were past their call date, and resecuritised them in a rare BofA principal deal, Auburn 15, in May 2024. However, the best bid for the residual notes seemingly came not from a third party, but from Cerberus itself, which has subsequently reoffered them in the BWIC last week.

After all that, it seems like Cerberus still owns Granite — we hear the Vantange and Auburn residuals traded, but Granite didn’t meet the reserve.

This may not matter too much. Multiple fully-levered re-racks and interest through the structures will have ensured Cerberus has done alright from the deals, though it’s beyond my capabilities and inclination to try to backsolve the returns (please write in if you have!).

A decade of investment might suggest that Cerberus has one of its funds coming to the end of its life, and needing to return capital and realize its investments. But the optics don’t quite fit here.

Cerberus did most of its early investments through Cerberus European Residential Holdings BV (if you wish to read a few MPs grilling Cerberus president David Teitelbaum about the tax arrangements of this vehicle, click here).

But the last Granite and Vantage refinancing had flipped to Cerberus European Residential Holdings II, which sure looks like a new fund vehicle that might extend the lifespan of these investments.

In other Cerb-related news, we had the pleasure of talking to Jonathan Klein-Strandberg of Brocc Finance, the fintech which has been transformed into one of the largest NPL buyers in Europe via an infusion of Cerberus capital and the transfer of Intrum’s back book. The firm announced last week that it would buy Intrum out of the 35% it held in the original Intrum-Cerberus joint venture.

Find the interview here.

Is Europe bad at funding data centres?

Huw van Steenis, former banks analyst at Morgan Stanley turned globe-trotting finance maven (currently vice chair at Oliver Wyman), certainly thinks so.

In this FT thinkpiece, published presumably as Davos food-for-thought, he draws a direct line from the parlous state of European insurance and securitisation regulation to the parlous state of data centre funding in European capital markets.

This graph certainly looks bad.