Our distressed debt and legal team are still looking at it, but the proximate cause is pretty obvious — even if there’s a solution short of nationalisation, it’s unlikely that the regulated and securitised Thames Water operating company is going to be upstreaming much in the ways of dividends to service the high yield and private placement holdco debt.

Beyond that, there’s an almighty fight over what went wrong, whose fault it was, and what role securitisation, in its whole business form, played in the current distress. I’m no fan of sewage in the water, but there are a few nuances to the debate which are being lost. 

I would highly recommend a read through the comments section on this FT Alphaville Post — an anonymous poster named “Securitisation” is patiently explaining WBS structures to all challengers and shedding some actual light on the situation. As of Thursday morning, he seems to have been invited to come above the line, so enjoy Tim Short’s “Seven Misconceptions about the Thames Water Crisis” post. 

The proximate cause of distress is kind of boring (everything about water companies should be boring!). 

All the big UK water companies finance their operating companies through RPI (Retail Price Index)-linked liabilities — even when they issued fixed rate debt they usually swap to RPI-linked. Other than the UK Debt Management Office, the watercos are the major source of inflation-linked bonds in the UK. 

For historical and market structure reasons, inflation trading, through either inflation swaps or index-linked bonds, is pretty much all linked to the RPI index, despite interminable consultations and considerations about trying to switch over to CPI (Consumer Price Index), the inflation index used by the Bank of England and increasingly large parts of the UK policymaking apparatus, crucially, pension calculations.

CPI usually comes in lower than RPI (the “wedge”), so the big Gilt investors have generally been against any change, and it’s been particularly far apart of late. That matters because in the latest round of regulatory discussions between the water companies and Ofwat, the regulator, the water companies were made to base price increases on CPI, and not on RPI. 

There are inflation basis swaps to be had to switch between CPI and RPI, but as far as I’m aware, the market is not huge (most inflation trading is based on RPI vs nominal), and it doesn’t look like the water companies rushed to hedge out this basis. Instead, they just ran the risk on a truly massive scale — enabled, of course, by the whole business securitisation structures which allowed heroic leverage levels on the regulated water company while still maintaining investment-grade ratings.

From first principles, this is not a dumb idea — water companies are heavily regulated monopolies that can, in theory, pass through all sorts of investment costs to customers through their bills. If investment is required, the regulator agrees it, the customers pay for it, the water companies are kind of neutral one way or the other. London customers will no doubt be familiar with the surcharge to pay for the £4bn “Super Sewer” down the middle of the Thames. 

Price increases should match cost increases, there should be minimum volatility in the business, so from first corporate finance principles, you should lever the hell of out it. Debt should be the cheapest form of capital, super secure whole business securitisations should be the cheapest form of debt, ship it in.

But it sets up a tension — Ofwat wants customers to have the lowest bills, water companies don’t really care one way or the other as long as they can pass on costs (but very much want to keep paying dividends out of their regulated structures), so there isn’t a clear constituency pushing for high levels of investment in water infrastructure. 

If the world changes, the inflation basis changes, or the decades of underinvestment become disgustingly clear, the structures don’t allow much flexibility. The regulator has limited powers outside the framework of the regulatory cycle; the current Thames crisis appears to have been precipitated by a fight over whether shareholders will put in more equity.

Also, let’s be clear, there has been some terrible behaviour. The combination of long-dated liabilities and massive appetite for inflation hedging makes the water companies very lucrative derivatives counterparties. 

Southern Water used to be owned by RBS (why???), and did a lot of trades with Fred the Shred’s rates trading desk; do we think they were all on competitive market terms? I’ve heard the word “piggy bank” bandied about.

As a aesthetic appreciator of financial structuring, I rather like WBS, and feel professionally disappointed that there isn’t much of it around these days (most of the WBS-worthy companies in the UK already did one in the early 2000s). 

But it does have clear flaws; what you gain in incremental leverage you lose in flexibility, and it’s probably better to have the option to renew and refresh capital structures every few years to match the environment prevailing at the time. Even the best structurers cannot envision all possible states of the world 30 years hence and bake those into deal documents; maybe a “once for all” debt issuance platform with ultra-long liabilities isn’t actually a great plan?

If there is to be a Thames Water nationalisation, by far the simplest and cheapest way is probably to keep the regulated WBS structure in place. Wipe out the shares (no tears for the Canadian civil servants and, uh, UK academics) and the holdco bonds, transfer ownership of the WBS, claim some management scalps and carry on as before.

The WBS bond docs kind of spell it out — there are two Change of Control clauses, one for TWUL (basically the WBS) and one for TWH (Thames Water Holdings), one of a layer of holdcos above it.

TWUL Change of Control is an event of default, TWH Change of Control just means you notify bondholders — so no unpleasant accelerations or make-wholes to jump through. This level is below where the Kemble Finance bond debt sits, so nationalisation at this level would leave the HY debt orphaned.

Does a CLO business need to make money?

You’d think the answer would be “obviously yes” — nobody is really in the investment banking industry for purely charitable purposes, despite the stubborn discount to book value across European banking. 

But that’s not necessarily how institutions are acting. There’s a relatively tight concentration of top shops who swallow the bulk of the mandates; success tends to beget success. If you do a lot of deals, you’re more connected to the market, and perhaps more able to place deals in difficult conditions.

Breaking into this magic circle is difficult, but not impossible — we’ve spilled much ink on the rise of Jefferies from a standing start to one of the Street’s top shops — but for every Jefferies, there’s a fair few also-rans, institutions which just didn’t quite get there. Even when there’s a tangible business in place, with a head of CLOs, some structured credit distribution, and a few junior structurers, doing one or two deals a year from that level of headcount isn’t super-profitable. 

But yet, there are plenty of shops who’ve been looking to get into the business. Société Générale and RBC have been publicly keen (both doing deals last year for CVC); it depends who you ask about HSBC. Mizuho has been around the business for a while; SMBC may be channelling capital this way. 

These efforts are unlikely to go full Jefferies and leap to the top of the table, but they may still be worthwhile for the broader franchise effects. A CLO operation tends to enhance loan trading profitability; although CLO warehouses can source collateral from anywhere, in practice the arranger’s traders tend to buy much of it. 

Having a ready-made source of demand, a list of loans to go and lift, means more flow and activity going through loan trading, more flow and more engagement means better pricing and better information, a better trading operation also means a better distribution capability for primary loans.

This is all very well at the bank level, but investment banks are rife with revenue attribution questions. The better an investment bank’s franchise works, the more cross-selling and cross-pollination between the different business units. But the more cross-pollination occurs, the more vicious the fights to claim the P&L.

Mortgages but worse

It’s an absurd fact that you have to take a forward rates view to get a house in much of Europe. I like to think I’m easily top 5% of the country in terms of how much time I spend looking at markets, but I don’t have strong conviction on whether UK swap rates are mispricing two-year or five-year rates. This is very much the argument made by Arjan Verbeek and all at Perenna, who want to get the long term fixed rate for life product off the ground in the UK.

Anyway, a way to make this bad situation worse is to add an element of FX risk as well! This was a popular product in eastern Europe before the financial crisis, with Swiss franc interest rates well below those prevailing in, say, Czech koruna and Polish zloty, and it’s basically been a slow burn disaster ever since.

But there was also a bit of it done in Italy, and there’s a packet up for sale as part of Barclays’ plans to exit its Italian resi exposures. The Barclays Swissie mortgages are synthetic or indexed — they’re euro-denominated but converted to Swiss francs — but they’re still a product with, um, character.

The vast bulk of the Barclays Italian portfolio, some €4.4bn, consists of performing Italian mortgages — but we understand that, as the bidding goes in to round two, there are no strategic buyers (i.e. Italian banks) in the running. It’s all securitisation-funded financial investors.

Barclays is providing some vendor finance, with a ready-baked securitisation structure, balance guaranteed swap, and an anchor ticket for the triple-A already in place, so we should at some point have a rather large and very cool transaction to look at. 

Italian RMBS is a fairly rare asset class these days, at least in distributed format; there are some massive retained shelves, but most Italian collateral coming out in public is ABS, either consumer loans or auto loans. 

But a deal of this size would need careful handling whatever the jurisidiction. Even with the senior spoken for, unless the deal disappears into the vast maw of Pimco, the mezz tranches represent significant tickets that will need some deftness of touch to place.

We don’t have any info about the synthetic Swissie portfolio, other than to note that this is some pretty weird collateral — presumably at this point it’s too late to unpick the currency element mortgage-by-mortgage, so perhaps the plan is to finance it in a Swissie-denominated structure? Maybe this is a step too far for a securitisation exit in public form. 

What you really need is a Swiss bank with a quality securitisation franchise and a penchant for structural complexity to get stuck in, but alas, where would you find such an institution?

Well done to the PayPal Four

I didn’t have time to dig in to the KKR-PayPal transaction much last week, beyond what was in the public domain, but congratulations to financing banks BNP Paribas, Citigroup, RBC Capital Markets and Société Générale for their involvement in a very cool and innovative transaction!

As a reminder, it was a €3bn forward flow facility securitising PayPal’s European buy-now-pay-later origination, with a capacity for $40bn-equivalent over the lifetime of the deal, across multiple European jurisdictions. 

I find it a little strange these banks weren’t disclosed at the outset (presumably KKR is pleased with its counterparties, and all the deal teams are taking an internal lap of honour?) but I don’t mind poking around a bit, and here we are. 

BNPP and Citi took the largest tickets, as I understand it, but the involvement of RBC is still worth noting — it’s not a huge player in European securitisation (despite the CLO buildout discussed in past publications), and this is a landmark deal, a statement of intent transaction. RBC is also hiring, so big things to come?

There was also another American bank around the situation, I understand, which didn’t end up lending — just showed up, cut prices for everyone else, and then left again. Maybe this institution deserves a special KKR shout out as well?

On structure, there’s quite a neat element baked in to incentivise good origination. BNPL originators have an incentive to make the whole origination process as painless as possible, meaning minimal checks at the point of sale, and maximum approvals to potential customers. PayPal might be the 800lb gorilla in this market, but it’s still quite a new business line without the years of performance data that you want in a securitisation.

So the purchase price for the receivables actually bakes in the credit performance experienced over the previous period. BNPL receivables pay 0%, so they are sold into the structure at a discount, and this discount depends on swap rates, plus a credit spread to give KKR a return and support the structure, plus a further discount reflecting the losses in the previous period. So if origination deteriorates, the next receivables will be bought from PayPal at a lower price, giving PayPal an incentive to keep up the standards. Should you wish to pore over this in more detail, take a look at PayPal’s 8K and the linked documents.

Sadly, these do not disclose the identity of the class B lender — I presume the investment banks above are in the class A, and class C is a KKR-linked entity — so there’s still a missing mezz fund I need to track down.

Also bidding for the PayPal portfolio were Pimco, Cerberus and Elliott, the latter of which is particularly interesting — Elliott’s activist position in PayPal’s equity prompted it to launch its buyback programme, management revamp and perhaps even to push towards the capital-lite model of selling out its loans. 

Equally, all of this activity might not have endeared Elliott to PayPal management, and a forward flow this size really does have to be a partnership. Onto the next deal though — as we hinted last week, Klarna is indeed in market, looking to sell on its UK BNPL receivables portfolio. A single jurisdiction deal is going to be technically easier to handle than PayPal’s cross-border effort (and perhaps more likely to come out in public), while Elliott already has an existing relationship with Klarna — according to a couple of sources, it bought the BNPL SRT trade Klarna priced last year.